Human face recognition using neural network

This paper aims to present a new method in combining feature extraction with neural network to recognize human faces. The system implements extracting features of human face from an image in three methods: Principal Components Analysis (PCA), Linear Discriminant Analysis (LDA), and Morphological processing. Then it mentions the neural network with the algorithm of back propagation to drill the categorization which is utilized to recognize the foregoing extracted features. When the face recognition is required, the features will be extracted from the input image. The features will be recognized by the trained neural networks soon after. The results of each distinct network will be compared later on to bring out the final systematic answer. Finally, by using statistical features and morphological features, the performance of the system is quite better than current systems. Especially, the system can recognize with noisy images.