Level Set Method for Image Segmentation in the Presence of Intensity Inhomogeneity and Noise

It is a great challenge to segment images distorted by severe noise and inhomogeneity. This paper presents a new level set method to segment images in presence of intensity inhomogeneity and noise. To mitigate intensity inhomogeneity, an improved multi-scale average filter is proposed to preprocess the image. With kernel metric and patch similarity measure, local and global image information is utilized to construct energy functional. In addition, the regularization term based on counting operator is incorporated into the final energy functional. Experimental results show that the proposed method has good performance for image segmentation in the presence of intensity inhomogeneity and noise.