A Novel Segmentation Algorithm Based on Level Set Approach with Intensity Inhomogeneity: Application to Medical Images

Abstract

Most image segmentation techniques are based on the intensity homogeneity. Intensity inhomogeneity frequently occurs in real word image like medical images. This type of images fails to provide accurate segmentation result; this is challenging issue. In this paper, we present a robust region-based method for image segmentation, which is able to deal with intensity inhomogeneities in the images. This method derives a local intensity clustering property of the image based on the model of images with intensity inhomogeneities, and then defines a local clustering criterion function in the neighborhood of each point. In a level set formulation, this criterion defines energy in terms of the level set function and a bias field. The level set functions represent a partition of the image domain whereas a bias field accounts for the intensity inhomogeneity of the image. Therefore by minimizing this energy, our proposed method is able to simultaneously segment the image and estimate he bias field, and the estimated bias field can be used for intensity inhomogeneity correction. Finally, experiments on some medical images have demonstrated the efficiency and robustness of the presented model.