A Binary Level Set Method Based on K-means for Contour Tracking on Skin Cancer Images

A great challenge of research and development activities have recently highlighted in segmenting of the skin cancer images. This paper presents a novel algorithm to improve the segmentation results of level set algorithm with skin cancer images. The major contribution of presented algorithm is to simplify skin cancer images for the computer aided object analysis without loss of significant information and to decrease the required computational cost. The presented algorithm uses k-means clustering technique and explores primitive segmentation to get initial label estimation for level set algorithm. The proposed segmentation method provides better segmentation results as compared to standard level set segmentation technique and modified fuzzy cmeans clustering technique.