期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2015
卷号：3
期号：9
DOI：10.15680/IJIRCCE.2015. 0309093
出版社：S&S Publications
摘要：Face Recognition is a very vast and challenging field. The limited research work has been done in thefield of face recognition by considering the aging attribute. The age effect brings certain changes in the face with theprogression in age. The mission of proposed methodology is to achieve the age invariant face recognition by usingfollowing steps: a) extracting facial features by using Viola Jones algorithm; b) the α-shape is constructed by using theextracted facial features; and c) extracting topological features for face recognition task. The result is evaluated byusing two public domain available FGNET and MORPH dataset. From both the dataset certain images are trained tosystem and test for the face recognition task. The result obtained after the evaluation is 50% for FGNET dataset when50 subjects are trained to the system and 90% for MORPH dataset when 60 subjects are trained to the system.Experimental results obtained demonstrate the efficiency of the proposed methodology named nonlinear topologicalcomponent analysis.
