Face Recognition and Tracking in Videos
S. V. Tathe, S. P. Narote, Sandipan Pralhad Narote · Advances in Science Technology and Engineering Systems Journal · 2017
Advancement in computer vision technology and availability of video capturing devices such as surveillance cameras has evoked new video processing applications.The research in video face recognition is mostly biased towards law enforcement applications.Applications involves human recognition based on face and iris, human computer interaction, behavior analysis, video surveillance etc.This paper presents face tracking framework that is capable of face detection using Haar fea-tures, recognition using Gabor feature extraction, matching using correlation score and tracking using Kalman filter.The method has good recognition rate for real-life videos and robust performance to changes due to illumination, environmental factors, scale, pose and orientations.