Abstract
We present a novel algorithm called crosspatch-based rolling label expansion for accurate stereo matching. This optimization-based approach can effectively estimate the 3D label of each pixel from huge and infinite label space and then generate a continuous disparity map. The algorithm has two obvious characteristics when compared with the traditional label expansion algorithms. The first feature is the cross-based multilayer structure, where each layer contains a series of cross patches with adaptive shapes, reflecting the edge structure of objects on the image. Besides, such cross patches are non-overlapping and independent, satisfying the submodular property for employing graph cuts. The second feature is the rolling optimization, that firstly generates new label proposal by expanding candidate labels within cross patches, then globally updates labels for the whole image using a proposed rolling move. The experimental results show the high matching accuracy of our method, both in pixel level and subpixel level. According to the latest ranking list of Middlebury 3.0 benchmark, our method is one of the best stereo matching algorithms.
Highlights
Stereo matching is one basic task of computer vision, whose goal is to estimate disparity when inputting an image pair [1], which is widely applied in navigation [2], 3D construction [3] and virtual viewpoint imaging [4]
Estimating accurate and continuously varying disparities is the key to stereo matching, and many related algorithms have been proposed and they can output dense and continuous disparity map [5]–[7]
In this paper, we have presented a novel global stereo matching approach for the accurate estimation of 3D label and sub-pixel disparity
Summary
Stereo matching is one basic task of computer vision, whose goal is to estimate disparity when inputting an image pair [1], which is widely applied in navigation [2], 3D construction [3] and virtual viewpoint imaging [4]. Estimating accurate and continuously varying disparities is the key to stereo matching, and many related algorithms have been proposed and they can output dense and continuous disparity map [5]–[7]. These methods estimate successive disparities by assigning continuous 3D labels [5] to neighboring pixels, and mapping the labels to disparities. Since the 3D label space (R3) is huge and infinite, two difficulties occur during label assignment, namely, how to reduce the search space of candidate labels to decrease the computational complexity, and how to assign labels accurately
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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
