Survey of Face Recognition using ICA

Abstract

Nowadays, face recognition is used in number of applications as a means of authentication and verification. But the task of face recognition is not easy for a computer system. There are various ways that can be used for face recognition. This paper reviews the problem of face recognition using edge information as independent component. Here, the edge information is obtained by using LoG(Laplacian of Gaussian) and Canny edge detection methods. On the obtained information, Principle Component Analysis (PCA) is applied for preprocessing. Then, the images are trained using Independent Component Analysis (ICA). Then the images are tested.