Algorithm 2. GASISUK Algorithm.

Let $I_p$ = preprocessed image (see Section 2.2.1) Let $o_w$ and $o_h$ be the original width and height of $I_p$ respectively Let $č$ = dominant color of $I_p$ If $č>\left(20, 20, 20\right)$   $B_r$ = predominantly light else   $B_r$ = predominantly dark if $B_r$ is predominantly light   assume gamma factor $g_f$ of 0.75   $I_{\gamma }$ = gamma of $I_p$   $I_{{\gamma }_g}$ = gray of $I_{\gamma }$   $I_{{\gamma }_t}$ = binary threshold of $I_{{\gamma }_g}$   $I_{{\gamma }_m}$ = morphology opening of $I_{{\gamma }_t}$ using a 7 × 7 matrix   $I_{{\gamma }_b}$ = initial segmentation using bitwise_AND of $I_{{\gamma }_m}$ and $I_p$   If $I_{{\gamma }_b}$ == [0] … depicting black image     $I_{{\gamma }_b}$ = $I_p$ else   assume gamma factor $g_f$ of 1.7   $I_{{\gamma }_b}$ = gamma of $I_p$ $k_p$ = keypoint features of $I_{{\gamma }_b}$ using ORB $k_{p_c}$ = clustering of keypoint features using DBSCAN $C_i$ = contour sketch of clustered keypoints $I_{{\gamma }_{f_1}}$ = first-level contour filtering   filter-off contours satisfying below characteristics     ∗ contour is identified as inner contour     ∗ contour has an area of less than 32.000     ∗ width and height of contour less than minimum width (variable, value of $0.2o_w$ or fixed value of 200) and minimum height (variable value of $0.2o_h$ or fixed value of 200)   morphology erosion of contours $I_{{\gamma }_{f_2}}$ = second-level contour filtering   filter-off contours satisfying below characteristics     ∗ ccontour has an area of less than 25.000 $I_s$ = segmented lesion image using filled mask of $I_{{\gamma }_{f_2}}$