Stereo Matching Algorithm Based On Binocular Stereo Vision Research And Application

Binocular stereo vision technology is an important research field of computer vision, which is widely applied to the intelligent robot navigation virtual three-dimensional reconstruction, etc. Stereo matching algorithm is a key technology for binocular stereo vision, and the accuracy of stereo matching directly influence applications of stereo vision technology.Common categories of stereo matching algorithms are studied and compared. Region-based local stereo matching algorithm with the advantages of fast and accuracy is the most common use of computer vision applications, but stereo matching in the disparity discontinuous pixels and weak texture region is still a problem of computer vision. In order to improve the matching accuracy in the mentioned regions, an Edge feature stereo matching algorithm based on image segmentation is proposed. Firstly, the reference image is segmented by Mean-Shift algorithm, extracting edge after image segmentation. Then, the pixels of the image are divided into three categories according to the edge feature: dynamically select the support window based on the type of pixel. Finally, the introduction of the weighting factor is put forward. The improved algorithm uses Middlebury’s pictures to stereo matching and the result is detected by its online system. The results show that this algorithm can reduce noise and mismatching points, effectively improve the matching accuracy.Extracting depth information of target in three-dimensional space is one of the main tasks of stereo vision. Therefore, this project builds a binocular vision platform. The paper research geometric model of binocular stereo vision and derive the Passive Ranging formula. Camera calibration is the foundation of binocular stereo vision. In order to improve the platform measurement accuracy, the experiments of camera calibration and the camera optical axis correction are designed. Two groups of text are designed to detect the accuracy of binocular stereo vision system. The results show that Binocular stereo vision system has high measurement accuracy. Finally, the paper analyzes the system error.