Stereo Matching Algorithm of Color Image Based on Weber's Law and Census
Qiang Wu, Xiangnan Zhao, Xuwen Li · 2014
The algorithm presented here is used on unmanned aerial vehicles system. In 2013,C.Stentousmis proposed a matching cost computation model(or function), This matching cost model combined three kinds of constraints, the cost function is formulated by combining three individual costs:a cost computed on an extended census transformation of the images, the absolute difference cost, taking into account information from colour channels, and a cost based on the principal image derivatives.Experiments show that the cost function is excellent performance in similar estimates.Apply it to the stereo matching of ancient buildings, it is favorable for practical. However, this experiment, in color mutated region, leaded to blurred border.Therefore, our proposed method added color space structural information, which based on HVS Weber's law, to the cost function. we use the differential excitation of WLD reflects the structure of color space.This robust cost measures can reduce percent of bad pixels in mutated color region. The proposed method use color image, which is used after epipolar line rectification.