One of the goals of the three-dimensional computer graphics is to create virtual realistic images of dynamically changing scenes. In this paper, we survey several passive acquisition approaches and design and implement a video-image based virtual volumetric scene reconstruction system, which acquire synchronized multi video images and render of dynamic real-world scenes. The system includes efficient image-based reconstruction algorithm design to computing and coloring visual hulls from silhouette image data and other required implementations e.g. robust background subtraction algorithm.
