Abstract
Despite progress made in the accuracy and robustness of the dense matching technique in past years, efficient occlusion detection remains an open problem. In this paper, we present a two-step occlusion detection method to remove false matches in dense matching fields. First, a statistical dense matching method is developed by considering the correspondence between the grids to identify most occlusion regions. Second, to handle the potential misjudgment match in the occlusion boundary, a double-threshold filtering method is first used to reduce the noise in the grid image, which ensures that the gradient operator can accurately extract the boundary grid in the grid image; then, misjudgment matches in the boundary grid region are corrected based on the triangulation with descriptors. The results of the experiments comparing the proposed method and existing occlusion detection methods by, respectively, using the MPI-Sintel and KITTI datasets’ test sequence show that the proposed method has higher accuracy and better robustness.
Highlights
Dense matching aims at determining the dense correspondence between two consecutive frames
PERFORMANCE ANALYSIS To get a better understanding of the performance of the proposed method, we evaluate the impact of different input matches, the procedure of misjudgment correction and the parameters of grid size m on the results of occlusion detection of the proposed method
It contains two major steps: the coarse detection based on the statistical matching constraint and the fine detection based on triangulations of an image
Summary
Dense matching aims at determining the dense correspondence between two consecutive frames. The results reveal that most false matches are obviously located in the occluded region compared to the ground truth of occlusions In this context, most matching algorithms [14]–[22] depend on the forward-backward consistency check method to detect occlusions. The flow of mutually corresponding pixels in the occluded region may be zero, while it will be considered as true correspondences using cross-checking; another case exists in the low-texture regionmultiple pixels may correspond to the same pixel-which means that there are many erroneous mutual correspondences and they are difficult to detect using cross-checking This method must calculate the bidirectional flow, which is highly time-consuming for some advanced matching algorithms. Aiming to solve those pre-mentioned problems, this paper combines the statistical matching constraint and triangulations of an image into a two-step occlusion detection method to remove false matches in dense matching fields.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
