A Stereo Matching Algorithm Based on Color Image Segmentation
Peng Li · Infrared Technology · 2009
This paper presents a new stereo matching algorithm.First,regions of homogeneous color are located by applying the meanshift color image segmentation method.Then according to the algorithm matching points based on variable windows sizes matching method,segments are divided into reliable and unreliable segments,the planes of reliable segments are computed.Finally,the global evaluation function by adding occluded constraints,a greedy strategy is used to minimize the global energy.The experiments results show our algorithm has better performance for processing the large low texture areas and the occluded pixels,a higher matching accuracy than,can overcome the shortcoming of local algorithms.