Face Recognition Based on Fusion Feature of LBP and PCA with KNN

Abstract

Face recognition mainly include face feature extraction and recognition (classification) two parts. The basic idea of local binary pattern (LBP) use texture description operator to build a number of local face description, then describe the local combination form global description. LBP operator works with the 3*3 neighborhood pixels for texture description. After scanning the whole image, we get a LBP response images. Then, we use principal component analysis (PCA) method to dimension reduction for response image, finally using K-Nearest Neighbor classifier to classify the test sample. The experimental results show that LBP-PCA-KNN method gains higher recognition rate than PCA-KNN method’s recognition rate, can be effectively used for face recognition.