Stereo vision with 3D coordinates for robot arm application guide

摘要

This paper presented the application of stereo vision system for robot arm guidance application. The stereo vision system comprises the process of camera calibration, image rectification, stereo correspondences and 3D reconstruction. Camera calibration process was carried out to permit the extraction of camera intrinsic and extrinsic parameters for the image rectification purposes. Next, the image rectification process will transform and align the images such that corresponding features in each image will appear on the same row. Then, stereo correspondences were performed using the method of Sum of Differences (SAD) block matching algorithm. Lastly, the 3D reconstruction was implemented by backprojecting the image pixel in 3D point. From the experimental results, the stereo vision system well performed as the correct shape of the object and the depth of the stereo scene are recovered. The location and the distance of the object can be obtained through the 3D point.