The accuracy and validity of the traditional image correlation-matching algorithms are degraded by the presence of the gray change, image clutter, object variation and occlusion. The reason of the degradation is that the method of adding the differences of the pixels′ gray value is used according to the corresponding position as degree of mismatching. The regions formed by image segmentation is a higher-level representation than pixels. Based on the technology of adaptive multi-thresholds image segmentation through potential function clustering, a new algorithm named weighted region template correlation (WRTC) is introduced, which constitutes a method to measure the various membership degree according to different regions. The results show that this new algorithm is robust to more image distortions and has its application value in image matching.