A characteristic that Normalized Cross Correlation Operating (NCCO) results independent of image gray levels in pattern matching for image feature positioning was presented.In the field of microscopic vision research,the majority of images can be approximated to an image with three gray levels,or even binary image.So with this attractive feature,a method of improved pattern matching algorithms was presented.Experimental results show that the characteristic of gray level independence can give guidance to designing template images,reducing time consumption of calculations,and improving illuminating conditions.