A Fast Quasi-Dense Matching Method

The paper presents a fast quasi-dense matching algorithm using an adaptive window. The algorithm starts from a set of sparse seed matches, then propagates to the neighboring pixels, finally the most points in the images are matched. This algorithm adopts an additional level of incremental calculation to enhance the speed of computing the sum of squared difference (SSD). The confidence coefficient is introduced and the search window is varied in inverse proportion to it, namely realizing the adaptive search window. The algorithm has been tested with three stereo images and the results demonstrate its accuracy and efficiency.