"In this paper, the authors present a face recognition system that identifies a person from the input image given, for authentication purposes. As a feature extraction technique, Linear Discriminant Analysis (LDA) is used. After the generation of features, the classification is performed using Euclidean Distance classifier. Recognition rates are calculated for varying sizes of training data and corresponding test data. The data set is the ORL face database which is a standard face database for face recognition systems. The database consists of 400 images of 40 people with 10 different poses for each individual."