• DocumentCode
    3392996
  • Title

    Geodesic Model of Human Body

  • Author

    Wu, Weihe ; Hao, Aimin ; Zhao, Yongtao

  • Author_Institution
    State Key Lab. of Virtual Reality Technol. & Syst., Beihang Univ., Beijing, China
  • fYear
    2010
  • fDate
    20-22 Oct. 2010
  • Firstpage
    391
  • Lastpage
    397
  • Abstract
    Anthropometry is widely applied to the research in skeleton extraction from surface meshes of human body. Especially the anatomical proportion can be employed as a benchmark in model segmentation and joint extraction. Unfortunately, the anatomical proportion is usually measured with the Euclidean distance, which makes it difficult to correlate it with the surface mesh. To bridge this gap, we take advantage of the property of the geodesic metrics that is invariance to rotation, translation, scaling and model pose, and propose an original geodesic model in which the length of each part of human body is measured by geodesic metrics, by which the anatomic proportions can be directly mapped to the contours of the mesh surface of human body in arbitrary pose. Combining the geodesic model with automatic extraction of feature points, we can determine the candidate scopes of joint positions and boundaries between the parts on meshes, and then refine the joint positions in the scopes using existing methods. And finally, we illustrate the utility of the geodesic model with an application to joint extraction.
  • Keywords
    anthropometry; bone; differential geometry; feature extraction; pose estimation; Euclidean distance; anthropometry; arbitrary pose; feature point extraction; geodesic metrics; human body; joint extraction; model segmentation; skeleton extraction; surface meshes; Biological system modeling; Computational modeling; Feature extraction; Joints; Level measurement; Shape; anthropometry; geodesic distance; human body model; skeleton extraction;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Cyberworlds (CW), 2010 International Conference on
  • Conference_Location
    Singapore
  • Print_ISBN
    978-1-4244-8301-3
  • Electronic_ISBN
    978-0-7695-4215-7
  • Type

    conf

  • DOI
    10.1109/CW.2010.16
  • Filename
    5655197