Mobile robot global localization using just a visual landmark
ISPRA'06 Proceedings of the 5th WSEAS International Conference on Signal Processing, Robotics and Automation
A novel approach to image matching using convex layers
ISCGAV'06 Proceedings of the 6th WSEAS International Conference on Signal Processing, Computational Geometry & Artificial Vision
Hi-index | 0.00 |
The paper discusses the 2DE gel image matching problem that occurs when define correspondence to detect spots position and their pairing correlations. It is expected that a proportion of these correspondences will in fact be mismatches. The feature description and matching method are used to find correspondences; these are refined by the algorithm that automatically computes a homography between two images. With the refined correspondences, the RANSAC is designed to deal with exactly this situation - to estimate 1) the homography and also 2) a set of inliers consistent with this estimate and 3) outliers. Experimental results show that the algorithm is efficient and it improves the robustness and accuracy of the automatic image matching.