Face Recognition Based on Steerable PyramidTransform and LS-SVM

摘要

An efficient local appearance feature extraction method based on S-P wavelet transform and Least Square Support Vector Machine (LS-SVM) is proposed for face recognition to reduce the dimensionality of facial image and improve the recognition rate.As a latest multi-resolution analysis method,steerable pyramid transform (S-P) has improved directional elements with anisotropy and better ability to represent sparsely edges and other singularities.By utilizing S-P wavelet transform to extract features from facial images and LS-SVM to classify facial images based on features,the proposed scheme has been evaluated by carrying out experiments on the wellknown ORL face database.Experimental results show that the proposed method provides a better representation of the class information,and obtains much higher recognition accuracies in real-world situations including changes in pose,expression and illumination.