Image Processing Basics: Contour-based Pattern Matching

Pattern Matching is a method to search for a given template or so-called model within an image. The procedure results in a similarity image with a score figure at every pixel. The higher the score, the higher will be the probability for a real match between the template and the image region in the vicinity of this individual pixel. There are some matching methods based upon the contours of objects rather than grey-level templates. These algorithms process edges in images and use quite abstract relations between geometric primitives to check geometric models of objects.