PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246201 (2023) https://doi.org/10.1117/12.2669818
This PDF file contains the front matter associated with SPIE Proceedings Volume 12462, including the Title Page, Copyright information, Table of Contents, and Conference Committee Page.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246202 (2023) https://doi.org/10.1117/12.2660822
With the boom in 5G technology, the bandwidth of communications is increasing while the coverage area of base stations is getting smaller and smaller, making it necessary to have more base stations to cover the same area. On the other hand, the variety of base stations and antennas is also increasing. This complicates the problem of site planning for 5G communication base stations. The core of the site selection problem is to select a certain number of coordinate points for the weak coverage areas of the existing network according to the coverage of the existing network so that the coverage of the weak coverage areas of the existing network can be solved by building new base stations on these coordinate points. Therefore, this proposes a 5G base station planning model based on the idea of the binary mask, combining differential evolution algorithm and Monte Carlo simulation to fully consider the correlation and synergy between new 5G base stations and existing base stations. According to the experimental results, 186 new micro base stations and 1930 macro base stations were built, and the coverage rate of the new base stations in the region reached 91.2% of the total service volume of the weak coverage points. This shows that the method proposed in this paper can effectively solve the problem of siting 5G communication base stations and achieve the rational utilization of urban spatial site resources and the minimization of economic costs.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246203 (2023) https://doi.org/10.1117/12.2660807
Heterogeneous accelerators play a crucial role in improving computer performance. General-purpose computers reduce the frequent communication between traditional accelerators with separate memory and the host computer through fast communication links. Some high-speed devices such as supercomputers integrate the accelerator and CPU on one chip, and the shared memory is managed by the operating system, which shifts the performance bottleneck from data acquisition to accelerator addressing. Existing memory management mechanisms typically reserve contiguous physical memory locally for peripherals for efficient direct memory access. However, in large computer systems with multiple memory nodes, the accelerator's memory access behavior is limited by the local memory capacity. The difficulty of addressing accelerators across nodes prevents computers from maximizing the benefits of massive memory. This paper proposes a contiguous memory management mechanism for a large-scale CPU-accelerator hybrid architecture (CLMalloc) to simultaneously support the different types of memory requirements of CPU and accelerator programs. In simulation experiments, CLMalloc achieves similar (or even better) performance to the system functions malloc/free. Compared with the DMA-based baseline, the space utilization of CLMalloc is increased by 2×, and the latency is reduced by 80% to 90%.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246204 (2023) https://doi.org/10.1117/12.2660961
As the control center of wireless sensor network, gateway plays a key role in the communication between internal network and external network. The wireless sensor network gateway mainly integrates the data of each sensor node and transmits the data to the remote terminal through wireless communication and Ethernet, and then performs corresponding processing according to the instructions issued by the monitoring center. However, the large amount of data transmission of gateway nodes in wireless sensor networks leads to the problem of high CPU occupancy, which has a negative impact on the resource allocation and utilization of the gateway. To this end, a wireless sensor network gateway node based on micro service architecture is designed. The hardware part selects the CC2530 chip integrated with 8051 CPU core and starts the gateway flash burning tool. The software part extracts the characteristics of wireless sensor networks and adjusts the layout of gateway and relay nodes based on the micro service architecture. Combined with the means of multi hop routing, set the grid spacing in the space, design a new address mapping layer according to the unidirectional transmission channel of communication data, and set the data communication standard of software application layer. The test results show that the average CPU occupation percentage of the designed WSN gateway node is 47.54%, which indicates that the designed WSN gateway node has more comprehensive functions after integrating the micro service architecture.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246205 (2023) https://doi.org/10.1117/12.2660793
The application of digital image processing technology in log volume gauge count greatly promoted the progress of forestry production to intelligent automation direction. However, based on digital image processing of intelligent log ruler algorithm for accurate ruler, the first premise is to obtain high-definition log end image, under the same log ruler algorithm, image clarity determines the final detection effect of the algorithm system, high-definition picture can improve the accuracy of algorithm recognition. In the natural environment, how to obtain the high-quality image of the log face effectively without changing the resolution of the camera is a problem. This paper presents a method based on the combination of laser ranging sensor and Yolov3 target detection algorithm, and an ZYNQ embedded system for automatic acquisition of higher quality log end face image is designed. Finally, the images obtained by the system are processed into the dense wood detection segmentation algorithm to obtain the log recognition results. Compared with the traditional image acquisition method, the number of logs that can be recognized by the images collected by the system designed in this paper increases by 36.6%. The experimental results show that, the system can obtain clear and higher quality target images in complex environment background and different illumination intensity. The problem of how to obtain high quality log end face image without changing camera resolution in natural environment is solved successfully.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Luoning Gan, Jinjia Ji, Shangping Kong, Zhaojun Yang
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246206 (2023) https://doi.org/10.1117/12.2660786
Circularly-polarized microstrip antenna is widely used in communication fields. However, conventional microstrip antenna is limited by its bandwidth and axial ratio characteristic. As a result, a 2×2 slot-fed circularly-polarized double-layer microstrip antenna array is proposed in this paper. The design of double-layer structure and corner-cut patch is applied to improve circularly-polarized performance. Impedance matching is well designed by H-shaped slot and microstrip line. The form of rotation array is used to decrease the axial ratio of the array. The simulated bandwidth for VSWR lower than 2.0 of the sub-array is 37.0% in the frequency range of 2.4GHz-3.5GHz, and 10.3dBi gain is obtained at the design frequency. Results obtained based on the proposed analysis can be used in the design of wideband circularly-polarized antenna.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Shanhong Yin, Haiming Zhao, Yang Bi, Qiqi Tang, Ying Wang
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246207 (2023) https://doi.org/10.1117/12.2660910
Multi-channel Active Noise Control (ANC) system can effectively eliminate low-frequency noise in large scenes, received widespread attention in recent years. First, we derived the multi-channel ANC algorithm, namely the multi-channel Filtered x Least Mean Squared (FxLMS) algorithm. Based on the multi-channel FxLMS algorithm, we analyzed the effect of the step-size parameter on the multi-channel ANC system. The simulation results show that increasing the step-size parameter can effectively improve the low-frequency noise cancellation speed. However, the system stability decreases dramatically when the step-size parameter larger than the threshold. Meanwhile, when the value of step-size is larger than the threshold, the multi-channel ANC system is very sensitive. A small change will have a drastic impact on the multi-channel ANC system.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246208 (2023) https://doi.org/10.1117/12.2660932
The fifth-generation mobile communication (5G) has the advantages of high bandwidth, low time delay and low power consumption, and it can play an important role in all aspects of power transmission, substation, power distribution and electricity consumption, effectively making up for the disadvantages of traditional optical fiber communication and profoundly transforming the power communication network. However, the information security problems brought by the application of 5G technology are gradually becoming prominent. Based on this, a quantum key distribution strategy based on the quality of service (QoS) is proposed to improve the confidentiality of 5G electric power private network application scheme. Finally, the feasibility of quantum communication application in power dispatching system is verified. The test results show that the service quality of 5G private network meets the communication requirements of power grid services, the actual transmission delay of the power grid simulation dispatching data is about 1s, and there is no packet loss.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246209 (2023) https://doi.org/10.1117/12.2660824
Aiming at the problems of background interference and incorrect angles in the images collected by inspection robots, a computer vision-based automatic inspection and reading system for pointer-type instruments is proposed, which can automatically obtain pointer readings on the basis of instrument image detection and correction. First, use the Centernet algorithm to detect the target of the pointer instrument image, and cut out the instrument image with the background removed according to the detected position information; then, perform key point detection, select a pair of symmetrical key points to rotate the instrument using affine transformation correction, and then select two pairs of symmetrical key points through template matching to correct the inclination of the meter using perspective transformation; finally, complete the reading of the meter by using Otsu segmentation and Hough transform circle and line detection. The experimental results show that the proportion of images to be corrected is 93%, the average error rate of the corrected instrument image is reduced by 10.19%, and the average accuracy of readings reaches 97.02%, which can meet the practical application.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620A (2023) https://doi.org/10.1117/12.2661010
In order to improve the independent innovation of embedded system experiments, an open experimental teaching platform that integrates experimental resources is built to provide comprehensive services for students to practice. Combining the engineering application background of the automation discipline, this paper discusses the construction scheme of a new experimental platform for embedded systems, explores the intelligent open laboratory management mechanism, and forms an experimental teaching reform idea characterized by informatization, service, and progressiveness, so as to stimulate students' potential for embedded system learning.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620B (2023) https://doi.org/10.1117/12.2662599
Frequency hopping synchronization is the premise to ensure the normal operation of the DS/FH communication system, and the frequency hopping synchronization process is the main object of communication interference at present. Based on the basic principles of frequency hopping synchronization, the principle and performance of the traditional synchronization method of DS/FH communication system are analyzed. First, the basic principles of synchronization are introduced; then the traditional self-synchronization method is introduced, its advantages and disadvantages are analyzed, and its performance in low signal-to-noise ratio environment is analyzed and derived in detail.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620C (2023) https://doi.org/10.1117/12.2660796
In order to solve the problem of doppler shift caused by high-speed movement of maglev train and high carrier frequency of train-ground wireless communication system, this paper studies the Doppler shift estimation algorithm of SC-FDE system and proposes a two-stage estimation algorithm based on the excellent correlation characteristics of gray sequence. Firstly, the coarse estimation and compensation of doppler shift are carried out by using the gray sequence in the preamble training sequence. Then the fine estimation and compensation of Doppler shift are carried out by using the gray sequence inserted in the data block. Simulation analysis and experimental verification are carried out. The results show that the proposed algorithm can significantly improve the throughput performance of train-ground wireless communication system in high-speed mobile environment. The algorithm has high estimation accuracy and is easy to be implemented by hardware. It can be extended to SC-FDE systems in other large doppler scenarios.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620D (2023) https://doi.org/10.1117/12.2661012
With the development of science and technology and the arrival of the era of big data, artificial intelligence is constantly developing, and related theories are maturing. One of the important branches of artificial intelligence, that is, natural language processing technology, is also constantly developing. In the Internet era, there are more and more text information. Therefore, we should seek a good way to effectively obtain and manage this information. Through experiments, this paper tests the feasibility and efficiency of using natural language processing technology for text classification and draws a conclusion that natural language processing technology is efficiently applied to text classification.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620E (2023) https://doi.org/10.1117/12.2661162
