Age Classification using Facial Image Features in DCT Domain

Abstract

Age estimation or age based classification can be achieved using different facial features both in spatial and transform domain. In this paper, transform domain features are used to classify facial images into appropriate age group. This paper provide technique which is helpful in classifying images efficiently using only transform domain features. This algorithm can be used in real time environment for different purposes in human computer interaction. Initially DCT cheeks’ energy are used as a threshold parameter to classify images into two broader age based categories. These two main categories are further subdivided into 6 age based sub categories based on Left Cheek wrinkle energy (E-I), cheeks wrinkle energy (E-II) and forehead wrinkle energy (E-III) energy.