The Individual Identification using The Facial Thermal Image

Abstract

The purpose of this paper is to construct the individual identification system by the facial skin thermogram (FST). FST gives a constant image under any illumination condition. Then, we extract the individual feature from FST with the image processing and examine the evaluation method to identify individual. We adopted neural networks with back-propagation (BP) learning algorithm. The identification experiment was done among 60 (20×3) registered patterns and 250 (50×5) input patterns, consequently the correct acceptance rate was 73% and the correct rejection rate was 73.3%.