Classification of Age Groups Based on Facial Features

An age group classification system for gray-scale facial images is proposed in this paper. Four age groups, including babies, young adults, middle-aged adults, and old adults, are used in the classification system. The process of the system is divided into three phases: location, feature extraction, and age classification. Based on the symmetry of human faces and the variation of gray levels, the positions of eyes, noses, and mouths could be located by applying the Sobel edge operator and region labeling. Two geometric features and three wrinkle features from a facial image are then obtained. Finally, two back-propagation neural networks are constructed for classification. The first one employs the geometric features to distinguish whether the facial image is a baby. If it is not, then the second network uses the wrinkle features to classify the image into one of three adult groups. The proposed system is experimented with 230 facial images on a Pentium II 350 processor with 128 MB RAM. One half of the images are used for training and the other half for test. It takes 0.235 second to classify an image on an average. The identification rate achieves 90.52 % for the training images and 81.58 % for the test images, which is roughly close to human’s subjective justification.