Modelling geiometric features for face based age classification

As one of the important research areas in face processing, face based age classification could be very useful for the index of large face database. Though a few works have been proposed in literature, many up to date face processing technologies have not been applied and more attention shall be drawn to this area. In this paper, we developed both efficient and robust facial feature location algorithms and proposed a geometric feature based age classification system. The proposed nose and mouth location algorithm was tested using images with variant illumination, expression and pose changes. Based on the robust eyes, nose and mouth location algorithm, simple modeling of the geometric features provides a reasonable classifier for baby and adult classification. The proposed system was tested using hundreds of images downloaded from internet. Though the classification task is much more challenging, significantly better performance than earlier work in literature has been achieved.