Face Recognition Using Handcrafted Features Extraction Techniques Under Occlusion Condition

Over the past three decades, many face recognition methods have been proposed in computer vision, and most of the methods recognized faces under constrained scenarios where no occlusion, illumination and pose variation exist. In a real-world application such as video surveillance, human faces might be occluded by other objects or some face accessories and it is difficult to identify a person’s face. To address this, a new face recognition approach develop to recognize a person from their unoccluded part of the face. The proposed method extracts more relevant features than the existing handcrafted features methods, which helps to improve recognition accuracy. The validation of the proposed method has been verified using existing major datasets, i.e. AR and LFW public datasets.