Dense Corresponding Pixel Matching Between Aerial Epipolar Images Using an RGB-Belief Propagation Algorithm

A new algorithm, RGB-belief propagation (RGB-BP), for dense corresponding pixel matching between aerial epipolar images is proposed in this letter. Evolved from the traditional belief-propagation algorithm, RGB-BP makes full use of all the three color components, R, G, and B, and simplifies the parameter settings. In order to reduce the impact of the obvious color differences between corresponding pixels, RGB-BP reduces the sensitivity of central pixels and increases the contribution of neighboring pixels. Three pairs of aerial epipolar images are tested using RGB-BP. The experimental results demonstrate the effectiveness of RGB-BP.