Active Contour Image Segmentation Method Based on Local Fitting Image

Abstract

The invention discloses an active contour image segmentation method based on a local fitted image. The method comprises the following major steps: providing a new local squared fitted image according to existing active contour models to approximate square of an original image and highlight a target region in the image; then, constructing a new energy functional based on two different local fitted images by utilizing the squared fitted image and the local fitted image in an LIF model; and meanwhile, in order to ensure curve smoothness and reasonable curve length of the segmented result, introducing two different regular terms to the energy functional, thereby improving segmentation accuracy and timeliness; and finally, introducing the segmentation algorithm to a variational level set solution framework, and realizing full-automatic extraction of a target contour. The method can enable the target object in the image to be extracted accurately under the condition of different mage backgrounds and poor gray uniformity.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an active contour image segmentation method based on locally fitted images. Background technique [0002] Image segmentation is an image processing technology that divides the image area into different sub-modules according to the grayscale characteristics. It plays an important role in the fields of image understanding and analysis, computer vision, and target detection, so it has been widely used and deeply studied. . In the field of image segmentation, there are various image segmentation algorithms, such as threshold segmentation method, watershed algorithm and active contour algorithm; among them, the active contour algorithm has obtained a large number of It is still a relatively active research branch in the field of segmentation. However, the active contour segmentation algorithm is closely related to the characteristics of the contour model, and effectively constructing...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the above technical problems, the present invention proposes an active contour image segmentation method based on locally fitted images, which solves the problem of rough segmentation results, low accuracy and incomplete segmentation caused by complex background information and serious gray level inhomogeneity in the image. and other issues, so as to accurately and completely extract target objects with different shapes and sizes in the image

Images