Feature-based correspondence: an eigenvector approach
Image and Vision Computing - Special issue: BMVC 1991
An Eigenspace Projection Clustering Method for Inexact Graph Matching
IEEE Transactions on Pattern Analysis and Machine Intelligence
Distinctive Image Features from Scale-Invariant Keypoints
International Journal of Computer Vision
A Spectral Technique for Correspondence Problems Using Pairwise Constraints
ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision - Volume 2
IEEE Transactions on Pattern Analysis and Machine Intelligence
Negative Samples Analysis in Relevance Feedback
IEEE Transactions on Knowledge and Data Engineering
Geometric Mean for Subspace Selection
IEEE Transactions on Pattern Analysis and Machine Intelligence
Deterministic Column-Based Matrix Decomposition
IEEE Transactions on Knowledge and Data Engineering
Signal Processing
Max-Min Distance Analysis by Using Sequential SDP Relaxation for Dimension Reduction
IEEE Transactions on Pattern Analysis and Machine Intelligence
Gabor-Based Region Covariance Matrices for Face Recognition
IEEE Transactions on Circuits and Systems for Video Technology
Robust Tensor Analysis With L1-Norm
IEEE Transactions on Circuits and Systems for Video Technology
Robust point pattern matching based on spectral context
Pattern Recognition
Hi-index | 0.10 |
Spectral decomposition subject to pairwise geometric constraints is one of the most successful image matching (correspondence establishment) methods which is widely used in image retrieval, recognition, registration, and stitching. When the number of candidate correspondences is large, the eigen-decomposition of the affinity matrix is time consuming and therefore is not suitable for real-time computer vision. To overcome the drawback, in this letter we propose to treat each candidate correspondence not only as a candidate but also as a voter. As a voter, it gives voting scores to other candidate correspondences. Based on the voting scores, the optimal correspondences are computed by simple addition and ranking operations. Experimental results on real-data demonstrate that the proposed method is more than one hundred times faster than the classical spectral method while does not decrease the matching accuracy.