KEYWORDS: LIDAR, Clouds, Sensors, Data modeling, Detection and tracking algorithms, 3D modeling, Visualization, Object recognition, Head, Signal attenuation
In order to detect the object and inspect the road conditions in real-time, the 2-dimensional (2D) and 3- dimensional (3D) data obtained from the onboard sensors, LiDAR and digital cameras are analyzed for object recognition to assist driving. Due to the uncertainties of the dynamic objects, such as pedestrians, animals or vibrated vehicles, extraction of complete and clear objects from LiDARs datasets requires complex post-processing since LiDAR data can be used for scanning at long distances, i.e., 300m, which can alarm the driver timely to take necessary actions. The dynamic and static objects from the LiDARs point clouds can be detected with the teacher-student framework algorithm along with the KITTI dataset. Furthermore, a semi-supervised theory is utilized to improve detection performance.
In multi-robot systems with dynamic and complex environments, robots are required to avoid not only the static objects but also other moving robots. To solve this problem, this paper presents an implementation of cooperative collision avoidance architecture based on optimized sampling-based collision avoidance paradigm. In our work, localization error is considered and bounded in adaptive Monte-Carlo localization process. Plus, we employ velocity obstacle paradigm in predicting collisions. Subsequently, by using Sampling-based planner and optimization theory, we get an optimizing velocity selection policy. Furthermore, we also introduce our distributed multi-robot system model in this paper. By applying the cooperative collision avoidance method in Gazebo self-driving car simulation environment and ROS mobile robots, it is illustrated that our approach is applicable and well-performed.
Helmet wearing is a major concern for the safety and protection of people on the construction site. Statistic data demonstrate that injuries and accidents occur mainly due to not following prescribed procedures, i.e., not wearing helmet. Camera-based surveillance system can conduct online monitoring task to detect such abnormalities through captured images with image processing system analysis. Although deep learning-based method can achieve higher image identification performance, it requires extensive hardware support of the computational resources. Therefore, it is imperative to design a lightweight network with lower hardware requirement to address such problem. In this paper, a GhostNet, YOLOv5 and a lightweight network are combined to design a model to analyze the image for online monitoring with faster processing speed. The performance of the proposed model is compared with those of the mainstream lightweight models. Experimental results have demonstrated that the proposed model has higher detection accuracy and flexible adaptability.
Rapid-exploration Random Tree (RRT) is an efficient algorithm to search non-convex and high-dimensional spaces via randomly constructing spatial filling trees. This algorithm has been widely used in autonomous robot path planning. However, the basic RRT algorithm has some shortcomings. In order to improve the defects of low search efficiency and poor path quality of the RRT algorithm, this paper proposes an A* based RRT path planning algorithm with the advantages of completeness and optimality of the A* algorithm and fast extensibility of the RRT algorithm. During the procedure of random node sampling of the RRT algorithm, A* path is used to formulate the sampling strategy. Meanwhile, the constraint of the path turning angle is added to the nearest neighboring search of the RRT algorithm, which can enhance the rationality of the search tree node selection and improve the obtained path quality. Simulation experiments have been performed to verify the effectiveness of the proposed method for unmanned aerial vehicle path planning.
Proceedings Volume Editor (2)
This will count as one of your downloads.
You will have access to both the presentation and article (if available).
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.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
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.