Finding Articulated Body in Time-Series Volume Data
This paper presents a new scheme for acquiring 3D kinematic structure and motion from time-series volume data, in particular, focusing on human body. Our basic strategy is to first represent the shape structure of the target in each frame by using aMRG, augmented Multiresolution Reeb Graph , and then deform each of the shape structures so that all of them can be identified as a common kinematic structure throughout the input frames. Although the shape structures can be very different from frame to frame, we propose to derive a unique kinematic structure by way of clustering some nodes of graph, based on the fact that they are partly coherent. The only assumption we make is that human body can be approximated by an articulated body with certain number of end-points and branches. We demonstrate the efficacy of the proposed scheme through some experiments.
KeywordsInitial Model Geodesic Distance Shape Structure Joint Point Kinematic Structure
Unable to display preview. Download preview PDF.
- 3.Nobuhara, S., Matsuyama, T.: Heterogenieous Deformation Model for 3D Shape and Motion Recovery from Multi-Viewpoint Images. In: 2nd International Symposium on 3D Data Processing, Visualization, and Transmission, pp. 566–573 (2004)Google Scholar
- 4.Masaaki, I., Yoshinari, K., Michihiko, M.: Estimation of the Location of Joint Points of Human Body from Successive Volume Data. In: ICPR 2000, pp. 699–702 (2002)Google Scholar
- 5.Hilaga, M., Shinagawa, Y., Kohmura, T., Kunii, T.L.: Topology Matching for Fully Automatic Similarity Estimation of 3D Shapes. In: Proc. of SIGGRAPH 2001, pp. 203–212 (2001)Google Scholar
- 7.Kenmochi, Y., Kotani, K., Imiya, A.: Marching Cubes Method with Connectivity. In: Proc. of International Conference on Image Processing, pp. 361–365 (1999)Google Scholar