Local Active Contour Image Segmentation Method Based on Betweenness Truth Degree Metric

Abstract

The invention discloses a local active profile image segmentation method based on measure of medium true value scale, mainly solving the problem that a current image segmentation method based on active profile cannot be used for segmentation of various images and is not accurate in the segmentation result. The implementation process of the local active profile image segmentation method based on measure of medium true value scale includes the steps: 1) input an image to be segmented, giving an initial profile on the image, and labeling internal and external pixels of the profile on the image; 2) taking the local area of the image, and respectively calculating the internal and external gray scale mean values of the profile on the local area; 3) calculating the similarity between the gray scale values of the pixels on the local area and the internal and external gray scale values of the profile; 4) according to the above result, constructing an energy function; and 5) using image segmentation to optimize the energy function, updating the internal and external pixel labels of the profile so as to drive the internal profile curve of the local area to be evolved to the target boundary, and completing image segmentation when the profile curve arrives at the target boundary. The local active profile image segmentation method based on measure of medium true value scale can improve the precision of image segmentation, can segment various types of images, and can be used for object identification.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to an active contour image segmentation method, which can be used in target recognition for natural images, artificial images, medical images, ultraviolet aurora images, multi-target images, images with uneven gray levels and images with complex backgrounds. segmentation. Background technique [0002] Image segmentation is a key technology in image processing and the first step in image analysis. Therefore, accurate image segmentation plays an important role in image processing. However, since image segmentation is a very difficult and deep task, there is no method suitable for various types of image segmentation so far. [0003] In recent years, active contour methods have been widely used in image segmentation, and have also achieved a certain degree of effect. Active contours are mainly divided into edge-based active contours and region-based active contours. The edge-...

AI Technical Summary

Problems solved by technology

However, this method cannot segment images with uneven gray scale and ultraviolet auroral images with extremely low contrast.

In order to overcome the shortcomings mentioned above, Yang et al. proposed a level set image segmentation method MELS embedded in Markov random field in 2015. This method can obtain better segmentation results when the parameters are optimal, but the search for the most Optimal parameters are not only time-consuming and labor-intensive, but also very difficult

Therefore, this method is not suitable for segmentation of a large number of images

Images