Feature Matching for Remote-Sensing Image Registration via Neighborhood Topological and Affine Consistency

Feature matching is a key method of feature-based image registration, which refers to establishing reliable correspondence between feature points extracted from two images. In order eliminate false matchings the initial matchings, we propose simple and efficient method. The principle our maintain topological affine transformation consistency among neighborhood matches. We formulate this problem as mathematical model derive closed solution with linear time space complexity. More specifically, can remove mismatches thousands hypothetical correspondences within few milliseconds. conduct qualitative quantitative experiments on different types remote-sensing datasets. experimental results show that general, it deal all kinds pairs, whether rigid or non-rigid deformation pairs various shadow, projection distortion, noise, geometric distortion. Furthermore, orders magnitude faster more accurate than state-of-the-art methods be used for real-time applications.