No abstract

No abstract

Template matching is considered to be a powerful image processing tool that is particularly useful in industrial inspection problems. In industrial inspection large numbers of objects have to be processed in short time. The processing itself is relatively easy. The defects that are to be found are known and ran be modelled with black and white patters. Binary images are usually sufficient. In this paper the binary template matching operation is explained and some design rules for these templates are presented. Furthermore some algorithms are presented that allow these templates to be learned from teaching by showing methods.However any template matching hardware offering this capability can be used for these techniques. Furthermore it is assumed that the images to be inspected are greyvalue images that can be segmented into a binary objectbackground image by simply thresholding the greyvalue image. The outline of this paper is as follows. In section I1 the template matching operation will be explained. In section 111, rules of thumb for the design of templates are given and solutions to some problems that arise in automatic learning of templates are discussed. Finally section IV is devoted to cellular logic operations that can be done with the same template matching hardware. I1The template matching operation and the RTR architecture.

No abstract

No abstract