Head Pose Estimation Based Robust Multicamera Face Recognition

In recent times, multicamera based face recognition has become an active research domain. Face recognition is utilized in multiple areas of applications such as surveillance, security etc. Several limitations and challenges like occlusion, pose changes, different illumination conditions, resolution, motion blur and sometimes subject will not be cooperative with the current face recognition systems. Due to the changes in pose, it becomes very difficult to recognize the person exactly by using a single camera. To overcome some of these limitations and challenges, a new approach of face recognition using multiple camera-based head pose estimation is proposed. The main motivation of this approach is to efficiently recognize the person by estimating the head pose. This estimation is used to select the best camera based on the direction and orientation of human head by using a network of multiple cameras.