Personal identification system through facial images based on the modified HSV color system and isodensity maps
Y. Hirayama, Osamu Nakamura · 2002
An algorithm which evaluates the extraction accuracy of a facial area and a personal identification system which is able to apply the facial image changes are described. Computer simulation using 91 pictures for extracting facial area shows a correct extraction accuracy of 90.1%. The experimental results using 79 pairs of mobile faces show a correct identification accuracy of 84.8%.