Fingernail detection in hand images using difference of nail-color pixels' density between vicinity areas of fingernails and skin

No research group has developed practical “fingernail detection” technologies in hand images including palm areas and/or wrist rotation angles, because the human palm has the similar color with fingernail areas. We propose in this paper a new and practical fingernail detection method, using the differences in nail-color pixels' density between vicinity areas of fingernails and skin. Experimental results showed that our system practically works in fingernail detection in hand images at least 80 percent of correction, even though users rotate their hands from -90 to -40 degrees and from 40 to 90 degrees, i.e. excepting that hand is not facing upward or downward.