Use Cross-correlation To Discern Content of An Image
Yang Hong-y · Microcomputer Information · 2009
It is a difficulty to discern the content of an image at the filed of Image Processing.The methods now used in most are implemented with the algorithm of Neural Network.This paper has made an exploration of determining the content of an image through another way that used the Cross-correlation of two images.To achieve the purpose,this algorithm computes the relation of two images to declare the similarity of them.Because of that the value is referenced with the size of these two images,a new concept named Deviation of Cross-correlation was produced in this paper.The Deviation of Cross-correlation made it more exact to descript the similarity.The experiment shows that this algorithm can discern the content of simple images exactly.