3D Detection-system Optimization Based On Binocular Stereo-vision

With the rapid development of visual theory and visual technology, visual measurement, as one of today’s high-techs, develops rapidly and has widely been applied in field of industry measurement. As one of the main visual measurement techniques, contact-type binocular stereo-vision measurement can obtain object’s three-dimension information by collecting a small count of point-clouds, so it plays an increasingly important role in area of industry and has broad applicability.The article makes research of 3D detection-system optimization on the binocular stereo-vision measurement platform had set up already. Contact-type binocular stereo-vision measurement technology includes image acquisition, camera calibration, feature extraction, image matching,3D reconstruction, and data post-processing, etc. On the basis of briefly introducing binocular stereo-vision measurement platform and its mechanical structure, measuring principle and realization of the process, the article particularly studies system architecture and hardware design,3D reconstruction, data post-processing and improvement of measurement accuracy.3D reconstruction of the contact-type binocular stereo-vision measurement mainly aims at NQS such as plane sphere cylinder and cone, which are common in the industry products. Combining the feature of measuring data, the article designs and completes the targeted reconstruction algorithm.By experimental verification, the article makes detailed analysis of factors affecting the measurement system accuracy and puts forward optimization method.The article completes the related algorithm program using VC++ and OpenCV.