Imaging simulated smeared fingers with a sensor based on scattered-light detection

Fingerprint identification is susceptible to the surface conditions of a finger. We studied the effect of image degradations with our sensor technology based on scattered-light detection. To simulate a smeared finger, we printed small black dots on a transparent sheet and inserted it between a finger and our sensor. With a green light-emitting diode ( LED) as its light source, the rate of successful identification decreased constantly as the number of black dots increased. With a near-infrared LED, the rate remained almost constant when the total area covered by the black dots was less than 1% of the input area. The infrared LED generated a clearer image for oily fingers than the green LED as well. We attribute these findings to the presence of an internal fingerprint pattern, which near-infrared light can probe better. (C) 2008 The Optical Society of Japan.