Classification of human age based on Neural Network using FG-NET Aging database and Wavelets

Face Aging has been an vital area of research for the past few decades. As the age increases, there are some visible changes in the face, making age classification simpler. Based on the facial growth, we can classify the human age into various kinds. Though there are various algorithms existed so far, a more sophisticated method is attempted for classifying facial age. Age Prototypes, Statistical models and Distance based technique have been widely used for classification of human face. The system can be improved by using the Wavelet Transformation (WT) for extracting the face features and Artificial Neural Network to classify the age group. The facial images are pre-processed and then the face features are extracted using Wavelet Transformation. The distance between each of features are evaluated using Euclidean distance and these values were given as input to Adaptive Resonance Network (ART). The Neural Network is trained using FG-NET (Face and Gesture Recognition Research Network) aging database. The human age is classified into four categories as Child (0-12 years), Adolescence (13-18 years), Adult (19-59 years) and Senior Adult (60 years and above) which is discussed in the paper.