Comparative study of camshift and KLT algorithms for real time face detection and tracking applications

Abstract

Face detection and tracking is one of the emerging research areas in the image analysis and computer vision systems. This face detection and tracking helps local security forces to investigate crime incidents. This paper describes a face tracking framework that is capable of tracking a face in real time rapidly frame by frame. Camshift algorithm and KLT algorithm implemented and a comparison study between these two algorithms has been described in this paper. A real time video is experimented using these two algorithms for face detection and tracking. The experimental results show that the KLT algorithm performance is better than the Camshift algorithm in detecting the face and tracking the face. In this article it has been shown how the KLT algorithm proved to be better tracking algorithm than Camshift algorithm.