The Non-local Mumford-Shah-TV Model for Texture Image Segmentation and Its ADMM Algorithm

To improve the accuracy of texture image segmentation,and overcome the difficulties of description of texture components and their boundaries for texture image segmentation,we propose a combined non-local Mumford-Shah-TV model under variational framework making use of the properties of TV(total variation)regularizer in image structure detection and non-local operators in texture descriptions.Meanwhile,a binary label function is used to divide different regions in the model.In order to improve computational efficiency,ADMM(alternating direction method of multipliers)algorithm is designed to decompose the original problems into a series of optimization sub problems.Some numerical examples are presented finally to demonstrate that our model is better in texture image segmentation and accuracy.