Abstract: Skin cancer is the most dangerous form of cancer. Images of melanoma taken from dermatoscope. Dermoscopy images have great potential in the early diagnostic of malignant melanoma, which is a type of skin cancer, but their interpretation is time consuming and subjective, even for trained dermatologists. Automatic lesion segmentation is an important part of computer-based image analysis of pigmented skin lesions. Melanoma is the deadliest form of skin cancer if left untreated. There is a need for an automated system to assess a patient's risk of melanoma using photographs of their skin lesions. In this project, for improving the segmentation accuracy, a segmentation algorithm based on Probabilistic Fuzzy C Means clustering (PFCM) is proposed. PFCM produces memberships and possibilities simultaneously, along with the cluster centers for each cluster. This approach combines the spatial probabilistic information and the fuzzy membership function in the clustering process. The proposed probabilistic fuzzy c-means method can deal effectively with image segmentation in a noisy environment. Next, regions in the image are classified as normal skin or lesion based on the occurrence of representative texture distributions.

Keywords: Skin cancer, Image segmentation, Dermatoscope, PFCM.