Enhanced Clustering Technique for Segmentation on Dermoscopic Images

Recently, there has been a drastically spiral-up amount of patients affected with melanoma the noxious kind of skin cancer. As a result, there is a need to build up a computer-aided diagnostic model to streamline its early detection for effective treatments. It is a challenging task for precise segmentation and classification of pigment of the skin due to the plentiful color and structures developed haphazardly inside the lesion. The present work is demonstrated about the segmentation of skin lesion images using Robust K-Means, the enhanced unsupervised clustering method. This method segregates disparity kinds of melanoma images successfully without using any additional parameters. Statistical methods are used to evaluate the performance of these methods.