When dealing with complex tasks, such as robots imitating human actions and autonomous vehicles driving in urban environments, it can be difficult to determine the reward function of the Markov decision-making process. In contrast to reinforcement learning, Inverse Reinforcement Learning (IRL) can infer the reward function through the finite state space and the linear combination of reward features, given the optimal strategy or expert trajectory. At present, IRL has many challenges, such as ambiguity, large computation and generalization. As part of this paper, we discuss existing research related to these issues, describe the existing traditional IRL methods, implement the model, and then propose future direction for further research.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620F (2023) https://doi.org/10.1117/12.2660983
The use of mobile phone cameras to capture and save reports shared on the screen at the meeting has become a major way for researchers to obtain information. However, a large number of screen-shot images obtained in this way have a lot of redundant information, and it takes a lot of time and energy to organize and save them later, therefore, it has become a realistic requirement to develop a software tool that can quickly extract the subject content of screen-shot images and realize automatic batch segmentation and compression storage. We use the self-made screen-shot image dataset SSD (Screen-Shot Document Dataset) composed of more than 1000 images for training, based on the improved U2 -Net network model to achieve automatic segmentation of screen-shot image subject area, combined with OTSU binarization, Canny edge detection and Hough Transform to extract the quadrilateral boundary of the subject area, and implement an Android-based screen-shot document automatic extraction system. The system can be used to automatically extract and save screen-shot document to PDF format in real time or at a later stage, significantly improving the efficiency of information collection and storage for researchers, reducing the financial and time loss caused by researchers not being able to find backups when they need information and discuss key content for meetings, and saving storage space on mobile phones.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620G (2023) https://doi.org/10.1117/12.2661048
Starting from the time series of factors, the level of analysis time, data types, and forecasting accuracy, based on the characteristics of the data sequence to be analyzed. ARMA model to predict sequence requirements must be stable, that factors in the time range of the study subjects must be subjected to the same requirements. If the given sequence is not stationary sequence, you must do on a given sequence of preprocess, smoothing it, then by ARMA model. Example is analyzed by Eview software, the validity of the model is verified.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Yujin Qin, Shuxia Wang, Qiang Zhang, Yao Cheng, Jiaxu Huang, Weiping He
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620H (2023) https://doi.org/10.1117/12.2660940
In this article, we demonstrate an implementation on Microsoft HoloLens, deep learning supported in the context of object detection. The main aim of the training system is to create the more accurate object detection model for augmented reality using deep learning models for image recognition directly on the HoloLens 2. In terms of the object detection approach, a deep learning model called YOLOv5 has been used for the implementation of this system. This article uses the Windows ML API to implement machine learning in augmented reality applications. A simple and easy method of drawing lines between specified 2D coordinates on a canvas is proposed. The module division and development steps of the development of augmented reality training system are given. Our system provides the annotation of augmented object detected and its bounding box via HoloLens. It allows to detect the new object in a few milliseconds. Preliminary results show a great rate of object detection and reasonable detection time.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Ning Wu, Hui Gao, Peng Wang, Xiaoyan Li, Zhigang Lv
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620I (2023) https://doi.org/10.1117/12.2660942
In crowded and complex scenes, it is easy to cause problems such as poor human pose estimation and low-key point positioning accuracy. In this paper, a high-resolution human pose estimation algorithm based on position awareness was proposed. The algorithm introduced the coordination attention (CA) in the feature extraction module, which realized the accurate acquisition of the spatial position information of key points, finally improved the human pose. Estimate the detection accuracy of the algorithm. The AP value of the improved algorithm was 76.5%, which was 2.1% higher than the original algorithm, the AP50 was increased by about 3.1%, and the AP75 was increased by about 2.8%. The experimental results showed that the proposed algorithm could effectively improve the detection performance in crowded and complex backgrounds and had higher detection accuracy.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Zongyang Wang, Yingchun Zhong, Heer Huang, Zhiyong Luo, Huiqing He, Bo Wang
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620J (2023) https://doi.org/10.1117/12.2661165
When the Unmanned Aerial Vehicle (UAV) is guided by the navigation system that only relies on Ultra Wide Band (UWB), the dynamic error of UAV is too large to avoid the obstacles. Additionally, the navigation system needs eight positioning base stations on ground, which is too expensive to widely use. This paper proposes a navigation system which fuses UWB information and height data in order to reduce the dynamic error and number of positioning stations. First, a height-finding radar is mounted on the UAV to obtain height data; second, a new positioning algorithm by fusing UWB information and height data is designed, and only three UWB ground base stations are needed for positioning. The experimental results show that this navigation system is able to guide the inspection UAV to land precisely at the landing area with a diameter of 60cm, at the same time, compared with the navigation system only by UWB, the average dynamic error is reduced by 10%.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Beibei Xu, Valeriy Kuzminykh, Shiwei Zhu, Junfeng Yu, Mingjun Zhang, Sisi Li, Dmytro Lande
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620K (2023) https://doi.org/10.1117/12.2660850
The deep mining of data and the application of technology integration are the main characteristics of a smart society. This paper is to explore the application of multi-technology integration in library smart construction in a smart society, and study the library construction route driven by multi-technology integration. It promotes the transformation of technology empowerment to service empowerment, and provides a reference for the sustainable development of libraries and the construction of a national smart society. This paper analyzes the connotation and research status of multi-technology integration in a smart society. Based on this, it studies the important role of multi-technology integration in promoting the development of library service innovation, and analyzes the specific application of key technologies, such as 5G, IOT, big data, artificial intelligence, block chain and other technologies. Finally, it proposes the route of smart library construction driven by multi-technology integration.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Ke-ke Liu, Guo-wang Gao, Fei Wang, Dan Wu, Zhao-xue Wu, Yu Gong
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620L (2023) https://doi.org/10.1117/12.2660957
Crude oil water content is an important technical indicator in oil extraction, transportation and oil trading. Real-time online testing of crude oil water content is extremely important in estimating crude oil production and evaluating the extraction value of oil wells. At present, most of the wells at home and abroad are in the middle and late stage of development, it is difficult and inaccurate to measure under the high-water content condition of crude oil, so it is necessary to adopt new detection means to improve the detection accuracy. In this paper, a study on the method of water content measurement using infrared spectroscopy was carried out. This study used S-G smoothing and normalization as the method of data pre-processing, selected the characteristic wavelengths using the continuous projection method (SPA) with a root mean square error of 4.4702, and then used partial least squares (PLS) to establish a water content detection model, and obtained a prediction root mean square error of 9.7131 and a correlation coefficient of 0.98527, which obtained a good accuracy. The feasibility of using spectroscopic detection technology to measure the water content of crude oil was demonstrated, providing a new method for oil extraction exploration and production processing.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620M (2023) https://doi.org/10.1117/12.2660767
According to the network text analysis method, the tourists' perception of the image they want to know about the Glorious Orient based on the user comments on the Ctrip website are collected through the data collection software "Octopus Collector", and the online text is analyzed with the help of ROST CM6.0 text analysis software. High-frequency word analysis, semantic network analysis and sentiment analysis to study tourists' image perception of tourist destinations. The results show that: (1) the Glorious Orient theme park attracts tourists by combining high-tech and stimulating experience items; (2) The semantic network diagrams are mainly positive or neutral words, indicating that tourists are opposite each other. The overall image perception of the Glorious Orient is better; (3) The sentiment analysis shows that tourists are highly satisfied with the Glorious Orient, but some problems that need to be paid attention to are also summarized in the negative emotions.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620N (2023) https://doi.org/10.1117/12.2660953
Video smoke detection benefits life safety and environment protection, its early warning is of great importance. In response of many disadvantages of traditional smoke detectors, a method of video smoke detection based on dynamic, color and texture features are proposed. Firstly, the motion area is extracted through improved Vibe algorithm. Then, the suspected area corresponding to smoke is identified in CIELAB color space and segmented using a color filtering method. Finally, uniform local binary mode and gray level cooccurrence matrix are extracted from the image within suspected area and used to form the input vector of machine learning classifier for recognizing smoke. The classifier is tested with 400 images, and the results show that the detection system based on random forest algorithm has better performance, the selected smoke features have high recognition accuracy.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620O (2023) https://doi.org/10.1117/12.2661159
Aiming at research on the coupled control of ejection seat attitude rocket under unfavorable attitude, we provide multi-parameter coupling mechanism of a dual-channel controlled ejection seat. We use a dual-channel (roll, pitch) controlled ejection seat as the research object. On the basis of deriving the mathematical model of its ejection motion, combining with the working principle of the ejection seat, and according to the preliminary drafted various control modes, together with both qualitative analysis numerical simulation methods, the multi-parameter coupling mechanism of the seat with specific roll and pitch ejection states is studied, and the research shows the dual channel controlled has a beneficial effect on the trajectory of the ejection process.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620P (2023) https://doi.org/10.1117/12.2661161
Industrial digital twin is the key support for the transformation of industry to intelligence. The realization of digital twin relies on the integration of whole-process data, which can provide real-time and intelligent decision-making optimization for production management. This paper proposes a digital twin data fusion method based on semantic data dictionary. Firstly, a modeling method of domain data dictionary is designed, and a semantic method is introduced to realize the construction of semantic data dictionary. Then, based on the semantic data dictionary, a semantic similarity calculation method based on multivariate distance weighting is proposed. Finally, the algorithm is tested and verified. The experimental results show that the method in this paper has a significant effect on automatic data fusion in the construction of industrial digital twins.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620Q (2023) https://doi.org/10.1117/12.2660821
University ranking has a positive impact on making rational use of educational resources and promoting deep-seated development of higher education and colleges and universities. This paper establishes two models to rank universities in the Yangtze River Delta: Model A uses AHP to analyze 8 main evaluation indicators; model B adopts the entropy weight method with reference to the data of the evaluation authority. In the results of the models, the ranking of top 20 universities in the Yangtze River Delta region is basically same. Comparing the two models, model B refers to the rankings of authoritative institutions, has obvious advantages over model A in terms of solution scale and model complexity, so the calculation results of model B are finally adopted. The models can not only rank the comprehensive strength of colleges and universities, but also be used to optimize the college applications and make better choices for the healthy progress of higher education.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620R (2023) https://doi.org/10.1117/12.2661031
Urban greenway planning is not only an important part of the construction of ecological cities and urban forest ecological network systems, but also a necessary means and an effective way to build urban ecological gardens and ecological green space systems. The link of the city plays a vital role in protecting the cultural and ecological civilization of the old city, and protecting the ecological environment, natural landscape, cultural relics and historical sites and local features along the line. This study selects the old urban area of Jinan City as the research area, uses GIS (geographic information system) remote sensing interpretation, conducts data analysis, conducts comprehensive performance evaluation of the area, selects the best connection paths and landscape nodes, and establishes the gravity model of urban greenways.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Communication Signal Processing and Software Design
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620S (2023) https://doi.org/10.1117/12.2660931
