Traditional matching algorithms based on intensity is slow,and sensitive to the change of intensity.To improve the speed of image matching,an algorithm based on searching strategy along the border and vector correlation are presented.From the viewpoint of vector correlation in high dimensional space,the difference of two images was analyzed.It uses vector to present the image information and calculates the angle of two vectors as the metric of correlation of candidate and template.After the simulation experiment of the introduced algorithm,it was theoretically analyzed and discussed，and the experiment show that this approach has high speed and accuracy in image matching.