Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/13321
Title: A triangulation-based hierarchical image matching method for wide-baseline images
Authors: Wu, B 
Zhang, Y
Zhu, Q
Issue Date: 2011
Publisher: American Society for Photogrammetry and Remote Sensing
Source: Photogrammetric engineering and remote sensing, 2011, v. 77, no. 7, p. 695-708 How to cite?
Journal: Photogrammetric engineering and remote sensing 
Abstract: This paper presents a triangulation-based hierarchical image matching method for wide-baseline images. The method includes the following three steps: (a) image orientation by incorporating the SIFT algorithm with the RANSAC approach,(b) feature matching based on the self-adaptive triangle constraint, which includes point-to-point matching and subsequent point-to-area matching, and (c) triangulation constrained dense matching based on the previous matched results. Two new constraints, the triangulation-based disparity constraint and triangulation-based gradient orientation constraint, are developed to alleviate the matching ambiguity for wide-baseline images. A triangulation based affine-adaptive cross-correlation is developed to help find correct matches even in the image regions with large perspective distortions. Experiments using Mars ground wide-baseline images and terrestrial wide-baseline images revealed that the proposed method is capable of generating reliable and dense matching results for terrain mapping and surface reconstruction from the wide-baseline images.
URI: http://hdl.handle.net/10397/13321
ISSN: 0099-1112
EISSN: 2374-8079
Appears in Collections:Journal/Magazine Article

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

31
Last Week
7
Last month
0
Citations as of Jul 7, 2017

WEB OF SCIENCETM
Citations

28
Last Week
1
Last month
1
Citations as of Aug 14, 2017

Page view(s)

46
Last Week
1
Last month
Checked on Aug 13, 2017

Google ScholarTM

Check



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.