In recent years, ocean pollution is becoming more and more serious, and plastic waste is entering the ocean and damaging the environment while causing the death of a large number of marine animals. Under the theme of marine animal protection, the game named “New Life Out of Plastic” is designed from the three levels of emotional design theory. In which the visual style is designed at the instinctive level, the interaction and logic of the game is designed at the behavioral level, and the metaphorical approach leads people to reflect on the harm caused by plastic waste to marine animals at the reflective level. In the end, the final implementation is by computer language. The author hopes to promote the concept of marine animal protection and raise people's awareness of protecting marine animals by designing games.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620T (2023) https://doi.org/10.1117/12.2660773
Micro motion detection is an essential research direction in urban underground space and energy prospection. Aiming at the shortcomings of the traditional system for seismic exploration, the upper computer software of the micro motion signal acquisition system based on TCP/IP protocol is designed, which can be distributed offline acquisition and control. By formulating the communication protocol of the application layer and adding the heartbeat packet mechanism, the stable connection with the collector, data transmission and status monitoring is realized. Meanwhile the received data are stored and visualized, and the project management and task configuration are implemented by using the file reading and writing function of QT to achieve the effect of offline control. Through the test, the results show that each function can get the rapid and correct response of the lower computer, and the upper computer software of the system meets the design requirements. It has the advantages of high acquisition efficiency, reliability and convenience which has strong practicability.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620U (2023) https://doi.org/10.1117/12.2660965
Cookies are essential to the modern internet. People use cookies and other tracking technologies to integrate the browsing experience of websites, present personalized content and targeted advertising, understand the origin of their audience, and analyze web traffic. In most cases, by clicking “yes” or “I accept,” people will agree to the use of tracking technologies and cookies. An investigation of cookies and cybersecurity is necessary and imperative. This thesis is going to thoroughly investigate the function that cookies present within the internet world, the actual challenges it poses to cybersecurity and its existence. In order to accomplish the investigation, the thesis analyzes the pros and cons of cookies on cybersecurity, the pros and cons of cookies and cybersecurity through consumer feedback. The findings indicate that there are gaps in privacy. A critical set of effective actions have emerged for organizations that are looking for methods to address the consumer protections and data security requirements. From security perspective, organizations ought to be aware of what data they actually need to serve consumers.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620V (2023) https://doi.org/10.1117/12.2661176
Aiming at the problem that the kinematics feature dataset of different traffic flow, average speed, acceleration, and traffic time are affected by high-dimensional, irrelevant, and redundant factors, and the kinematics feature dataset is a multi-objective, multi-constrained and complex nonlinear optimization system, the improved Multi-Objective evolutionary Soft Subspace Clustering algorithm (iMOSSC) is proposed to mine the micro-stroke segments with different kinematic characteristics and realize data classification. The algorithm uses iNSGA-II as the base algorithm and performs local search operator and repair operator operation in the feature space to accelerate convergence and improve the accuracy of the solution. The feasibility and effectiveness of the algorithm are verified by 12 sets of UCI standard dataset. The classified kinematics characteristic data is used to construct the Xi'an urban road trajectory database. Compared with the iMWK-HD algorithm in the collected kinematics feature data of circulation condition, the feature importance degree of the iMOSSC algorithm is more reasonable, the stability is better, the accuracy is higher, and the classification effectiveness is more obvious than the iMWK-HD algorithm. The excavated kinematics data is imported into the Optimumlap simulation software to construct the actual road circulation condition trajectory database. Based on the ADVISOR commercial software platform for the simulation module.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620W (2023) https://doi.org/10.1117/12.2660923
In mobile ad-hoc networks, OFDM technology is widely used, and in order to improve system capacity and reduce the problem of multiple retransmissions caused by collisions, non-orthogonal multiple access technology is usually used, which leads to non-orthogonal aliasing at the receiver of the OFDM system. In order to solve the aliasing signal separation problem of OFDM system, SIC(successful interference cancellation) technology, a non-orthogonal demodulation method based on OFDM system is proposed in this paper, according to the channel error vector magnitude (EVM), selecting sub-carrier blocks with better channel conditions for transmission, focusing on the use of subcarrier block SIC solution when there is non-orthogonal aliasing in the power domain at the receiver under strict synchronization. The simulation results show that, in this case, the aliasing signal separation technology for OFDM system adopted in this paper can not only improve the system capacity and improve the bit error performance, but also easy to implement in engineering.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620X (2023) https://doi.org/10.1117/12.2660781
A campus network with a certain scale will be divided into multiple functional areas according to different business functions, and these functional areas are distinguished by different VLANs. In order to ensure the reliability of each VLAN connection and achieve efficient collaboration, this paper proposes link backup and load balancing technology of route design to realize the construction and operation and maintenance of reliable and efficient Park networks under multi VLANs.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620Y (2023) https://doi.org/10.1117/12.2660845
At present, the detection of dim targets has become an important research direction in the field of infrared imaging, as the traditional single-lens dim target detection imaging system is limited by cost, size, and performance, which cannot support the requirements of many targets detection tasks. With the development of scientific and technological research, infrared array cameras have achieved improved detection capabilities through more imaging units, which are expected to break through the shortcomings and limitations of traditional single-lens imaging systems. To achieve this goal, this paper proposes a refocusing-based infrared array camera dim target signal-to-noise ratio enhancement method, which can effectively improve the detection distance of the imaging system for dim targets while achieving the improvement of the signal-to-noise ratio. Through experimental verification, the method achieves the expected results and can be applied to relevant subsequent processing tasks.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124620Z (2023) https://doi.org/10.1117/12.2660838
The control software running on the traditional single-core processor cannot meet the requirements of multi-task parallel computing and high real-time performance. An aircraft control software architecture design based on heterogeneous multi-core processor is proposed in this paper. The control software functions are divided into multiple tasks, which are decoupled and distributed on different CPU (Central Processing Unit) to achieve parallel computing and meet requirements of high performance in complex tasks. In this paper, Asymmetric multi-processing (AMP) is taken as an example, and inter-core communication mode based on shared memory is adopted to complete the design of aircraft control software architecture.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246210 (2023) https://doi.org/10.1117/12.2660992
Community topic classification is the basis of hot topic discovery. Existing graph models ignore the importance of key information to the text when performing text classification and increase the influence of irrelevant data. To address these problems, we propose a community topic classification model DGAT that incorporates key information as well as information about the topic itself. An integrated algorithm is proposed to extract keywords to avoid the problem of inaccurate keyword extraction. Then a composite complex network model containing both topic and keyword nodes is built. Finally, the graph attention mechanism is used to update node features and incorporate semantic-level attention to learn the effect of different graph structures on the current node classification. An example validation on the Qingdao community topic dataset proves the effectiveness of the method and outperforms the baseline models.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246211 (2023) https://doi.org/10.1117/12.2660989
Along with the deepening of human exploration of the sea, the demand for underwater vehicle is increasing day by day, and the quality of underwater vehicle needs to be further improved. In this paper, we propose a management model based on process nodes and carry out information system design based on this model to realize system function design, system architecture design, hardware design and software design, and provide a basis for the development of the whole process management system for underwater vehicle assembly and commissioning.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246212 (2023) https://doi.org/10.1117/12.2660972
In order to meet the conversion requirements between various types of transmission lines in microwave systems, an ultra-broadband single-ridge waveguide-coaxial converter spanning X and Ku bands is designed in this paper by using a single-ridge waveguide plus a stepped impedance converter. The operating bandwidth is 7.31Ghz-15GHz, and the return loss is better than -20dB. The design uses coaxial probe back-feeding, which is conducive to the cascading of the system, and also sets the tuning screw to facilitate the subsequent adjustment. The design provides a new idea for an ultra-broadband waveguide coaxial line converter across frequency bands.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246213 (2023) https://doi.org/10.1117/12.2660920
Aiming at the problem of seismic data sharing occlusion, a seismic information management system is designed. Using Eclipse development platform, B/S architecture based on Internet technology (Browser/Server structure), based on Java language and MySQL database, information technology, digital technology as the core, according to the work needs to set six modules, so as to achieve the purpose of data sharing. The system fully collects, excavates and integrates earthquake information resources to form a social security system, in order to grasp earthquake information timelier. By registering and logging in, users can quickly view the earthquake information and give feedback. It is also very convenient for people to gather information about earthquakes to improve their knowledge about the disaster reduction and prevention. Doing so will help them become more aware of the various steps involved in the planning and implementation of a disaster.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246214 (2023) https://doi.org/10.1117/12.2660791
With the rapid development of artificial intelligence and big data, informatization has become an irresistible trend of the times, so how to acquire and process geographic information has become one of the most important strategic resources. Named entity recognition is also the task of sequence labeling. Aiming at the problem of accurate identification of place names and addresses in geographic information, this paper proposes a method to first classify and segment place names and addresses, and then label them through a bidirectional cyclic long-short-term neural network and a conditional random field model. It solves the problem of classification and recognition of place names and addresses, and realizes the accurate identification of Chinese place names. The research results based on word segmentation and labeling and recognition through deep learning neural network model show that not only accurate word segmentation can be completed, but also place-name addresses can be accurately identified. The research results can be applied to the semantic measurement of placename addresses and the placename address labeling database. construction has great practical significance.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246215 (2023) https://doi.org/10.1117/12.2662564
AES has replaced DES to become a widely used encryption algorithm since 2000s. This paper may insight the basic theory of AES and safety analysis. In this paper, the whole process of AES encryption is described. The procedures of plaintext processing firstly, substituting bytes secondly, then shifting rows, mixing columns and adding round keys. AES-128 is taken as the study example in this paper to introduce the features and operating steps of AES. The construction and use of a simple S-box is also mentioned to help understand the procedure of substituting bytes. Safety analysis of AES is also taken into consideration to test AES’s resistance against different kinds of attacks. The results show that AES is free from brute force attack with time security analysis. AES with 128 or more bits of key length can resist square attacks according to reviews on research. A way of differential cryptanalysis attack with concrete operating steps is introduced as a potential attack method against AES encryption standard. The paper also casts view on an improved AES algorithm to increase efficiency and security proposed by other research.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246216 (2023) https://doi.org/10.1117/12.2661094
