Face Detection and Tracking System Using Hybrid Technique—A Design Implementation on DRiT

摘要

Face detection and tracking using image processing technique is a progressive research in the last few years. This biometric technology revolutionizes many industries like robotics, surveillance and security. This is due to the need of having a robust system operating in a real-timeenvironment. Despite the advancement of this technology, it remains as a challenging field of study as it requires solving many aspects like pose, illumination, scale and background. To circumvent this problem, a real-time face recognition and tracking system (DRiT) was developed for telepresencerobot. The DRiT system does not only detects a human face but ultimately recognize and tracks the face. The system consists of several modules designed with hybrid algorithms. A comprehensive algorithm for face detection module is discussed in this paper. A new approach using circle detectionmethod was introduced to increase the detection rate. The new method developed for face detection module was tested on DRiT system with a set of experiments and it demonstrates a good accuracy.