Feature Matching under Region-Based Constraints for Robust Epipolar Geometry Estimation

Outlier-free inter-frame feature matches are important to accurate epipolar geometry estimation for many vision and robotics applications. We discover a set of high-level geometric and appearance constraints on low-level feature matches by exploiting reliable region matching results. A new outlier filtering scheme based on these constraints is proposed that can be combined with traditional robust statistical methods to identify outlier feature matches more reliably and efficiently. The proposed filtering scheme is tested in an real application of outdoor mobile robot navigation based on far-field scenes and of scenes that contain repeated structures.

[1]  Rafael C. González,et al.  Digital image processing using MATLAB , 2006 .

[2]  Andrew Zisserman,et al.  Viewpoint invariant texture matching and wide baseline stereo , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[3]  Lutz Priese,et al.  Fast and Robust Segmentation of Natural Color Scenes , 1998, ACCV.

[4]  Luc Van Gool,et al.  Wide-baseline multiple-view correspondences , 2003, 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings..

[5]  Hai Tao,et al.  A global matching framework for stereo computation , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[6]  Cordelia Schmid,et al.  A Performance Evaluation of Local Descriptors , 2005, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Robert C. Bolles,et al.  Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[8]  Andrew Zisserman,et al.  Multiple View Geometry in Computer Vision (2nd ed) , 2003 .

[9]  H. Opower.  Multiple view geometry in computer vision , 2002 .

[10]  Christopher G. Harris,et al.  A Combined Corner and Edge Detector , 1988, Alvey Vision Conference.

[11]  Luc Van Gool,et al.  Wide-baseline stereo matching with line segments , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[12]  Wei Zhang,et al.  A New Inlier Identification Scheme for Robust Estimation Problems , 2006, Robotics: Science and Systems.

[13]  Volker Rehrmann.  Object Oriented Motion Estimation in Color Image Sequences , 1998, ECCV.

[14]  Bernhard P. Wrobel,et al.  Multiple View Geometry in Computer Vision , 2001 .

[15]  Mark Ross,et al.  Segment Cluster Tracking , 2004, CGIV.

[16]  Jianbo Shi,et al.  Image Matching via Saliency Region Correspondences , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[17]  Zhengyou Zhang.  Estimating Motion and Structure from Correspondences of Line Segments between Two Perspective Images , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  Jiri Matas,et al.  Robust wide-baseline stereo from maximally stable extremal regions , 2004, Image Vis. Comput..

[19]  Matthijs C. Dorst.  Distinctive Image Features from Scale-Invariant Keypoints , 2011 .

[20]  Alexander Hornberg.  Handbook of Machine Vision , 2006 .

[21]  Philip H. S. Torr,et al.  The Development and Comparison of Robust Methods for Estimating the Fundamental Matrix , 1997, International Journal of Computer Vision.