Disparity Map estimation using semi-global matching based on image segmentation
Eunsang Ko, Yo‐Sung Ho · 2016
In this paper, we propose a semi-global matching method based on image segmentation. We perform a k-means clustering algorithm in only left image as image segmentation. Then, to improve result of image segmentation, we integrate adjacent and small labels along edges of objects. After that, we extract feature points to estimate the disparity range in each label, and add weights to the disparity range to reduce disparity errors when choosing the final disparity that minimizes the aggregated cost. The cost aggregation in semi-global matching is applied in each selected label only, which allows parallel computation of the cost aggregation as well as reducing disparity errors along discontinuity regions. As a result, the proposed method generates better results at lower runtime than the original semi-global matching method.