The purpose of blockchain technology is to solve the trust problem between people or institutions and make the communication data and network communication of the Internet. In the past, cryptography lost money. Passwords are used to protect data, and the cost is relatively high. But with blockchain, cryptography becomes valuable. The formation of blockchain has made new contributions to cryptography and done something we could not do in the past. With blockchain, cryptography is "valuable". In fact, there are many cryptographic primitives used in blockchain, such as hash, digital signature, and etc. Moreover, digital signature not only uses standard digital signature, but also uses ring signature, connectable ring signature, one-time signature, borromer ring signature, multi signature, homomorphic encryption, homomorphic commitment, accumulator, zero knowledge proof, etc. As well as the recently popular password signature toss. As mentioned above, the popularity of blockchain technology will completely break the centralized pattern, indicating the advent of a new era in the future - Web3.0. This paper focuses on the application of encryption technology in blockchain and expounds in detail the applications of such as hash function and ring signature in blockchain. This study analyzes the application of cryptography in blockchain and discusses to the development of encryption technology in the future.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246217 (2023) https://doi.org/10.1117/12.2660780
In this paper, under the background of big data era, the theory and operation methods of data analysis and visualization and their important value in the field of data science are described, and the relationship between them is necessary. Besides, the convenience and shortcomings of using python as a tool for data analysis are analyzed. Finally, the development trend of data analysis and visualization is prospected.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246218 (2023) https://doi.org/10.1117/12.2660775
At present, artificial intelligence has become a hot topic, and the development of its related fields is also developing rapidly, among which visual ranging and image processing are particularly important for the development of artificial intelligence. Now there are many ranging methods are not fast enough, and measurement accuracy is not high, leading to the resulting estimates there are large deviation distance and the actual distance, such as unmanned vehicle, unmanned aircraft in operation process, cannot be accurately driving distance and obstacle avoidance, and existing deviations in robot grab, which cause personnel life safety is threatened, economic property damage. In order to solve the above related problems, this paper uses SIFT algorithm and ORB algorithm to extract feature points, and then through BFmatcher, FlannBasedMatcher, KnnMatch to match, finally get the corresponding distance, from the algorithm and accuracy of the two aspects of relevant research. It is concluded from the experimental measurement that FlannBasedMatcher takes into account both speed and accuracy after SIFT algorithm extraction, while ORB algorithm is faster than SIFT algorithm.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Shaokang Duan, Yun Zhou, Linhan Li, Lin Li, Dingwen Kang
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246219 (2023) https://doi.org/10.1117/12.2660970
A large span ultra-broadband microstrip planar structured diplexer are presented. Firstly, this paper describes the general steps for designing filters based on the synthesis of N+2 coupling matrices, based on microstrip lines. Secondly, this paper implements a dual-way frequency combing of C and X band satellite downlink signal based on a cross-finger filter structure and a microstrip T-junction. Finally, the diplexer was tested for practical processing. The diplexer operates in the frequency range of 3.7GHz-4.2GHz and 7.25GHz-7.75GHz, with in-band insertion loss better than 1.6dB and 2.6dB respectively, return loss better than 15.6dB and 15dB respectively, and passband isolation greater than 60dB. The test results are consistent with the simulation results and meet the design specifications.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621A (2023) https://doi.org/10.1117/12.2660783
This paper describes a scheme for the ECC Error Correction IP. The common circuit is extracted based on the traditional BCH (Bose, Ray-Chaudhuri, Hocquenghem) codes innovatively, which make the logic gate requirement reduced by 33%; at the same time, the interleaving idea is used for multi-channel ECC error correction, which can correct 4-bit adjacent errors at most; the p-channel parallel chien search circuit is used to replace the serial search circuit to reduce the error correction delay; the time-sharing error correction makes the reading and writing transparent to ECC error correction process. The simulation and tape show that the design can effectively ensure the correctness of DRAM internal data without affecting its access speed.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621B (2023) https://doi.org/10.1117/12.2661046
At the present stage, the informatization and intellectualization of the medical industry have greatly increased people's attention and developed medical informatization, but there are still problems such as inconsistent data and ineffective protection of patients' privacy. In terms of privacy protection of medical data storage and sharing, this paper first uses CiteSpace visual knowledge map analysis to learn the current research hotspots in the field of medical informatization. This paper addresses the problems of access barriers and information sharing and circulation among medical institutions, and protect patients' privacy data. Therefore, this paper designs a medical data privacy protection scheme based on the blockchain technology. Through the analysis of the practice results, it is verified that this scheme can achieve safe and reliable storage and sharing under the premise of effective protection of the patient's medical data privacy.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621C (2023) https://doi.org/10.1117/12.2661004
In terms of 3D target detection, a voxel-based detection method is proposed: voxel index R-CNN, aiming at the balance between the accuracy of detection results and detection efficiency. In order to improve the timeliness of target detection, this paper proposes a voxel index query method, which uses the index difference constraint to reduce the computational loss of the query process for quantified spatial voxels. On this basis, a voxel feature extraction module suitable for it is designed. apply index query to optimize the ROI pooling layer to speed up voxel feature extraction. The experimental results on the 3D dataset of KITTI show that the results are 90.63%, 81.74%, and 77.23% on three different detection difficulty levels of cars, and the frame rate of detection processing is 28.5FPS. Compared with other methods, this method of a faster detection speed can be achieved while maintaining a higher accuracy rate.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621D (2023) https://doi.org/10.1117/12.2660952
In order to solve the problem that the handwritten font of the power operation ticket is not easy to recognize and improve the work efficiency of the dispatch operation ticket, this paper proposes a text recognition-based approach to power operation ticket review. First, a text recognition method based on handwriting feature recognition convolutional neural network (CNN) is proposed, which introduces imaginary strokes and multi-directional features as input and uses a simple average calculation method to obtain the classification results. Then use the image samples of operation tickets in the actual operation and maintenance to conduct experiments to verify the effectiveness of the method proposed in this paper. Finally, a power operation ticket review system based on text recognition is designed.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621E (2023) https://doi.org/10.1117/12.2660809
In order to improve the effect of crew training on board, this paper establishes a crew training management system based on B/S architecture. "Crew training system" is a set of system for information management of crew training on board. Through the use, it is found that the system improves the training effect of the crew, provides a management means for the shipping company, and also helps the career development of the crew.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621F (2023) https://doi.org/10.1117/12.2661171
Spectral clustering has been successfully used in the domain of pattern recognition and computer vision. Kernel subspace clustering has become a hot research topic because it can reveal the nonlinear structure. However, the performance of exiting single kernel subspace clustering relys heavily on the choice of kernel function. To address the problem, we propose a novel method called multiple-kernel based subspace clustering method (MKSC) by combining kernel block diagonal representation with multiple kernel learning. The proposed MKSC algorithm firstly obtains the optimal kernel matrix by using multiple kernel clustering method, then replace the kernel function in single kernel subspace clustering model with the optimized kernel matrix, finally the clustering result is got by optimizing the MKSC model. Experimental results on three datasets testify the effectiveness of our proposed MKSC method.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621G (2023) https://doi.org/10.1117/12.2660790
This e-commerce system adopts HTML+Javascript+Python+Django, following the mode of modern social trading platform1. This data is found on the Internet and imported into the database for easy calling and viewing. E-commerce products are divided into different types on the website, and each product has its own brand. This makes it convenient for consumers to select the products they need and improve product sales. The simplicity of Python language makes this design can be completed efficiently. There is a clear division of labor between the front end and the back end. The front end uses HTML + JavaScript, and the back end uses Python + Django 2. Unlike the traditional Django framework, the advantage is that it does not require third-party libraries and tools to build the website, making it easier to use.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621H (2023) https://doi.org/10.1117/12.2660933
Microservices is an emerging software development architecture. The orchestration and optimization of microservices can effectively reduce the operating latency and improve the service quality. This paper investigates the problem of microservice deployments in the edge side of the intelligent factory. The microservices make up applications with different topologies. The goal is to minimize the weighted summation of the runtime of applications. The problem is formulated as a mixed-integer programming model and solved by the CPLEX solver. The experimental results show that the microservice deployment scheme presented in this paper is superior to the greedy deployment strategy.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Mingyuan Hu, Yi Zhang, Jun Guo, Yazhou Chen, Feng Wei, Mengjiao Gou
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621I (2023) https://doi.org/10.1117/12.2660959
Oil production plant A has entered the stage of high water cut in the middle and late stages of development, and its benefits are getting worse and worse. In order to formulate a reasonable production plan for the enterprise, avoid blind investment and improve production efficiency, based on the current situation, this paper introduces the modeling process of the ARIMA model in detail. Based on the oil production data of block T of the A oil production plant from January 2007 to September 2021, the ARIMA model is used to analyze the oil production. Four different ARIMA models were obtained through judgment, and after comparison, the study found that the prediction error metrics of the ARIMA (3,1,1) (2,0,0) [12] model were all small, and the model The effect of the short-term prediction of the daily oil production sequence in block T of the A oil production plant is relatively accurate, and it is of great significance for the short-term prediction of oil production.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621J (2023) https://doi.org/10.1117/12.2661052
The workload of the block painting stage accounts for the largest proportion of the ship painting, so its operation scheduling directly affects the painting quality and efficiency. In this paper, the scheduling problem of ship block painting operation is investigated. The problem can be regarded as a variant of the flexible job-shop scheduling problem. The mathematical model is developed considering a variety of practical constraints, including space size constraints, multiple spray-painting constraints, and waiting time constraints between processes. To effectively solve the problem, we propose a hybrid meta-heuristic to combine the merits of genetic algorithm and tabu search. We conduct numerical experiments on actual production data instances of different scales. The results show that our hybrid approach has significant advantages over the baseline algorithms.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621K (2023) https://doi.org/10.1117/12.2660802
Millimeter-wave radar has become one of the mainstream sensors applied in the field of target tracking research because of its high accuracy, low cost, and non-intrusive nature. Millimeter-wave radar can solve the user's concern about privacy leakage. In this paper, a millimeter wave radar target tracking and recognition method based on an interframe fusion algorithm is proposed. The fusion algorithm is the combination of the DBSCAN algorithm and GMM algorithm. The method first combines multi-frame point cloud data and then compares the fusion algorithm with the DBSCAN algorithm. The multi-frame merging method with frame sequence features solves the problem of sparse radar point cloud data. The fusion algorithm not only solves the problem of noise and target recognition confusion but also solves the problem of DBSCAN identifying multiple clusters into one cluster in an indoor multi-target scene. The result shows that the fusion algorithm improves the recognition accuracy of target tracking compared with the original algorithm and can be applied to the indoor multi-target tracking and recognition scene.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621L (2023) https://doi.org/10.1117/12.2660801
With the application of target detection technology in UAVs and other aircraft, the demand for the analysis of aerial images is increasing year by year. Focusing on the single-stage target detection algorithm in deep learning, the current mainstream deep learning single-stage target detection algorithm is first introduced; the basic principle and optimization process of the single-stage series target detection algorithm are sorted out, and a systematic summary of the research progress of aerial image detection by Single-Stage series algorithm is analyzed. Finally, the future direction of the Single-stage series algorithm in aerial image detection prospects.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621M (2023) https://doi.org/10.1117/12.2660986
