Abstract: Since few years Face Recognition has become quite popular because of its various applications in several domains. In automated face recognition a very difficult challenge is to compute the facial similarities between face images obtained in alternate modalities. To address this challenge Heterogeneous Face recognition (HFR) implementation is considered. This procedure helps to match the two face images which can vary from infrared image to a photograph like, driver’s license, passports, etc. or to a face image from different modalities like infrared images, aged images, etc. In this paper we will discuss how HFR can be implemented using SIFT and MLBP algorithm to extract various image features. The main objective is to extract the different invariance features from heterogeneous images that help to give a correct match between probe and gallery images.

Keywords: Face recognition, heterogeneous face recognition, gaussian, DOG, CSDN, SIFT, MLBP, LDA, KLDA.