Circle detection and location based on binocular stereo vision

Abstract

In this paper, a binocular stereo vision system is developed for recognition and localization of a spherical object. The system uses two cameras to acquire images of object, shows and processes the images from two cameras simultaneously. It then applies point Hough transform to recognize the object and get real-time coordinates through the two images. Experiments with the proposed method yielded good results, as the system can detect and position a spherical object successfully.