Based Face Recognition System Of The Gabor Wavelet

Face recognition has been an active research area for more than two decades. Various exciting results have been achieved in fields of image processing, pattern recognition, machine learning and computer vision. The paper first discusses a face recognition algorithm based on PCA and LDA.then an implementation of a Face Recognition System based on Gabor wavelets is presented.In the Face Recognition System, a face is represented by a set of feature points. Each of these feature points is described by a Jet, which is a set of filter responses of 40 complex Gabor filters that are applied at that point. Thus a jet describes a local neighborhood of gray-values around a feature point. The filter set is fixed and usually (but this may vary) contains Gabor filters that have eight different orientations and five different central frequencies. For training images, relevant feature points can be selected by hand and jets are extracted from the positions. For novel images, the corresponding feature points have to be found automatically. The coarse positions of the feature are located by the training face structure knowledge, then we use Elastic Bunch Graph to adjust the accurate positions. After the accurate positions of the features are located, Gabor jets are extracted from novel images. Similarity between novel images is expressed as function of similarity between localized Gabor jets corresponding to facial landmarks.The experiments show that the face recognition system tested on the BioID face Database is efficiency. The recognition rate is 90.5% and the relative speed of the similarity measures is 2800 faces per second.