Skin-pigmentation-disorder detection algorithm based on projective coordinates

Our skin is the first line of defense from external environment. It covers whole body and protects the tissues and organs from various external assaults. Skin also keeps a stable internal environment for body’s metabolism, like regulating temperature and saving necessary water. Furthermore the nerve of skin gives us tactile sensation, which makes us to better adapt to our living environment. When skin diseases occur, the partial functionality of skin cannot be accomplished. One of fatal diseases is skin cancer. With skin cancer progression, the color of skin will change on different stage. Paper proposes an algorithm to detect and quantize this kind of visual change. The program tries to assist people to diagnosis skin cancer, even self-exam. The proposed algorithm mainly contains three parts. Initially paper segments the skin area from images. Then two main components of skin color are retrieved by ICA. Finally paper estimates if there is pigmentation disorder through comparing the ratio of two main components, and a novel comparing model, projective coordinates is used in this process.