Aiming at the problems of single camera color measuring system, a method of 3D object color measurement based on convergent binocular stereo vision is proposed. By taking a pair of 2D images, an 3D image is reconstructed in a 3D point cloud model, in which color of each point is restored by fusing colors of corresponding point of 2D images. Based on color charts with 240 and 24 colors, a distinctive 11-term polynomial is trained to convert colors from image RGB to CIELAB. An experiment was conducted to test the proposed method. The results show that the color prediction accuracy for the proposed model was good enough.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621N (2023) https://doi.org/10.1117/12.2660839
As the classical analysis tool, Poincaré plot is widely applied to various nonlinear signals. Via Poincaré plot, the physiological status of heart can be shown by different distribution pattern of scatter points. In this paper, a novel quantitative index is presented, which is used to analyze these distributions patterns. In this study, the micro change of consecutive R-peak time intervals is considered in the novel index and experiment results demonstrate that the three physiological statuses of heart can be classified effectively. Hence, the novel index can be used in the recognition and classification of cardiac diseases.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621O (2023) https://doi.org/10.1117/12.2661047
Aiming at the problem that the traditional scene matching navigation algorithm needs to manually design features, a scene matching navigation algorithm based on improved Siamese network is proposed. First, the space transformation network module is fused with the original Siamese network to improve the fitting ability between the scene features, and then the improved network is applied to the location and orientation algorithm of aircraft. The experimental results show that the Siamese network image matching navigation algorithm based on the fusion space transformation module enhances the ability to deal with the rotation and translation transformation between the real-time image and the reference image. Compared with the original Siamese network, the algorithm in this paper optimizes the similarity of two scenes from different angles by an average of 9.04%, thus expanding the adaptability of the algorithm. Compared with the traditional template matching algorithm, this algorithm has higher matching accuracy and stronger robustness when the angle of the real-time image changes and has certain practical application value in navigation algorithms.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Hongqing Zeng, Min Xie, Yong Dong, Zhenwei Wu, Haoming Lin
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621P (2023) https://doi.org/10.1117/12.2660798
Dynamic binary translation is widely used in cross-ISA (Instruction Set Architecture) binary migration. With the technology, computers can directly execute programs compiled in other platforms, which helps to reduce the cost of software migration between different architectures. Efficient condition code emulation is critical for achieving higher performance. However, existing binary translation systems typically have to emulate the guest flags register with memory variables to hide the ISA difference, which is less-efficient due to code expansion. In this paper, we propose a way to substitute guest conditional instructions with counterpart instructions from the host ISA to achieve higher performance in flags register emulation. We recognize flag-defining and flag-checking instructions respectively through scanning the guest code sequence and then reproduce the condition pattern in the dynamically generated host code with the corresponding flag-related instructions in the host ISA. The evaluation results show that our approach manages to reduce the dynamically generated host instructions by 7.53% and improve the performance by 14.3% on average.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621Q (2023) https://doi.org/10.1117/12.2660994
The existing 3D human pose estimation method in exploring the relationship between root joint and other key points, to some degree reflects the local movement, but they ignore the trajectories of the whole of the human body, and to some extent, it reflects the global movement, even explore the method of the root node and other related points made significant progress, However, it is not robust to global motion, and it is not perfect to small local motion. For the above two kinds of situations, we proposed representation of temporal and position enhancement. Position enhancement is to maintain the consistency of the input and output distribution by position encoding the correlation of 2D coordinates of human pose. Temporal enhancement occurs by making connections between the current posture of the same person over time and other postures, as we focus more on the current posture and changes in motion in the forward and backward posture.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Shengjie Luo, Zhigang Liu, Yiting Wang, Jialiang Liu
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621R (2023) https://doi.org/10.1117/12.2660794
We conduct research on a practical task called Armored Vehicle Object Detection (AVOD), which is designed to real-time locate and identify armored vehicles in the armored cluster. Considering the scarcity of available datasets under the background of AVOD, we carefully collect a new dataset named AVOD2K, containing 2,000 pictures which consist of nine types of armored vehicles in various environments. AVOD2K complements the missing armored vehicle dataset, and it could drive special vehicle object detection in complex scenarios. In addition, we use YOLOv4 as the baseline of the task and propose LK-CSPDarkNet to modernize YOLOv4 by combining depthwise separable convolutions and large kernels. As the parameters decreased by 7% and FLOPs a little drop, the modernized YOLOv4 outperforms the baseline in 2.4 AP.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621S (2023) https://doi.org/10.1117/12.2661186
With the development of the 5th Generation Mobile Communication Technology (5G), the 5G network can more accurately locate the user's location. This enables location-based services to provide users with more personalized services, but it also brings more challenges to user location privacy protection. Aiming at the leakage of location privacy and query privacy in LBS, a LBS privacy protection scheme based on searchable encryption is proposed. The scheme uses thresholds to replace the private information queried by users, thereby reducing the risk of location information leakage during user service requests; it reduces the storage space of the server through attribute-based encryption and improves query efficiency. Finally, the correctness of the ciphertext search of the scheme is proved.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621T (2023) https://doi.org/10.1117/12.2661030
Image classification is a basic task in the field of computer vision, and general image classification task training requires a large amount of labeled data to achieve good generalization performance. However, in practical applications, the cost of obtaining labeled data is expensive. In contrast, unlabeled images are easy to obtain, so semi-supervised image classification is more meaningful for research. This paper pro- poses a framework for semi-supervised classification utilizing multiple self-supervised methods. Our approach is divided into three steps, firstly, pre-train multiple models on unlabeled data using different self-supervised methods. Then use the labeled data to fine-tune these models except the model pre-training by Contrastive learning to obtaining multiple self-supervised teacher models. Finally, the multi-teacher knowledge distillation framework is used to transfer the knowledge of multiple self-supervised teacher models to the model pre-training by contrastive learning to help it achieve further performance. We conducted experiments on cifar10 and miniimagenet60. Our method achieves further results than using only a single self-supervised method, and also achieves superior performance compared to other semi-supervised methods.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621U (2023) https://doi.org/10.1117/12.2660792
The atmosphere contains many tiny, suspended particles, and due to the scattering and absorption of these particles, images can show reduced visibility, distorted colors’, blurred details and other situations. Many computer vision applications are unable to accept these degraded images and therefore require high quality input images to ensure accurate work, provided by a defogging method. Single image deblurring utilizes physical models where transmission estimation is an important parameter in obtaining a fog-free image. The fog image is analyzed and pre-processed to highlight details and make it more suitable for human and machine recognition. The analysis of different deblurring methods divides them mainly into methods based on image a priori recovery, image enhancement and deep learning. The content of defogging-related algorithms is described, and future directions are analyzed.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621V (2023) https://doi.org/10.1117/12.2660782
With the continuous development of information technology, digital signal processing has been paid more and more attention. Aiming at the visualization of spectrum analysis and power spectrum estimation in digital signal processing, this paper designs and develops a simulation platform with the help of MATLAB GUI. The simulation platform has simple operation, intuitive display, comprehensive functions, good interactivity and strong intuition, and can visualize abstract and difficult algorithms.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621W (2023) https://doi.org/10.1117/12.2661090
On the aircraft, the utilities management computer completes the information interaction with the utilities systems and the avionics system, which plays a vital role in ensuring the safe flight of the aircraft. This paper mainly discusses the voice alarm mechanism of utilities systems based on utilities management computer. For the voice alarm that occurs during the flight, the fault inspection and mechanism analysis are carried out. For the reported bus access timeout fault, the fault location is completed, and a solution is proposed. It will provide important support for the development and guarantee the computer to ensure the safe flight of the aircraft and the utilities management.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621X (2023) https://doi.org/10.1117/12.2661053
Emotion will have an important impact on human behavior, and emotion management needs to be paid attention to, and speech is a significant way to reflect human emotion. The speech information rich in speech can show the inner emotional information of the speaker, and the speech signal analysis is an effective method to achieve. Through speech recognition system, the feature parameters of speaker speech are extracted, and the speech samples are classified by support vector machine (SVM) algorithm, so as to obtain the emotion data of speaker speech. Use the acquired emotional data to build a personal emotional database, so as to show the long-term emotional state of individuals. In this way, we can observe people's long-term emotional trend and avoid mental illness.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621Y (2023) https://doi.org/10.1117/12.2661063
As an important part of license plate recognition system, research for the license plate detection has made great progress in recent years, it is still affected by complex environments such as weather, distance, angle and brightness. Therefore, a MFA-UNet model is proposed in this paper, which is based on the UNet model structure and combines the multi-scale convolution feature fusion module and the spatial attention mechanism. In the last two layers of the up-sampling stage, the multi-scale dilated convolution feature fusion module is used to cancel the pooling operation, which ensures that the receptive field can be increased without losing the image resolution, and the image features can be enhanced. The attention of the license plate area is increased by introducing a spatial attention mechanism; the learning and training process have been optimized by using the focal loss function. Based the experiments results, accuracy of the model algorithm mentioned in this paper is 4.5% higher than the original UNet in IoU (Intersection over union), and average detection accuracy of the MFA-UNet model on the Chinese City Parking Dataset (CCPD) dataset is 97.8%, which is a great improvement compared with the target detection algorithm.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124621Z (2023) https://doi.org/10.1117/12.2661099
According to the maintenance process of complex equipment and the requirement of maintenance system, the general process of maintenance decision of complex equipment is analyzed. The maintenance decision system model of complex equipment is designed according to the complexity and distribution of maintenance process. The structures of Agent in the system are analyzed.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246220 (2023) https://doi.org/10.1117/12.2660819
In order to realize the intelligent extraction of Chinese medicine production enterprises, the old equipment can be transformed. The method of transformation is to use camera technology to obtain data. In view of the situation that the collected images are not clear enough, this paper establishes a step-by-step correction method for photographic images. In the field of visual inspection, cameras are used to collect images, and the images taken are photographic images. The photographic images may be distorted so that the parallel lines on the object become non-parallel lines in the captured image. In order to restore the true shape of the object, the image needs to be corrected. Image processing system includes image capture module, image recognition module and image correction module. Image correction first performs parallel line intersection calculation and then performs image step-by-step correction. Image step-by-step correction includes affine correction, similarity correction and metric correction. This paper summarizes the computer running logic of each correction module through experiments. The corrected camera image enters the monitoring system of traditional Chinese medicine extraction, which makes the detection of traditional Chinese medicine extraction more timely, comprehensive and effective. The successful application of step-by-step correction software for photographic images has broken through the technical bottleneck of traditional Chinese medicine extraction equipment, realized real-time and effective monitoring of key indicators in the production process of traditional Chinese medicine extraction, and formed an advanced and effective quality control system for the production process of traditional Chinese medicine extraction. Stepwise image correction technology can also be applied to process monitoring in other industries to achieve intelligent production management.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246221 (2023) https://doi.org/10.1117/12.2661312
