“…In most of the existing literature, certain preprocessing steps (color correction/brightness, contrast enhancement) played a significant role in an accurate border detection, which leads to accurate classification (Nasir et al, ). Lately, several research studies are giving a special attention on color correction obtained or color space transformations (A. C. F. Barata, ; Fahad, Ghani Khan, Saba, Rehman, & Iqbal, ; Iqbal, Khan, Saba, & Rehman, , Iqbal, Ghani, Saba, & Rehman, ; Mughal, Muhammad, Sharif, Rehman, & Saba, ). Similarly, several machine learning techniques are also adopted in literature for lesion detection and classification such as adaptive thresholding, k‐means clustering (Agarwal, Issac, Dutta, Riha, & Uher, ), fuzzy c‐means (FCM) clustering (Masood & Al‐Jumaily, ), inutile fragment removal methods (Majtner, Lidayova, Yildirim‐Yayilgan, & Hardeberg, ), region growing methods (Mohamed et al, ), gradient vector flow (GVF) snakes (Flores & Scharcanski, ), Markov random field, convolution autoencoder NN (Chen, Shi, et al, ), and region fusion based multimodal technique (Yuan, Situ, & Zouridakis, ).…”