Toward realizing Automatic High Beam, we are working on a study to classify the beam mode of headlights as "High" and "Low" from night-time in-vehicle camera images using deep learning. Deep learning requires a large amount of training data, and the more effective data are used for training, the better the accuracy of the model is. However, since creating training data is very time consuming, it is often not possible to prepare sufficient data. One of the approaches to perform any task with limited data is to use the learned weights of another task as initial weights of the model and finetune the model with actual limited data. In the image classification, the weights learned on large-scale dataset such as ImageNet are used as initial weights to perform another task. However, in such initialization, if the dataset domain differs from the target domain, there is a report that it is not a factor of the accuracy improvement and it is necessary to learn a model on the pre-training dataset which is suitable to the target domain. Therefore, in this study, we propose a method to create a pre-training dataset easily suitable for the target domain by utilizing public dataset. Experiment results show that proposed method improves the classification accuracy of "High" and "Low" of headlights beam from nighttime in-vehicle camera image.
Photon counting type X-ray, gamma-ray detector and imager were developed by using CdTe compound semiconductor.
The detector / imager could be applied for practical application and we tried to apply material identificated X-ray CT.
The imager has photon energy discriminate function and high linearity between number of incident photons and output
counts. It make high contrast image for X-ray penetration image and material identification in X-ray computed
tomography (CT) measurement.
The image quality cannot be easily applied to a color management system because the quality cannot be quantified by objective methods in general. In this paper, two subjective experiments are performed to quantify the relation between the printed image quality and the color-change of the images. In these experiments, observers are requested to estimate the color-changed images that are caused by the saturation decrease of three primary colors. The results of the experiment show that decreasing color saturation of the primary colors affects the quality of printed images. In addition, the decreasing saturation of B on image quality is significantly larger than the effects of decreasing saturation of other primary colors.
The present paper describes a query-by-sketch image retrieval system aimed at reducing the semantic gap by adopting relevance feedback. To reduce the semantic gap between low-level visual features and high-level semantics, in this content-based image retrieval system, users' sketches play an important role in relevance feedback. When users mark
similar images of output images with "relevant" labels, the "relevant" images are relevant to the sketch image in positive feedback. This method was applied to 5,500 images in Corel Photo Gallery. Experimental results show that the proposed method is effective in retrieving images.
512-pixel CdTe super-liner imaging scanner was developed. This device was consist with 512 chips of M-π-n CdTe diode detector fabricated by excimer laser doping process, 8 chips of photon-counting mode 64ch ASIC with FPGA circuit, USB2.0 interface with 1-CPU. It has 5 discriminated levels and over 2Mcps count rate for X-ray penetration imaging. This imaging scanner has 512 discrete CdTe chips for detector arrays with the length of 2.0mm, width of 0.8mm and thickness of 0.5mm. These chips were mounted in four plover array rows for high-resolution imaging with 0.5mm-pitch, therefore the pixel pitch was over the pixel width. When images were taken with scanning system with this arrays, we could obtain over-resolution than pixel width. In this paper, this "over-resolution" imaging will be called "super resolution imaging". In high-resolution imaging device, the pixel devices on one substrate were formed by integrated process, or many discrete detector chips were installed on circuit board, usually. In the latter case, it is easer to make each detector chips than former case, and it are no need to consider charge sharing phenomena compare with one-chip pixel devices. However, a decrease in pixel pitch makes the mount to the detector chip to the ASIC board difficult because the handling will also be difficult The super-resolution technique in this scanner by pixel-shift method for X-ray imaging is shown in this paper
An image capturing system for the reproduction of high-fidelity color color was developed and a set of three optical filters were designed for this purpose. Simulation was performed on the SOCS database containing the spectral reflectance data of various objects in the range of wavelength of 400nm ~ 700nm in order to calculate the CIELAB color difference ΔEab. The average color difference was found to be 1.049. The camera was mounted with the filters
and color photographs of all the 24 color patches of the Macbeth chart were taken. The measured tristimulus values of the patches were compared with those of the digital images captured by the camera. The average ΔEab was found to be 5.916.
KEYWORDS: Computer programming, Color reproduction, Signal processing, Image processing, Chemical elements, Data conversion, Cameras, RGB color model, Zirconium, Ytterbium
Multiprimary displays are needed to reproduce the most of visible color. However, while using four or more primary colors, it is difficult to convert color uniquely, because multiprimary displays have variability of color conversion. Even though multiprimary colors can be converted, there are some false edges in the reproduced image by using the most of the current conversion methods.
We introduce linear programming to the color conversion in order to get high accuracy and eliminate false edges. Although the linear programming can help multiprimary displays to reproduce images without false edges, it takes long time to convert high-definition images. Hence, a fast color conversion method that keeps the advantage of the linear programming is proposed. This method is constructed of decision tree and linear programming. Because the decision tree is a discrete classification method, color conversion by the decision tree must be implemented as a discrete classification problem. Therefore, we feed the results of the conversion by linear programming to the decision tree. As the result of the conversion by the method, fast color conversion was obtained in comparison with that obtained only by using linear programming. In addition, our method almost eliminated the false edges in the reproduced image.
The authors developed a new method for automatically and efficiently estimating the boundaries of soft tissue and amniotic fluid and to obtain a fine three dimensional image of the fetus from information given by ultrasonic echo images. The aim of this boundary estimation is to provide clear three dimensional images by shading the surface of the fetus and uterine wall using Lambert shading method. Normally there appears a random granular pattern called 'speckle' on an ultrasonic echo image. Therefore, it is difficult to estimate the soft tissue boundary satisfactorily via a simple method such as threshold value processing. Accordingly, the authors devised a method for classifying attributes into three categories using the neural network: soft tissue, amniotic and boundary. The shape of the grey level histogram was the standard for judgment, made by referring to the peripheral region of the voxel. Its application to the clinical data has shown a fine estimation of the boundary between the fetus or the uterine wall and the amniotic, enabling the details of the three dimensional structure to be observed.
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.