Comparison Between Eigenface Epace and Wavelet Technique as Methods of Face Recognition

In this paper we compare between two methods which are used efficient in extracting feature spaces of human faces. These two methods are wavelets technique and eigenface space. Three Multi layer perceptron (MLP) neural networks with conjugate gradient learning are used in this paper as a recognizer. The inputs to the neural network are the features (coefficients) of these two methods extracted from face images at a particular scale. An accuracy of average of 95% is observed for test images under different environment conditions not included during training.