Abstract
This paper describes a method to reconstruct in real-time the entire shape of the object by fusing the depth map obtained from three depth sensors set up around the moving object, for the generation of Free Viewpoint Video. For heavy processing of reconstruction of the entire shape of the moving object, in this paper, we realize high-speed by the shape processing and visualization unique method using GPU. Moreover, we implemented a observing system of the shape that is reconstructed, using a stereoscopic HMD.
