A new approach towards facial recognition system using DCT & LVQ

This paper introduces a novel means of neuron coding allowing us to recognize the face of an unknown individual. The algorithm involves the application of segmentation and Discrete Cosine Transform (DCT) and is used to perform the facial recognition after training and classification from a neural network. The algorithm determines features´ position and pixel value using image segmentation and DCT which is then applied to the neural network. The neural network used in the present discussion is Linear Vector Quantization (LVQ). The main aim here has been to incorporate varied constraints of pose and expressions of inputs facilitating insensitivity to system fluctuations.