Wavelets and Neural Networks Based Face Recognition System
Mohamed R. M. Rizk, O. Said, Rania Salah El-Sayed · 2006
This paper introduces a neural network-based face recognition system. The discrete wavelet transform was used for feature extraction and dimensional reduction method as a preprocessing step to the ORL database before training the network. A new algorithm is introduced for performance improvement. The algorithm trains two neural networks, each with a different input vector, the first is the horizontal. Image elements and the second is the vertical image elements. The recognition rate is highly improved.