Optimization of stereo vision depth estimation using edge-based disparity map

Stereo Vision based depth estimation has become a major research topic due to its enormous importance in applications such as industrial automation, Advanced Driver Assistance Systems. Stereo Vision approach of estimating depth is usually a complex process due to issues such as lighting, reflections, camera distortion and alignment, image complexity, pixel noise etc. A simple but an efficient Edge-Based Disparity Map algorithm to optimize the performance of a stereo vision system and accurately relate computed disparity map to true depth information has been presented in this paper. Additionally, a comparative study between the proposed Edge-Based Disparity Map algorithm, conventional Block Matching algorithm and Semi Global Stereo Matching algorithm performances with varying block window size was investigated and discussed. MATLAB demonstration results obtained from the investigated algorithms for corresponding matching results and disparity calculations which are important building blocks for depth (distance) estimation tasks were evaluated based on accuracy and efficiency (computation time).