“…15,28,75,91,98,101,118,136,143 The features extracted from the skin map include average skin probability of the whole image, height and width of the largest region of skin, average skin probability inside the largest skin region, number of skin regions in the image, distance from the centroid of the largest skin region to the center of the image, edge of connected components of skin, percentage of pixels detected as skin, and number of connected components of skin. The extracted skin regions and feature vectors are useful for image classi¯cation and will help the generic algorithm to decide.…”