Face tracking and recognition by using omnidirectional sensor network

In recent years, security camera systems have been installed in various public facilities. More intelligent processes are needed to track people in image sequences for security camera systems. In this paper, we propose a face tracking and recognition method based on a Bayesian framework. We assume that an observed space is three-dimensional, and we estimate the 3D position of a person. We use facial 3D shape, movement, and texture models for face tracking and recognition. Omnidirectional image sensors are used to acquire image sequences of a walking person because the sensors have a wide view and are suitable for object tracking. Our system generates 3D positional hypotheses based on the facial movement model and these positional hypotheses are projected onto an image plane. Image features are extracted from projected hypotheses and the system distinguishes faces using these image features. Our evaluation experiments show that our proposed method is effective for face tracking, and that tracking accuracy is proportional to the number of cameras used.