期刊名称：International Journal of Computer Science Issues
印刷版ISSN：1694-0784
电子版ISSN：1694-0814
出版年度：2013
卷号：10
期号：1
出版社：IJCSI Press
摘要：Abstract An accurate and robust transformed face descriptor that exploits the capabilities of filtered backprojection applied on Discrete Cosine Transform (DCT) and kernel Principal Component Analysis (PCA) methods is proposed. The method is invariant to rotation, variations in facial expression and illumination. Filtered backprojection constructs transform parameters from a set of projections through an image enhancing feature patterns that provide an initialization for subsequent DCT computations. DCT discards high-frequency coefficients that form least significant data to retain a subset of lower frequency coefficients visually significant in the image. The resulting coefficient features are mapped to lower dimensional space using PCA which extracts principal components that form the basis for the neural network classifier. Experiments were carried on JAFEE database and computed results compared with PCA and DCT approach. The results demonstrate significant improvements in results compared to other approaches.
