Robust Wide Baseline Feature Point Matching Based on Scale Invariant Feature Descriptor

Authors:
Sicong Yue;Qing Wang;Rongchun Zhao
Affiliations:
Northwestern Polytechnical University, Xi'an, P.R. China;Northwestern Polytechnical University, Xi'an, P.R. China;Northwestern Polytechnical University, Xi'an, P.R. China
Venue:
ICIC '08 Proceedings of the 4th international conference on Intelligent Computing: Advanced Intelligent Computing Theories and Applications - with Aspects of Theoretical and Methodological Issues
Year:
2008

Citing 7
Cited 0

A robust technique for matching two uncalibrated images through the recovery of the unknown epipolar geometry

Artificial Intelligence - Special volume on computer vision
Modal Matching for Correspondence and Recognition

IEEE Transactions on Pattern Analysis and Machine Intelligence
A direct method for stereo correspondence based on singular value decomposition

CVPR '97 Proceedings of the 1997 Conference on Computer Vision and Pattern Recognition (CVPR '97)
Wide Baseline Stereo Matching

ICCV '98 Proceedings of the Sixth International Conference on Computer Vision
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
A Performance Evaluation of Local Descriptors

IEEE Transactions on Pattern Analysis and Machine Intelligence
Wide-baseline multiple-view correspondences

CVPR'03 Proceedings of the 2003 IEEE computer society conference on Computer vision and pattern recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper proposed a robust point matching method for wide baseline in order to achieve a large number of correct correspondences and high accuracy. To cope with large variations of scale and rotation, a feature descriptor, that is robust to scale and view-point, is added to the feature detection phase and it is included in the equations of the correspondence matrix that is central to the matching algorithm. Furthermore, the image window for normalized cross correlation is modified with adaptive scale and orientation. At the same time we remove from the matrix all the proximity information about the distance between points' locations which is the source of mismatches. Thus, the proposed algorithm is invariant to changes of scale, rotation, light and partially invariant to viewpoint. Experimental results show that the proposed algorithm can be used for large scene variations and provide evidence of better performance.