Stereo matching algorithm using census transform and segment tree for depth estimation

This article proposes an algorithm for stereo matching corresponding process that will be used in many applications such as augmented reality, autonomous vehicle navigation and surface reconstruction. Basically, the proposed framework this is developed through a series of functions. The final result from disparity map which has information depth estimation. Fundamentally, input image represents left right images respectively. four steps total, starts with cost computation using census transform, aggregation utilizes segment-tree, optimization winner-takes-all (WTA) strategy, post-processing stage uses weighted median filter. Based on experimental results standard benchmarking evaluation system Middlebury, produce average low noise error at 9.68% nonocc 18.9% all attributes. On average, it performs far better very competitive other available methods benchmark system.