Entertaining video warping

While various techniques of image deformation have been developed and extensively applied in animation and morphing, there are few works to extend these techniques to handle videos, especially real-time warping of a meaningful moving part in the video like human face. An efficient online algorithm is proposed in this paper to implement real-time face warping for video sequence. We employ AdaBoost to detect the sixteen human facial feature points and implement a fast face warping frame by frame while maintaining both temporal and spatial continuity of the warped video. In order to reduce the shaking of the detected facial feature points due to the noise in the video, we develop a novel model called Frame Buffer. All these procedures are designed efficient to guarantee the real-time performance of our system (15 fps). In addition, many other types of warping functions are also compatible with our framework. It is shown that our algorithm can be applied in real-time special effect editing in video as well as other entertainment applications.