Implementation of memory-free skin-detection system for mobile display device

This paper proposes a simple yet extremely good technique for skin color extraction in images. This technique also allows for changeable mode options so that significantly better skin detected images are able to be obtained. The correlation between Cb and Cr is analyzed, and then a novel human skin color space is created to process 1-D skin detection properly. This system uses only a distribution of Cb and Cr from human skin, which is transformed from RGB color space, so that it is able to be applied to relatively small hardware, like mobile apparatuses.