A Reconstruction Model for 3D Human Motion Based on Images Processing

Abstract:

Traditional 3D reconstruction algorithms use fixes shape base which hardly expresses the change parameters of complex movement and motion law of large-scale dynamic features, thereby leading to non-realistic reconstruction results. The paper proposes a new reconstruction algorithm for 3D motion images that corrects the neighborhood system of feature points by motion parameters and reasons number base K to ensure accuracy. The simulation results show that, the proposed algorithm avoids drawbacks of sports reconstruction results caused by the great randomness of motion state, thereby complete 3D motion images' reconstruction.

• Periodical: Applied Mechanics and Materials (Volumes 556-562)
• Pages: 5021-5023
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.556-562.5021
• Online since: May 2014
• Authors: Zhi Yuan Yang*
• Keywords: 3D Motion Reconstruction, Body Movements, Feature Points

References

[1] O. R. Musin. Properties of the delaunay triangulation[J]. In Proc. 13th Annual ACM Symposium on Computational Geometry, pages 424–426, (1997).

[2] Ting Chen, Wei-Chung Lin. Artificial Neural Networks for 3-D Motion Analysis-Part II: Nonrigid Motion[J]. IEEE Trans Neural Networks, 1995, 6(6): 1394-1401.

DOI: 10.1109/72.471368

[3] Chen T, Lin W C, Chen C T. Artificial neural networks for 3-D motion analysis I: Rigid motion[J]. IEEE Trans Neural Networks, 1995, 6(6): 1386-1393.

DOI: 10.1109/72.471369

[4] Tzovaras D, Ploskas N, Strintzis G. Rigid 3-D motion estimation using neural networks and initially estimated 2-D motion data[J]. IEEE Trans Circuits and Systems for Video Technology, 2000, 101, 10(1): 158-165.

DOI: 10.1109/76.825869

[5] H. Derin and H. Elliott. Modeling and segmentation of noisy and textured images using Gibbs random fields[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, PAMI-1987. 9(1): 39-55.

DOI: 10.1109/tpami.1987.4767871