Age Identification System Based on Bivariate Gaussian Mixture Model
- Venkatarao Rampay
A method of automatic face recognition for personal authentication is proposed considering the age progression between the registered face image and the captured one. In face recognition, the features of the captured face image are compared with those of the registered faces in database and the closest registered face is selected as the person's. However, if age progression exists between the captured face and the registered one, the features changes with the person's growth. In this paper, a method of face recognition over age progression is proposed using the block matching method, especially for the case that the registered face is taken in the childhood and the person is to be authenticated in his youth. Since block matching compensates the position shift of face components, this method can avoid the harmful effect of the shape change of faces caused by growth. The experiment results show advantage of the proposed method to the conventional holistic one.