A Face-based Age Estimation System Using Back Propagation Neural Network Technique
M. O. Oladele *
Department of Computer Science and Engineering, Faculty of Engineering & Technology, Ladoke Akintola University of Technology (LAUTECH), Nigeria.
E. O. Omidiora
Department of Computer Science and Engineering, Faculty of Engineering & Technology, Ladoke Akintola University of Technology (LAUTECH), Nigeria.
A. O. Afolabi
Department of Computer Science and Engineering, Faculty of Engineering & Technology, Ladoke Akintola University of Technology (LAUTECH), Nigeria.
*Author to whom correspondence should be addressed.
Abstract
Age estimation is the determination of a person’s age based on biometric features such as face, finger print etc. it is a hard problem for both humans and the computer system. In this paper, back propagation neural network was used to classify face images into eight different age groups ranging from babies, young teenagers, mid teenagers, teenagers, young adults, mid adults, young old and old. The process is divided into three stages: Image preprocessing, Feature Extraction and age classification. The images were cropped, resized and converted into grayscale in the preprocessing stage. Principal Component Analysis was used to extract the facial features that will be fed into the neural network classifier. Finally, back propagation neural network was used to classify the face images into any of the eight groups. The developed system was experimented with 630 face images with different ages from the FG-NET database. 450 samples were used for training while 180 were used for testing. The results showed a training time of 110.914 seconds, Mean Absolute Error (MAE) of 3.88 years and an overall accuracy of 82.2%.
Keywords: Age estimation, biometric, preprocessing, feature extraction, back propagation neural network.