照明変化にロバストな顔画像を用いた個人識別(機械力学,計測,自動制御)

概要

論文の詳細を見る
In this paper, we proposed the face recognition method by using the information of intensity magnitude and direction in the input image. In generally, it was said the nose and eyes are important features, when people recognize each person. In previous works, these features, such as the edges and intensity etc. of the face parts, are used. On the other hand, it is thought that the intensity magnitude and direction are important features in the face image. In this paper, we proposed a face recognition method by using the information of intensity magnitude and direction. But these features are strongly influenced by the illumination conditions. Therefore, in our method, we established a robust recognition system against the illumination changes. In this system, combining the intensity direction information, we set a larger weight in the region near the edges where the effects of the illumination conditions are slighter. By the experiments, the system was verified to be effective.
2006-05-25