Face Feature Age Prediction through Optimized Wavelet Back Propagation Network

With the advancement in technology, one thing that concerns the world and especially in the developing countries is the tremendous increase in population. With such a rapid rate of increase, it is becoming difficult to recognize each and every person because we have to keep up photos either in digital or hard copy format of every person at different time periods of his/her life. Sometimes database has the required information of that particular person, but it’s of no use as it is now obsolete. Deciding age of a person from digital photography is an intriguing problem. Age changes cause many variations in visible of human faces. Many aspects affect the appearance of a person’s face during the process of growing older. The aging process will explain with many factors such as health, living style, living place and weather condition etc. Face is a non-intrusive recognition, without user co-ordination able to recognize the person. Age classification system is generally composed of feature extraction and classification. It is used to estimate the age of a person from his/her face features. For the aging feature extraction, face images interpreted as decomposition of optimized wavelet transform with 49 feature vectors using Daubechies wavelet and the classifier of supervised neural network to discriminate the ranges of ages. The work is to classify the age range into child (1-10), teenage (11-20) young (21-30), middle aged (31-50) and old (51 and above).