Generating Images with Interactive Manipulation and its Model Capture

In this paper, the methods of interactive new image generation using processed video contents, and its 3D model capture are proposed; we mainly stress the key technique of the system and the implementation of moving 3D object extraction. In order to reflect a user's intention for generating new multimedia contents, we discuss the methods of storage, retrieval, and presentation. This system is an example of the novel creation of multimedia contents by using the technologies of computer vision and virtual reality. The user is able to generate novel multimedia contents without shape modeling by a human. The proposed low-cost extraction and interaction algorithm can track objects at 8 frames/sec with sufficient accuracy.