Robust template matching method based on minimum maximum circular similarity measurement

To solve the problem of image matching under complex conditions including strong noise corruption, partial occlusion etc., a robust image matching algorithm was presented. Distance between points was redefined by use of circular restriction. Based on classical Hausdorff distance and the new definition of distance, a new similarity measure, i.e., minimum maximum circular measure was defined. By applying the new measure to scene matching, and by using circular windows in the searching process, a minimum maximum circular measure based robust image matching approach was proposed. The algorithm may be used to match images with rotation, intensity contrast change, noise corruption, partial occlusion, and intensity saturation etc. The feasibility of the algorithm is analyzed theoretically, and the effectiveness of the algorithm is illustrated by multiple experiments.