Similarity determination method and similarity determination device

Abstract

The invention relates to a similarity determination method. The method comprises the steps of obtaining an original template image and an original to-be-detected image; determining edges in the original template image and the original to-be-detected image, and extracting the to-be-detected image from the original template image and the original to-be-detected image according to the edges; determining a to-be-detected connected region in the to-be-detected image according to the neighborhood gray average value of the pixels in the template image and the to-be-detected image; obtaining a template contour image of the template communication area and a to-be-detected contour image of the to-be-detected communication area; and calculating the similarity between the template contour image and the to-be-detected contour image to serve as the similarity between the original template image and the original to-be-detected image. According to the method and the device, the similarity between theoriginal template image and the original to-be-detected image is determined on the basis of the contour, and the contour is not greatly different due to different types of the original template imageand the original to-be-detected image, so that the similarity between the medical images can be accurately determined.

Description

technical field [0001] The present disclosure relates to the technical field of medical images, and in particular to a similarity determination method, a similarity determination device, electronic equipment, and a computer-readable storage medium. Background technique [0002] Image similarity measurement is a very important basic problem in the field of computer vision. Commonly used image similarity calculation algorithms mainly include the following two types: [0003] Histogram matching: Calculate the histograms for the two images separately, and use the normalized correlation coefficient (including Bhattacharyachian distance, histogram intersection distance, etc.) of the two histograms as the similarity. This type of similarity measurement is based on the difference in simple mathematical vectors, and the difference in image resolution can be ignored for similarity calculation, and the calculation amount is small. [0004] Feature point extraction: extract the feature...

AI Technical Summary

Problems solved by technology

However, based on the histogram matching method, the comparison is only the distribution probability of the gray value of the image pixel, which cannot reflect the change of the internal information of the image, so the similarity between medical images cannot be accurately determined.

[0006] Moreover, medical images that need to be compared with similarity can be obtained through a variety of imaging methods, and the feature information that can be extracted for images obtained by different imaging methods will be different. If the feature point extraction method is used, even if the actual similarity Two very high images will also have a low degree of similarity due to different feature information, and the number of feature points and the way of selecting feature points will also affect the calculation of similarity, so it is impossible to accurately determine the difference between medical images. similarity

Images