A Study of Geometric Features for Aging Face Recognition

Human facial shapes undergo significant variations from infancy to teenage, while they change limitedly as people age into adulthood. In this paper, facial shapes of males across ages using Euclidean distances are studied and a D-A standard (Distance related to Ages) is proposed for this purpose. Further, we determine the distance features that significantly contribute to the aging face recognition. Experimental results using the MORPH database containing age separated facial images are encouraging and illustrate Euclidean distance features improve the performance of face recognition across ages.