An Image Segmentation Method Based on Statistical Active Contour and Texture Dictionary

Abstract

The invention proposes an image segmentation method based on statistical active contour and texture dictionary, which is used to solve the technical problem that the active contour model based on sparse texture cannot clearly represent the structure and texture of the image, and the model has a large amount of calculation. The invention establishes a level set energy function and a level set updating equation based on the Gaussian mixture distribution under the statistical framework. The steps are: firstly, use the dictionary learning algorithm to obtain the binary sparse matrix; secondly, initialize the level set and obtain the probability label through the linear transformation of the binary sparse matrix; then use the probability label to obtain the statistical parameters of the current segmentation; then combine the current level set functions, probability labels, and statistical parameters to predict new segmentation curves. The level set function evolves driven by the probability labels, which are updated from the level set based binary labels by line transformation. Compared with traditional methods, the present invention can effectively segment complex textures while greatly reducing calculation costs.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image segmentation method based on statistical active contour and texture dictionary, which can efficiently complete complex texture image segmentation and can be widely used in the field of image analysis. Background technique [0002] Active contour model is a class of methods for object segmentation in the field of computer vision, and has an irreplaceable position in applications such as medical image analysis. This model is usually implemented by minimizing the energy function and evolving the level set function into a time-dependent partial differential equation with the smoothness of the zero level set constraining the bounds of the objects coupling the image data. Most of the existing research focuses on designing a class of effective energy functionals, bringing various constraints and constructions about images into it, and proposing some effective methods. ...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies in the above-mentioned background technology, the present invention proposes an image segmentation method based on statistical active contour and texture dictionary, which solves the problem that the active contour model based on sparse texture cannot clearly represent the structure and texture of the image, and the model calculation amount big technical problem

Images