Comparison of Skin Lesion Segmentation Methods

S. Binu Sathiya and Dr.S.S. Kumar

This paper deals with different segmentation algorithms for skin lesion segmentation. Skin Lesion segmentation is essential for proper diagnosis of skin cancer. Initially, the dermoscopic images are preprocessed for denoising and hair removal. Skin lesions from the preprocessed images are segmented employing Fuzzy C-Means (FCM), Fast FCM and ARKFCM. The segmented results are compared by means of performance measures like Jaccard, Dice, FPR, FNR, Visual overlap. The results show that ARKFCM segments the lesions better than other methods.

Volume 11 | 04-Special Issue

Pages: 102-111