立松直倫・○大谷　淳（早大）・Davis Larry（University of Maryland）

This paper improves the authors' conventional method for reconstructing the 3D structure of moving and still objects that are tracked in the video and/or depth image sequences acquired by moving cameras and/or range finder. To solve our problems concerning the 3D reconstruction: lack of accuracy of segmenting into each object's region from sparse classified points, this paper proposes a new 3D segmentation method that utilizes a multi-label graph-cut method utilizing object boundary map. Experiments using multiple moving objects demonstrate the effectiveness of our method.