Pht and kelm based face recognition

Abstract

Recognition of human face images is getting much attraction in pattern recognition since last few decades. Artificial intelligence and machine learning always tries to get more and more accurate for recognizing the face images. Only pixel based information of the face image can be helpful in recognizing the face images. This recognition rate can be increased if some feature of the face image is also added up with the pixel information of the face image. Based on this phenomenon, polar harmonic transform is utilized as the feature extraction technique for the feature based information. With this feature based information, kernel extreme learning machine (KELM) is utilized as the classification tool. It can be seen from the results obtained on the ORL, YALE and Georgia Tech face databases that more accurate results can be obtained using the feature based information.