A novel method for human age group classification based on Correlation Fractal Dimension of facial edges

In the computer vision community, easy categorization of a person's facial image into various age groups is often quite precise and is not pursued effectively. To address this problem, which is an important area of research, the present paper proposes an innovative method of age group classification system based on the Correlation Fractal Dimension of complex facial image. Wrinkles appear on the face with aging thereby changing the facial edges of the image. The proposed method is rotation and poses invariant. The present paper concentrates on developing an innovative technique that classifies facial images into four categories i.e. child image (0-15), young adult image (15-30), middle-aged adult image (31-50), and senior adult image (> 50) based on correlation FD value of a facial edge image. (C) 2015 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University. This is an open access article under the CC BY-NC-ND license.