As an important branch in the field of image processing, image fusion has become one of the hot issues in research. For pixel-level image fusion, multi-scale multi-resolution decomposition has been widely used, but in the processing of low frequency sub-bands, because the averaging method is easy to lead to blurry fused images, contrast degradation and other issues, a low-frequency sub-band fusion method based on regional energy is proposed; for high-frequency sub bands, the influence of their neighborhood coefficients is also considered while the coefficients themselves are considered. The fusion experiment of the registered image and the evaluation of the fusion image show that the method in this paper fully preserves the high-frequency edges and details of the image while preserving the image contour.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246222 (2023) https://doi.org/10.1117/12.2660941
In order to realize the development path planning of education-industry-education integration under the vision of big data information fusion, the development path planning model under the vision of big data information fusion is proposed based on information resource integration scheduling. This paper constructs a big data statistical average analysis model of educational production-education integration development planning under the big data information fusion vision is constructed, and the big data feature extraction and relevance description is taken combined with descriptive statistical analysis results. Pattern recognition and feature screening are carried out by using the average mutual information clustering method for the extracted statistical features of the development planning of education-production-education integration under the vision of big data information fusion, so as to realize the development path planning of education-production-education integration under the vision of big data information fusion based on big data analysis. The results of empirical analysis show that this method has a high level of confidence in the development path planning of the integration of education, production and education from the perspective of big data information fusion, and the evaluation results are accurate and reliable.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246223 (2023) https://doi.org/10.1117/12.2661083
Existing point-based, sparse voxel-based, or hybrid point cloud processing methods require time-consuming neighborhood searches or sparse 3D convolutions, which consume a lot of time and computational resources. Therefore, it is difficult to run at high speed in real time on mobile devices. To this end, we reconstructed the internal structure based on RangeNet++1 and designed an efficient and lightweight network named AMBrnet, still using the encoder-decoder architecture. An asymmetric multi-branch aggregation module is designed to directly aggregate the input information and provide rich information for subsequent step encoding. The dual-branch structure is introduced to combine with the encoder and decoder to strengthen the information encoding and decoding capabilities of the network. Weighted cross-entropy loss combined with Lovász-Softmax loss2 is used to directly optimize the Jaccard index (IoU). In the network inference stage, structural reparameterization is introduced to ensure that the inference speed is improved based on the same accuracy, reducing the number of network parameters. We evaluate the proposed model on the SemanticKITTI dataset. The prediction accuracy is better than most existing networks, notably its inference speed is up to 43.9 Hz. Experimental data show that AMBrnet is well suited for real-time high-speed point cloud processing on outdoor mobile devices.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246224 (2023) https://doi.org/10.1117/12.2660973
An expression synthesis method based on Progressive Generative Adversarial Network (PGGAN) and StyleGAN2 is proposed to create customized datasets in order to address the problems with traditional expression dataset creation methods, such as their high cost, length of time, and difficulty in avoiding the influence of subjective annotation factors. In order to separate identity features from expression features in expression images, this paper first uses PGGAN to generate images of various resolution layers. It then builds a feedback network for this image and obtains the feature latent codes corresponding to the generated images of each resolution layer. The target face image is then produced with specific expressions based on the image generation direction dictated by the fusion result, using the StyleGAN2 network model to fuse the latent codes of identity information in the target face image with the latent codes of expression information in the original expression image. Finally, the Fréchet onset distance (FID) and structural similarity (SSIM) of the synthesized image and the original image are compared. The mean values of FID and SSIM compared with the original image in this paper are 34.61 and 0.90, respectively. It is difficult for human visual perception to distinguish the real from the fake, proving the authenticity of the synthesized image and the efficacy of the synthesis method.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246225 (2023) https://doi.org/10.1117/12.2661032
Aiming at the declining trend of the overall binding quality of adhesive binding publications, a defect detection system for signature collation based on machine vision is designed. The signature image is collected by the CCD sensor, and the traditional image algorithm is studied by using OpenCV open-source vision software. After preprocessing the signature image to be detected, the Hough transform and affine transformation algorithms are firstly used to extract and describe the features of the target image, calculate the tilt angle of the signature, and correct the tilted image. Then, the template image is matched with the signature image for template matching to locate and extract the region of interest ROI; finally, gray statistical features, perceptual hash and image difference methods are selected to process the template and ROI images, and calculate the matching similarity. The experiment realizes the real-time monitoring and error processing of many kinds of signatures such as word pages, image-text pages and graphic pages, and completes the detection task. The results show that the accuracy of the system defect detection can reach 99%, which has good practicability and stability.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246226 (2023) https://doi.org/10.1117/12.2660805
With the rapid development of big data and Internet technology, the tourism industry is actively combining with big data and new technologies. As a service industry, how to accurately understand the concerns of tourists and create a good tourism destination image has become the key to the management and development of scenic spots. Taking Sanwei Bookstore -Lu Xun's residence as the research case, this paper uses Octopus software to collect Ctrip network comments and use RostCM6 to make a qualitative analysis of the online comments made by tourists. Through big data analysis, the paper discusses the image perception of domestic tourists for the destination from the cognitive image, emotional image and overall image, and puts forward targeted suggestions on this basis, in order to improve the image perception of Sanwei Bookstore- Lu Xun's residence and speed up the development of its study tourism.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246227 (2023) https://doi.org/10.1117/12.2660967
Since the appearance of bitcoin, blockchain technology has been widely treated as an opportunity for businesses and the technology revolution. Until now, blockchain technology has been applied to several fields such as energy system evolution, Cryptocurrency, and social media. In this paper, the author mainly explored the current application of blockchain technology in social media, the algorithm behind blockchain technology, and the existing problem in blockchain technology. For the application in social media, this paper explores fake news detection, user trust framework, and decentralized online social network. For the algorithm behind blockchain technology, the Raft algorithm and Practical Byzantine Fault Tolerant algorithm are discussed in this paper. Moreover, one of the innovative consensus algorithms will be explained in the subsection. The following topics will be discussed in the existing problems section: Internet of Things issues, delayed confirmation, and information security. Finally, the author concluded that the current application of blockchain in social media is still developing, and more possibilities for blockchain application in social media will be found with further development.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246228 (2023) https://doi.org/10.1117/12.2660806
The high covertness of APT attack is an important feature that is different from traditional cyber-attacks, and it also can reflect the attacker's ability. The existing evaluation methods for the covertness of APT attack usually analyze the malwares used in the APT attack process. To evaluate the covertness of APT attack systematically, we propose a quantitative method. For multiple phases of APT attack, we construct two covertness metrics of each single attack phase considering the probability of the attack detection and the degree of disturbance to the target system, and then merge multiple phases to analyze the covertness of APT attack process. According to the results of simulation experiments, the method can evaluate the covertness of APT attack effectively and accurately. Quantifying APT attack covertness can help defenders to understand the specific process of APT attacks more clearly and provides a method to learn about the ability of attackers. It also can give some significant suggestions for the distribution of defense resources and the defense strategies.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246229 (2023) https://doi.org/10.1117/12.2660975
In recent years, a large number of mixed products, counterfeit products and adulterated products of traditional Chinese medicine powder have often appeared on the market. And because the traditional image processing algorithm cannot meet the application requirements for such complex background, large amount of data, and multi-category problem processing accuracy and speed. In view of the above problems, this paper proposes an improved method of deep convolutional neural network based on texture extraction and improved attention mechanism model. First, the MobileNetV3 network is unable to fully extract the texture information of fiber powder, and an inception-like structure is added to significantly improve the final recognition accuracy; secondly, the use of hole convolution minimizes the impact of the inception-like structure on the overall parameter quantity. Influence: finally, an improved attention mechanism module is added to the network, which significantly suppresses the noise texture in the background image. The fiber characteristics of 32 different kinds of Chinese herbal medicine powder images are used for experimental comparison. The added inception-like structure and the improved attention mechanism module increase the accuracy rate by 4.1% to 92.82%. The experiments show that the improvement method proposed in this paper is better than other classification algorithms.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622A (2023) https://doi.org/10.1117/12.2661115
The outbreak of COVID-19 makes people feel distant from each other, and masks have become one of the indispensable articles in people's daily life. At present, there are many brands of masks with various types and uneven quality. In order to understand the current market of masks and the sales of different brands, users can choose masks with perfect quality. This paper uses Python web crawler technology, based on the input of the word "mask", crawl JD website sales data, through data visualization technology drawing histogram, pie chart, the word cloud, etc., for goods compared with the relationship between price, average price of all brands, brands, average distribution of analysis and evaluation of user information, In this way, the sales situation, price distribution and quality evaluation of each store of the product can be visually displayed. At the same time, it also provides some reference for other users who need to buy the product.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622B (2023) https://doi.org/10.1117/12.2660978
System integration is a necessary solution to the database explosion and data silos and is a major trend in the development of web technologies. This paper examines system integration in terms of both breadth and depth, detailing the concepts, methods, typical application areas, and challenges faced by system integration. It also focuses on the most popular SOA-based integration methods. Finally, a summary and outlook on system integration are given.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Neural Network Application and Algorithm Modelling
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622C (2023) https://doi.org/10.1117/12.2661172
When testing software functions, it is necessary to adopt scientific and reasonable test cases to ensure the accuracy of test results, so as to lay a good foundation for the operation and use of software. Based on this, this paper designs a new algorithm FICBPR on the basis of the traditional abstract test case set optimization sorting algorithm (FICBP) based on repeated one-dimensional combination coverage and analyzes the application effect of the algorithm through empirical analysis. Through verification, it can be found that the FICBPR algorithm has lower error rate and shorter time overhead, and its performance and efficiency are significantly better than the traditional FICBP algorithm.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622D (2023) https://doi.org/10.1117/12.2660808
