Identifying any type of disease at an earlier stage became an essential thing in the medical field. Cancer especially skin cancer is a major disease that affects humans at a higher rate in the current scenario. Early detection is a significant method to prevent death; treating at an earlier stage leads to the cure of cancer. Researchers proposed a different technique to detect skin cancer. This paper proposed an enhanced DCNN for classifying melanoma (skin cancer) as benign and malignant. The proposed method involves preprocessing, and enhanced fuzzy clustering for detecting melanoma, followed by enhanced DCNN(E-DCNN) for the classification of dermoscopy images. Enhanced fuzzy clustering is a method that combines modified region grow image segmentation along with fuzzy Kmeans clustering to provide more accurate classified results than other methods proposed by researchers.

DCNN, Fuzzy K-means clustering, Modified Region-grow segmentation, Dermoscopy, Melanoma, Classification.