Reliable Template Matching Using First Derivation Edge-Direction

Abstract

This paper proposes a new and more reliable method for template matching that is using the first derivation edge-direction. A template image is separated into local block areas and sum of inferior angle between edge-directions of target image and edge-directions of template image are calculated in all block areas. For each block, a coordinate that is the highest similarity of edge-directions gets a vote and a target coordinate is decided by its result. The experimental results show that the proposed method was robust to the debased target image with the conventional methods.