Skin diagnosis algorithm for analysis of skin images captured from a fingerprint sensor

This paper presents a skin diagnosis algorithm installable in cell phones that have fingerprint identification function. The algorithm can analyse low contrast skin images captured from a fingerprint sensor based on two dimensional linear prediction analysis. The authors' devices can be smaller than conventional skin diagnosis devices having a camera, without an optical lens or a focus adjustment apparatus. Especially, it can be installed in fingerprint identification devices as a bonus application. The authors' algorithm presented in this paper has been embedded in Pantech PG-6200 mobile phone.