Wavelet Face Recognition Using Bayesian Classifier
Li Ping Niu, Shi Tao Yan · Advanced materials research · 2012
A novel wavelet face recognition using Bayesian classifier technique is presented. Wavelet-based and 2DLDA method have to down sample the original image dataset. In the preprocessing stage, the system uses 2DLDA to select the N top-ranked candidate images but forges making the final choice among those candidates, secondly all the candidate images are decomposed by applying wavelet transform, Bayesian recognition is parallel processed using these sub-band images. Final decision is made by use of weighed average. Our experiments on FERET have shown the effectiveness of the method.