Image retrieving comparison method

The invention belongs to the technical field of processing methods of data of specific function, and solves the problems of storing, expressing and organizing images and rapidly and accurately finding the required image information. The image retrieving comparison method is as following: the cleavage line of the maximum cleavage weight value between two contiguous columns or rows of pixel combinations is used for differentiating the image into M image areas which can not be cleaved in an image area R, and then the M image areas are combined into N image areas which can not be combined to calculate the visual interesting weight value of the image areas which can not be combined to determine the distinct area and the indistinct area. A first image step and a second image step are constructed according to the distinct area to determine characteristic points. The characteristic points and a gradient block image are used for determining the main direction of the characteristic points to carry out normalization pro