Accurate and efficient image segmentation and bias correction model based on entropy function and level sets

With the development of society, image segmentation plays a pivotal role in real-world life. However, the images we obtain are often distorted or contaminated by noise and shade, causing low construct and weak boundaries. In this paper, we propose a new model suitable for segmenting and correcting images with inhomogeneous intensity simultaneously. According to the characteristics of the entropy function, we use it as the coefficient of the global and local terms of the energy functional, related to the intensity of the image. Most importantly, the compression function is added to control the global item and the local item in the same range, reducing parameters to be tuned. In addition, our model can be extended to the multi-objective model and the colour model to deal with a more difficult situation. Furthermore, by utilizing the split Bregman method to solve the energy functional, we reduce computational cost and maintain the convergence of the algorithm. A large number of experimental results demonstrate the superiority of our model over previous models. Our model has promising performance for the segmentation and correction of inhomogeneous medical images and colour images.