Parkinson's disease (PD) is a common neurodegenerative disease, with a high probability of Parkinson's disease dementia (PDD) in patients with intermediate and advanced PD. Gait disorders and cognitive disorders are common symptoms of PD patients and PDD patients. It is of great clinical significance to identify healthy elderly (HC), PD patients and PDD patients with gait characteristics under cognitive tasks. This study found that stride length, toe-off angle and heel-strike angle are important gait markers for identifying HC and PD as well as HC and PDD. Gait characteristics of multiple 7 task gait consumption can preliminarily identify PD and PDD. The gait features under multiple 7 task were used as input variables of machine learning, and the classification model was modeled by training random forest (RF) and support vector machine (SVM), and the accuracy of machine learning classification was evaluated by using the five-fold cross-validation method. The results found that the classification accuracy of all machine learning can reach more than 80%, and RF has a better classification effect. To further improve the recognition accuracy, this paper introduces recursive feature elimination (RFE) for important feature selection. By screening important features, it is found that the accuracy and AUC value of machine learning are improved to a certain extent. The highest classification accuracy of HC and PD is 91.25%, and the AUC value is 0.9127. The classification accuracy of HC and PDD was up to 97.5%, and the AUC value was 0.95. These findings have important application value for clinical diagnosis of PD and PDD. It also paves the way for a better understanding of the utility of machine learning techniques to support clinical decision-making.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622E (2023) https://doi.org/10.1117/12.2660936
In order to improve the computational efficiency and accuracy of the neural network algorithm, this research establishes an optimized neural network algorithm. Firstly, the optimal training function and the optimal number of hidden layer nodes of the neural network are obtained by using the empirical formula method; secondly, the prediction accuracy of the neural network is optimized, and the science of the neural network is improved. Finally, with the network as the core algorithm, the quantitative adjustment rate is used as the weight coefficient to improve the evaluation and calculation of the impact factor MIV.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622F (2023) https://doi.org/10.1117/12.2660840
This paper proposed a multi-strategy adaptive artificial bee colony (MAABC) algorithm for solving the multi-objective flexible job shop scheduling problem. This approach optimizes the combination of the different search strategies to produce a highly efficient variant of the artificial bee colony (ABC) algorithm. An adaptive mechanism is incorporated into ABC, and therefore, during the process of evolution, based on previous experiences, each search strategy's selection probability is dynamically adjusted. Furthermore, construct a local search strategy oriented to object characteristics to further improve the algorithm’s exploitation ability to exploit different objectives. In this study, the proposed algorithm is tested on a well-known benchmark instance, and comparisons with other state-of-the-art algorithms are conducted. Detailed analysis of experimental results reveals the highly effective performance of the proposed algorithm.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622G (2023) https://doi.org/10.1117/12.2661154
At present, blockchain is in a period of rapid development. However, there are many hidden dangers in its development process. This paper completed the analysis and verification after studying various indicators of the Bitcoin network and putting forward hypotheses based on experience. It is found that the correlation between bitcoin network trading volume and the amount of account funds remains stable. Most of the early saving accounts are actually still transaction participants in the current ecosystem. At present, most bitcoin transaction network entity recognition and classification methods are still based on network graph structure features. This paper proposed an entity recognition and classification method based on the relevance features by starting from the characteristics of bitcoin’s own multi-transaction association. Some data are obtained from public websites, and it shows that the method in experiments could improve the performance of entity recognition effectively. In addition, this paper also analyzes the importance of features and the relevance of classification. In order to deal with the above transaction network feature analysis and entity recognition of the actual demand, this paper also designed a set of highly automated, timely interruption and recovery, and abnormal alarm function of the system. It can be migrated to other Bitcoin networks, and it is very convenient and efficient. This paper can improve the performance of entity recognition and make our understanding of Bitcoin deeper. This system has great theoretical value and application value for the research of bitcoin de-anonymization.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622H (2023) https://doi.org/10.1117/12.2660996
Handwritten digit recognition is a process of identifying zero to nine ten digits handwritten by human hands, and its related research has always been a hot topic in the field of machine learning classification. In order to explore the accuracy of the classification recognition of handwriting bodies by K nearest neighbor classifier and MLP multilayer perceptron, this paper first introduces the relevant algorithm principle and its research progress, and then experiments on K nearest neighbor classifier and MLP multilayer perceptron and summarizes the relevant experimental data. Experiments show that in the K nearest neighbor algorithm, the classification accuracy is the highest when the number of neighbors K=3; for the MLP multilayer perceptron algorithm, the classification rate is higher when the number of neurons is larger, the number of iterations is 1000, and the learning rate is smaller.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622I (2023) https://doi.org/10.1117/12.2660777
With the rapid development of artificial intelligence, deep neural network (DNN) has been widely used in industrial defect detection, intelligent driving, medical research, etc. However, DNN is still limited in the implementation of edge computing and mobile devices due to its characteristics of high model complexity and high computing resource consumption. Therefore, we designed a neural network hardware accelerator based on Field Programmable Gate Array (FPGA) for printed circuit board (PCB) defect detection. In this paper, firstly, since structure re-parameterization can improve the network's accuracy without increasing the inference model's complexity, we introduce structure re-parameterization to improve the YOLOv2 model and propose RepYOLOv2. Secondly, a low-bit quantization method based on integer type is adopted to quantify the model data to 6-bit. Then a specific convolutional computing module and neural network hardware accelerator are designed according to the characteristics of the model. Experimental results on Xilinx ZCU102 FPGA show that the real-time processing speed of the system reaches 2.12 FPS, the throughput is 68.53 GOP/s, and the power consumption is only 1.12 W. Compared with similar work, better performance is obtained.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622J (2023) https://doi.org/10.1117/12.2660925
Some intelligent detection methods for ultra-dense network attacks are likely to generate false alarms in the application process. In order to improve the security in ultra-dense network, an intelligent detection method based on edition learning is designed. Considering the SRP change rate, different thresholds are set, the node switching structural features of ultra-dense networks are extracted, the function sets that can effectively control error detection are selected, the host recognition algorithm is designed, the function field selection model based on joint learning is constructed, the iteration points are created in the feasible domain, real-time network traffic is collected, and the doattack intelligent detection model is optimized. Experimental results: in the paper, the average probability of non-intelligent detection methods for attacks in ultra-dense networks is 24.864%, which shows that when combined with federated learning algorithm, it has more advantages in practical performance.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622K (2023) https://doi.org/10.1117/12.2661076
In view of the complex underground structure and harsh environment of coal mines, it is easy to cause problems such as low pedestrian detection accuracy and missed detection. An improved mine pedestrian detection algorithm based on YOLOv4-Tiny was proposed. The algorithm introduced spatial pyramid pooling (SPP) module after 13×13 features, which realized the extraction of global and local features of image information, and finally improved the detection precision of the system. The mAP of the improved algorithm was 91.98%, which was 2.32% higher than the original algorithm, the precision increased by 3.87% and the recall by 0.93%. The experimental results showed that the algorithm improved the detection of the mine pedestrian system effectively and had a better detection effect.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622L (2023) https://doi.org/10.1117/12.2660934
In order to solve the problems of traditional manual cone valve seal detection, such as time-consuming and laborious, low detection accuracy and low efficiency, etc., we propose a method based on YOLOv3 framework for cone valve seal detection. In this paper, we propose a cone valve seal inspection method based on YOLOv3 framework. Under the self-built simulated test environment, we use a high-definition camera to collect image data samples of high-resolution bubbles and make our own bubble dataset, on which we realize the accurate detection of cone valve seal test under the inspection scenario. The experimental results show that the detection accuracy of yolov3 model can reach 95.5%, and the model can detect 15 pictures per second. Moreover, the method can provide reliable supervision information for whether the sealing test is qualified in the process of cone valve detection, which can not only effectively avoid the time-consuming and laborious manual detection, but also improve the timeliness of safety management.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622M (2023) https://doi.org/10.1117/12.2661039
Recently, coverless image steganography (CIS) has become a hot topic in the field of image steganography. However, currently most CIS methods do not build an invertible two-way mapping between the secret information and image, so that the steganography capacity of CIS is limited. Moreover, the generated stego-images have a visible distinction between real-images, and it is difficult to deceive the deep-learning-based steganalysis, which makes CIS unsafe. To address the above issues, we propose a coverless image steganography method based on invertible neural networks, we call it as CIS-INN. Our method uses an invertible flow-based model, Glow model, and we combine it with adversarial generative network (GAN) to improve generated images quality. In encryption phase, secret information is encoded into latent variables by proposed Gray-code-based coding method, then the Glow model takes these latent variables as a prior and generates stego-images. To against steganalysis, we make CIS-INN more secure by introducing adversarial examples when generating stego-images. In decryption phase, secret information can be recovered through the reverse process of the glow model. Experiments demonstrate the CIS-INN achieves significant improvement of the steganography capacity (4 BPP) and maintains reliable security when confronted with multiple steganalysis methods.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622N (2023) https://doi.org/10.1117/12.2660939
Recommendation system is becoming increasingly important in various fields of life. To guarantee the accuracy of recommendation, the detection of shilling attacks must be considered. However, the performance of the existing detection techniques for shilling attacks is relatively low, especially for unknown types of attacks, the existing detection techniques are not universal. In this paper, we propose an improved clustering algorithm-based shilling attacks detection method. This method uses information entropy to select a feature and uses the selected feature to calculate the similarity between two users in the clustering algorithm. Experiments show that the algorithm has good detection performance in the detection of shilling attacks.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622O (2023) https://doi.org/10.1117/12.2660935
It is of great concern to ship designers to choose schemes with excellent comprehensive performance among many feasible ship types. The existing decision-making methods are generally highly subjective, and they need to be improved in algorithms and other aspects. In addition, for different ship types, the evaluation factors vary greatly. The optimization of sea-river-through container ship scheme is mainly studied here. Firstly, the main influencing factors of the preferred scheme are discussed. Secondly, the ideal solution neural network optimization model is established. Finally, the feasibility and effectiveness of the model application are verified through the example analysis of 250TEU sea-river-through container ship. The research results can provide a reference for the design of sea-river-through container ships and similar vessels.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Neng Sheng Bao, Zhao Peng Luo, Sheng Bao Guo, Yu Chen Fan, Jia Hua Jiang, Li Wei
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622P (2023) https://doi.org/10.1117/12.2660851
In the field of facial expression recognition, problems exist in traditional dataset acquisition, including high economic cost, time-consuming process, and difficulties in avoiding subjective factors. However, expression synthesis methods have provided solutions for such problems. In expression synthesis, facial expression trait decoupling is a key technology that affects the image quality. As such, in targeting incomplete separation of the two main types of features, identity features and expression features, new methods for finishing trait decoupling were proposed in the present study. ResNet50 was used to obtain the initial features and the latent codes were obtained by mapping network. A progressive generative adversarial network was used to generate images of different resolution layers. Subsequently, a feedback network was constructed to obtain latent codes. Thus, the separation of identity features and expression features was achieved, and the FID and SSIM between the generated images and the raw images were smaller. Through the proposed method, the accuracy of facial feature editing can be improved.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622Q (2023) https://doi.org/10.1117/12.2660987
