Performance Comparison for Face Recognition using PCA, DCT & Walsh Transform of Row Mean and Column Mean

Face recognition has been a fast growing, challenging and interesting area in real-time applications of image processing. A large number of face recognition algorithms have been developed for decades. Principal Component Analysis (PCA) [2][3] is one of the most successful techniques that have been used in face recognition. Here four criteria for gray image pixel selection to create feature vector were analyzed, the first one has all the pixels, the second one is based on taking row mean of the face image, the third one is based on taking column mean of the face image and the fourth criterion is based on taking row and column mean of the face image. These face image pixel distributions are used to generate feature vectors with the help of Principal Component Analysis (PCA), Discrete Cosine Transform (DCT) and Walsh Transform. Experimental tests on the ORL Face Database [1] achieved 99.60% of recognition accuracy, with lower computational cost. To test the ruggedness of proposed techniques, they are tested on our own created face database where 84.60% of recognition accuracy is achieved. From experimental results it can be observed that the proposed Row and Column Mean Technique gives similar performance compared to using all image coefficients at much reduced complexity, so proposed technique is better for real time applications.