Template matching algorithm based on gradient search

Abstract

In order to improve the tracking performance of the conventional template matching method, a novel template matching method is proposed. The round template is used to ensure the rotation invariance of the method. The differential image can be got according to the original image in the polar coordinate system. The matching value is composed of the grey matching value and the differential matching value. The differential information contains the detail information of the template, which can enhance the accuracy of recognition. The gradient search is used to reduce the matching times, which can ensure the real time of the tracking system. Simulation results show that the template matching method proposed in this paper has better recognition and tracking performances than the conventional method.