Sparse representation over discriminative dictionary for stereo matching. (November 2017)

• Record Type: Journal Article
• Authors: Yin, Jihao; Zhu, Hongmei; Yuan, Ding; Xue, Tianfan
• Is Part Of: Pattern recognition. Volume 71(2017:Nov.)
• Page Start: 278
• Page End: 289
• Publication Date: 2017-11
• Subjects: Computer vision -- Stereo matching -- Data-driven -- Sparse coding -- Dictionary learning; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2017.06.015
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: Presenting a novel data-driven matching cost function based on sparse coding. Learning a dictionary by using weighted sparse coding and discriminative learning. The number of atoms in the dictionary is automatically decided during training. Computing the matching costs based on the sparse representations. The proposed method achieves the best accuracy on 30 test stereo images. Graphical abstract: Abstract: We propose a novel data-driven matching cost for dense correspondence based on sparse theory. The ability of sparse coding to selectively express the sources of influence on stereo images allows us to learn a discriminative dictionary. The dictionary learning process is incorporated with discriminative learning and weighted sparse coding to enhance the discrimination of sparse coefficients and weaken the influence of radiometric changes. Then, the sparse representations over the learned discriminative dictionary are utilized to measure the dissimilarity between image patches. Semi-global cost aggregation and postprocessings are finally enforced to further improve the matching accuracy. Extensive experimental comparisons demonstrate that: the proposed matching cost outperforms traditional matching costs, the discriminative dictionary learning model is more suitable than previous dictionary learning models for stereo matching, and the proposed stereo method ranks the third place on the Middlebury benchmark v3 in quarter resolution up to the submitting, and achieves the best accuracy on 30 classic stereo images.