Improved local information-based CV model image segmentation method

Abstract

The invention discloses an improved local information-based model image segmentation method, and mainly solves the problem of false segmentation caused by non-ideal non-uniform grayscale image segmentation effect of an improved CV model and relatively low robustness of a local information-based model to an initial contour at present. The method is implemented by comprising the steps of inputting an original image and setting the initial contour; setting default parameters and important parameters; combining a global grayscale fitting value and a local grayscale fitting value of an improved kernel function into a new grayscale fitting value of a weighted target and a background; obtaining gradient descent flow by utilizing an energy functional of a CV model in which penalty terms are introduced; and evolving a level set function according to a level set iteration formula, and through iteration, outputting a segmentation result. According to the method, a non-uniform grayscale image is effectively segmented; the robustness to the initial contour is enhanced; the initial contour is converged to a target contour more quickly; compared with other related models, the method has higher segmentation precision and efficiency; and the method is used for segmentation of artificially synthesized images, non-uniform grayscale images and infrared images.

Description

technical field [0001] The invention belongs to the field of image information processing, and mainly relates to image segmentation, in particular to an image segmentation method based on a CV model of improved local information, which is used for artificially synthesized images, uneven gray scale images and infrared image segmentation. Background technique [0002] Image segmentation is a fundamental problem in the fields of computer vision and image processing. The purpose of image segmentation is to divide the image area into several disjoint sub-areas, and a certain property of the image on each sub-area is consistent. [0003] Active Contour Model (ACM) has attracted people's attention because of its ability to handle local discontinuous edges better. The level set method replaces the parametric curve in the ACM method with the level set function defined in the high-dimensional space, and successfully solves the problem of topological structure change in the curve evol...

AI Technical Summary

Problems solved by technology

The most famous of these is the CV model, which uses the grayscale statistical information of the image and no longer relies on the gradient information of the image to segment images with fuzzy boundaries, weak boundaries, and discontinuous boundaries, especially for images with uniform gray levels of the target and background. Segmentation effect, but this method cannot segment images with uneven gray levels, and the evolution speed is slow when segmenting larger-sized images or noisy images

[0005] In view of the shortcomings of the CV model, Li Chunming et al. proposed the LBF model based on local information, and changed the global binary fitting energy functional of the CV model to the local binary fitting energy functional based on the kernel function, which better solved solves the segmentation problem of uneven grayscale images, but the model is sensitive to the initial contour, different initial contour positions will affect the segmentation results, and the evolution speed of the LBF model is difficult to meet the high real-time requirements of the application

In the paper, Wang Li and others combined the energy functionals of the CV model and the LBF model into a new energy functional, which solved the problem of image segmentation with uneven gray levels to a certain extent, but compared with a single model, the calculation amount was larger

Combining local and global information, Wang Xiaofeng and others proposed a local CV model (LCV). By making a difference between the convolved image and the original image, the gray contrast between the target and the background can be enhanced, and the image with uneven gray scale can be segmented. The segmentation speed and The number of iterations is faster than LBF, but because it does not use the relationship between the gray value of local pixels and is limited by the size of the convolution window, the segmentation effect for the complex gray image of the background is not ideal

[0006] In summary, the segmentation speed of the model based on global information is fast, but it cannot successfully segment images with uneven gray levels

Models based on local information tend to fall into local minima, leading to over-segmentation

The hybrid model combining global and local information can better segment grayscale heterogeneous images, but the calculation is more complicated

Images