Advanced model-based image coding scheme

Studies on 3-D model-based image coding have concentrated on facial image sequences to implement low-bit rate communication and other applications. In this paper, this scheme is extended by including human-computer interfaces through the recognition of facial gestures and interactive control of a face object on an Internet browser, and by digital watermarking. Human faces or faces of anthropomorphic agents are synthesized with the real and/or recognized facial motion on an Internet browser using virtual reality modeling language(VRML). Digital watermarking is employed to prevent real face images from unwanted use and copyright infringement. By extending previous model-based image coding schemes for facial image sequences, an interactive, effective, safe and low-bit rate communication can be carried out. Experimental results show the usefulness and effectiveness of the proposed scheme.