Please use this identifier to cite or link to this item:
http://hdl.handle.net/1942/18341
Title: | Iterative Refinement for Real-time Local Stereo Matching | Authors: | DUMONT, Maarten GOORTS, Patrik MAESEN, Steven DEGRAEN, Donald BEKAERT, Philippe LAFRUIT, Gauthier |
Issue Date: | 2014 | Publisher: | IEEE Xplore | Source: | Proceedings of 3D Stereo Media | Series/Report: | International Conference on 3D Imaging (IC3D) | Abstract: | We present a novel iterative refinement process to apply to any stereo matching algorithm. The quality of its disparity map output is increased using four rigorously defined refinement modules, which can be iterated multiple times: a disparity cross check, bitwise fast voting, invalid disparity handling, and median filtering. We apply our refinement process to our recently developed aggregation window method for stereo matching that combines two adaptive windows per pixel region [2]; one following the horizontal edges in the image, the other the vertical edges. Their combination defines the final aggregation window shape that closely follows all object edges and thereby achieves increased hypothesis confidence. We demonstrate that the iterative disparity refinement has a large effect on the overall quality, especially around occluded areas, and tends to converge to a final solution. We perform a quantitative evaluation on various Middlebury datasets. Our whole disparity estimation process supports efficient GPU implementation to facilitate scalability and real-time performance. | Keywords: | local stereo matching; disparity estimation; iterative disparity refinement; edaptive aggregation; Windows, real-time | Document URI: | http://hdl.handle.net/1942/18341 | ISBN: | 9781479980239 | DOI: | 10.1109/IC3D.2014.7032581 | ISI #: | 000380557000010 | Rights: | (c) 2014 IEEE | Category: | C1 | Type: | Proceedings Paper | Validations: | ecoom 2017 |
Appears in Collections: | Research publications |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
mdumont-IC3D14.pdf | 502.78 kB | Adobe PDF | View/Open |
SCOPUSTM
Citations
1
checked on Sep 5, 2020
Page view(s)
130
checked on Aug 6, 2023
Download(s)
278
checked on Aug 6, 2023
Google ScholarTM
Check
Altmetric
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.