Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
530103 | Pattern Recognition | 2015 | 10 Pages |
•We propose a novel cross-trees structure to perform the non-local cost aggregation.•The trees׳ constructions are unique and independent on the image itself.•We propose and incorporate two priors into the non-local framework.•New edge weights function is designed according to the trees and the priors.•Optimal support regions can be chose by cutting the cost aggregation flow on paths.
In this paper, we propose a novel cross-trees structure to perform the non-local cost aggregation strategy, and the cross-trees structure consists of a horizontal-tree and a vertical-tree. Compared to other spanning trees, the significant superiorities of the cross-trees are that the trees׳ constructions are efficient and the trees are exactly unique since the constructions are independent on any local or global property of the image itself. Additionally, two different priors: edge prior and superpixel prior, are proposed to tackle the false cost aggregations which cross the depth boundaries. Hence, our method contains two different algorithms in terms of cross-trees+prior. By traversing the two crossed trees successively, a fast non-local cost aggregation algorithm is performed twice to compute the aggregated cost volume. Performance evaluation on the 27 Middlebury data sets shows that both our algorithms outperform the other two tree-based non-local methods, namely minimum spanning tree (MST) and segment-tree (ST).
Graphical abstractFigure optionsDownload full-size imageDownload high-quality image (282 K)Download as PowerPoint slide