Pressure transmitters have a large number of applications in process industry sites, and the stable operation of pressure transmitters is related to the stability and safety of the entire process industry site. Therefore, fault prognosis of the pressure transmitter can greatly reduce the unplanned shutdown of the plant due to pressure transmitter damage. This paper proposes a fault prognosis method for pressure transmitter based on artificial neural network (ANN). According to the pressure value measured by the pressure transmitter, we construct a time series sequence, and segment each group of ten measured values, and label each segment of data according to whether the pressure transmitter is damaged. Then we build a 4-layer neural network, which is trained using shuffled segmented data. The validation accuracy of the final training can reach 0.98, which can effectively distinguish fault data from normal data.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Le Ding, Haoyu Wang, Chao Chen, Jiayan Xu, Pingping Zhou
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622R (2023) https://doi.org/10.1117/12.2660914
The topic of non-contact diagnosis became a hot topic during COVID-19 and online consultation gained popularity. In this research, a deep learning-based autonomous limb evaluation system is developed for online consultation and remote rehabilitation training for people with physical limitations. Its main goal is to collect and analyze information about limb states. The patient can evaluate the limb state at home using the mobile app, and the doctor can view the data and connect with the patient via the web's chat module to offer diagnostic opinions. Deep learning is used for the Start/End Attitude Determination Model and OpenCV for the limb and hand evaluation model, with the results being uploaded to the server.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622S (2023) https://doi.org/10.1117/12.2660854
The Convolutional Neural Network (CNN) enables deep neural networks to be deployed to resource-constrained mobile devices via model compression and acceleration. At present, channel pruning methods select channels based on channel importance or designed regularization, which are suboptimal pruning and cannot be automated. In this paper, a channel pruning algorithm is proposed to get the optimal pruned structure via automatic searching. By setting the super-parameter constraint set, the combination number of pruning structures is reduced. The number of channels for each layer of the CNN is determined using the sparrow search algorithm, and the optimal pruned structure of the model is found. The results of extensive experiments show that the proposed method can improve the model's parameter compression ratio and reduce the number of FLOPS within the acceptable range of model accuracy loss.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622T (2023) https://doi.org/10.1117/12.2660982
Heuristic optimization algorithm and game theory are the traditional methods of task offloading and resource allocation. In practice, Mobile Edge Computing (MEC) system is a dynamic system with high real-time performance, in which the influencing factors have two characteristics: uncertainty and time variation. Traditional methods can achieve local optimality, but it is difficult to achieve global optimality. In view of the above situation, in order to let the MEC system reached the highest efficiency is put forward based on deep reinforcement learning (DRL) task offloading algorithm, using Markov Decision Process (MDP). The optimization objective is to minimize the average delay and total energy consumption of the system, so that the task unloading optimization can be achieved under the condition of satisfying the delay constraint.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622U (2023) https://doi.org/10.1117/12.2660921
In order to find the optimal solution of task partitioning for data flow graph of cryptography algorithm based on reconfigurable logical array, a task partition algorithm ACP based on ant colony algorithm is proposed. The algorithm first improves the cluster partition algorithm according to the hardware structure of reconfigurable array. On the basis of the ant colony algorithm, many ants in the ant colony algorithm are used to traverse the nodes in the data flow graph, and the number of divided blocks and the cost of inter-cluster communication are used as cost functions to continuously optimize the initialized pheromone concentration matrix and improve the algorithm. The experimental results show that the algorithm has the ability to obtain the optimal solution more quickly and stably than the CBP algorithm and the LSCBP algorithm.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Xingwang Fan, Han Xu, Pengyu Chen, Jing Hu, Tiecheng Song
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622V (2023) https://doi.org/10.1117/12.2660971
Edge computing and network slicing are two key technologies to reduce communication latency and improve network flexibility in fog radio access network (F-RAN). Due to the existence of the massive potential offloading decisions, in this paper, we develop a joint computing offloading and resource allocation strategy to minimize the total energy consumption of the cloud-edge system. In order to meet the quality of service (QoS) of different devices, two different radio access network (RAN) slices are designed. Besides, considering the curse of dimensionality caused by the explosive growth of the UEs, we propose a deep Q-learning (DQN) algorithm, which uses value function approximation to compress the status dimension. Moreover, to reduce the complexity of the algorithm, the problem is divided into two subproblems, which are joint radio resource allocation and fog access point (FAP) selection problem and cloud side task forwarding problem and solved by DQN and greedy algorithm separately. Through simulation, we demonstrate that the method proposed in this paper can effectively reduce the total system energy consumption and shorten the convergence time.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622W (2023) https://doi.org/10.1117/12.2660984
The visual SLAM algorithm based on the assumption of static environment will be affected by the motion feature points in dynamic environment, resulting in the system robustness and accuracy degradation. To solve this problem, a visual SLAM algorithm for dynamic feature removal based on instance segmentation is proposed. Firstly, Yolact network is used to detect potential moving objects, and the obtained semantic information is combined with epipolar geometric constraints to accurately eliminate dynamic feature points. Considering the failure of example segmentation network, this algorithm combines the semantic information in the historical key frame with the depth map and uses the region growth algorithm to segment the dynamic objects that are not detected by Yolact, which further improves the robustness of the system. The experimental results show that compared with ORB-SLAM3, the absolute trajectory error and relative trajectory error of the algorithm in the dynamic environment are significantly reduced.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Binhong Ma, Tingwei Zhang, Mingliang Shen, Jun Tang
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622X (2023) https://doi.org/10.1117/12.2661021
The data association problem in cluttered environments is one of the difficult problems in the field of object tracking. When probabilistic data association algorithms are used in motion model deviation scenarios, incorrect tracking occurs. Combining the adaptive robust Kalman filter (ARKF) with the probabilistic data association (PDA), this paper presents the adaptive robust probabilistic data association (ARPDA) algorithm for estimating the target state in cluttered environments. The results of the experiment indicate that the proposed algorithm has higher tracking accuracy under the condition of motion model deviation compared with the traditional probabilistic data association algorithm.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622Y (2023) https://doi.org/10.1117/12.2660778
At present, due to the COVID-19, China's social and economic development has slowed down. Some life service e-commerce platforms have successively launched "contactless delivery" services, which can effectively curb the spread of the epidemic. Robot distribution is the current mainstream, but robots are different from people and need to have accurate program settings. Both path planning and obstacle avoidance are currently top issues. This requires the mobile robot to successfully arrive at the destination while minimizing the impact on the surrounding environment and pedestrians, and avoiding encroachment on the movement space of pedestrians. Therefore, the mobile robot needs to be able to actively avoid moving pedestrians in a dynamic environment, in addition to avoiding static obstacles, and safely and efficiently integrate into the pedestrian movement environment. In this paper, the path planning problem of unmanned delivery robot is studied, and the path of mobile robot in the crowd is determined by global planning and local planning, and the matlab simulation is used for verification.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 124622Z (2023) https://doi.org/10.1117/12.2660841
Aiming at the low detection speed and poor detection accuracy of traditional image processing algorithms, this paper introduces a surface defect detection method for capsule based on an improved YOLOv3 algorithm. Firstly, to obtain important information, the key parts of the image with more refined features are extracted by introducing a dual attention mechanism into the network. Secondly, considering the feature variability of different defect scales of the capsule, the feature detection network is optimized by using adaptive convolution to improve the detection ability of targets. Thirdly, we use the K-means++ algorithm to cluster the anchor box of the target samples to obtain more suitable anchor boxes for the detection task. The experimental results show that although the algorithm loses a little detection speed, it improves the accuracy of capsule detection and meets the demand for real-time detection. Compared with YOLOv3, the average detection accuracy increases from 84.3% to 88.5%.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246230 (2023) https://doi.org/10.1117/12.2660788
Pooling is an important part of modern convolutional neural networks, which can expand the perception field, reduce the parameter matrix, and avoid overfitting. Currently used maximum pooling, average pooling and various subsequent improved pooling algorithms cannot take into account the contour and background information of the feature map at the same time. In addition, the performance of different pooling algorithms varies greatly on different models and datasets. In this paper, we propose a learnable pooling algorithm. The introduction of learnable parameters allows the pooling layer to adaptively optimize the selection of key feature information that is beneficial to improve model performance during model training. It is experimentally verified that the pooling algorithm has superior performance over the existing maximum pooling and average pooling on several classical models and public datasets for image classification and text classification. The pooling algorithm with the introduction of learnable parameters can better prevent overfitting, steadily improve the accuracy of the model, and is highly generalizable.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246231 (2023) https://doi.org/10.1117/12.2660779
This paper attempts to improve the classical encryption algorithm by using Lagrange interpolation method. A large positive integer is used to represent plaintext, that is, we can use plaintext as a positive integer to carry out Lagrange calculation. Because ciphertext is a big positive integer, not repeatable, it is impossible to analyze ciphertext by probability. The algorithm increases the key space and the difficulty of brute force cracking. This algorithm has the security of one-time pad. The length of the key can be less than the length of the plaintext, and the choice of the key is more flexible.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246232 (2023) https://doi.org/10.1117/12.2660787
The nominal projection data of FY-4A satellite is used for cloud derived wind inversion. Firstly, the cloud image is radiometric calibrated and denoised by the median filter. Then the optical flow is solved by the CLG optical flow method combined with global and local and multi-resolution image deformation calculation technology. The real wind speed is obtained according to the optical flow value. Finally, the rationality of the algorithm is verified by radiosonde data. The results show that CLG optical flow algorithm can be used for the inversion of cloud derived wind, and the average deviation and root mean square error of wind speed at cloud height of 300-700hp are less than 5.5 m/s.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Wenyuan Zheng, Miaohua Huang, Ruifeng Wang, Tianyou Jiang, Guohang Li
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246233 (2023) https://doi.org/10.1117/12.2661160
In recent years, with the increase of the number of pure electric vehicles, the phenomenon of battery spontaneous combustion during charging is also emerging in an endless stream. For early fault early warning, we use a machine learning-based model to predict the probability of battery failure. We also use intelligent algorithm to optimize the hyperparameters of the model so that it can accurately predict the probability of battery failure after different time periods. Through our model, the driver or vehicle safety system can perceive the danger in advance and solve or avoid it in time.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
Proceedings Volume Third International Symposium on Computer Engineering and Intelligent Communications (ISCEIC 2022), 1246234 (2023) https://doi.org/10.1117/12.2660743
In the supply chain, inventory management is a very key problem, and with the change of enterprise organization and operation mode, inventory management has many new characteristics and requirements. From the perspective of system theory and integration theory, combined with the mathematical analysis of genetic algorithm, this paper discusses the new inventory management strategy and new methods. On this basis, this paper puts forward an improved idea: in commercial production, while the inventory is based on the existence of various economic reasons, it is also a kind of inevitable consequences. This is a response to external environment, and cannot anticipate the needs of the future. If humans do not have the ability to put every piece of work to the best, it will create a redundant and uncoordinated environment. In the supply chain, a holistic and agile response to the supply chain can be achieved by means of strengthening inventory based on genetic algorithms. Its new thinking, new ideas in the enterprise has been fully applied.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.