HCFD-based Facial Recognition System for User Identification

Facial recognition systems have their accuracy based on the intra-class variations between two stages in particular, enrollment and identification. These intra-class variations are affected by the lighting conditions with other reasons such as, facial expressions, pose, occlusion, poor sensor quality, and illumination quality. To identify a face or an object with accuracy, all such errors need to overcome. In this paper we are proposing a new approach about the effects of varying light conditions and hence, in order to curb these effects, face images are pre-processed to normalize intra-class variations. Traditional principal component analysis with Haar Cascade Face Detector is used for the facial recognition process for user identification. Keywords—eigenfaces, neural networks, OpenCV library, facial recognition, Haar cascade face detector (HCFD)