A New Binocular Vision Stereo Matching Algorithm

Abstract

With the shortcomings of the large data amount and long time consuming in the conventional binocular vision stereo matching algorithm, a new fast stereo matching algorithm is presented in this paper. It is fast computed based on the line moments and angle vector .Firstly, the method detects image edge using canny operator, and finds the target objects according line moments, then the object contours feature points are extracted. Finally, using angle vector to realize the corresponding pixel matching in the left and right image .The algorithm effectively reduces the computational complexity, computational cost is decreased greatly. The experimental results show that the algorithm is possible and valid.