SIFT(scale invariant feature transform) descriptor usually leads to mismatching because it uses the gradient information in the neighborhood of one feature point

when the extracted feature points locate in some similar structures of one image. So a method to correct SIFT mismatching based on a kind of spatial distribution descriptor is proposed. Because the spatial distribution of pixels on the image contour are different aiming at different matching points

each matching point obtained by SIFT can be described again to generate a more distinctive descriptor. Then the method corrects two kinds of mismatching using corresponding correction strategies by the new descriptor. Through the experiments on the real images

the comparing results between the algorithm and RANSAC indicate that the correction method can improve the percentage of correct matching under the condition of remaining the original right matching.