Image segmentation processing method and device

Abstract

The embodiment of the invention provides an image segmentation processing method, and the method comprises the following steps: firstly carrying out the preprocessing of an obtained optical tissue image of a to-be-analyzed object, and obtaining a preprocessing image of the to-be-analyzed object; based on the threshold value matched with each pixel, performing adaptive binarization processing on the pixel to obtain a binarization image corresponding to the preprocessed image; then, determining a boundary contour of the binarized image, determining a main body contour based on an inter-contour relation of the boundary contour, segmenting a main body component image from the preprocessed image according to the main body contour, and finally, performing adaptive clustering processing on the main body component image based on a color space to obtain a color quantization result of the main body image. According to the invention, over-segmentation and under-segmentation conditions existing in a traditional image segmentation method are reduced, the technical effect of reducing the situation that pores are mistakenly divided into pore walls is achieved, and the accuracy of classification and recognition of optical tissues of objects with anisotropic structures can be improved by applying the invention.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image segmentation processing method and device. Background technique [0002] The traditional image segmentation method is generally divided manually, which has the disadvantage of time-consuming and labor-intensive. Now, the use of image processing technology to segment the image has liberated the labor to a certain extent, and at the same time, it has guaranteed the accuracy of the segmentation to a certain extent. accuracy. [0003] For some images with unclear light and dark resolution, such as the optical tissue image of the object, although the light and dark difference between the pore and the pore wall of the object is obvious in the image, in the optical tissue image of the object, there is a certain overlapping relationship among the components of the object tissue. In the prior art, the pores of some images will be misclassified as pore walls, whi...

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present invention provides an image segmentation processing method and device, which solves the problem in the related art that the accuracy of segmenting the optical tissue image and the non-optical tissue image of the object to be analyzed is not high, resulting in over-segmentation and under-segmentation of the image. Happening

Images