Skip to main navigation Skip to search Skip to main content

3D human body shape and motion tracking by LBS and snake

  • Jia Chen*
  • , Xiaojun Wu
  • *Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

A new framework combining Snake deformable model with LBS (Linear Blend Skinning) is proposed for both 3D human body shape and motion tracking. First, skinned mesh model of object is established. Then surface mesh with highly improved quality is obtained by using our new silhouette-based visual hull reconstruction method for each frame of multi-view videos. After that, articulated ICP registration method is used to capture the 3D human body shape and motion under Skeletal Subspace Deformation. Furthermore, the multi-view silhouettes are applied again for Snake external force computation, and then the vertices of surface mesh are moved to approach the target object under silhouette and internal forces. Synthetic data with ground-truth is used for quantitative comparison, and experimental results show that the tracking accuracy is improved since both 3D and 2D error constraints are used.

Original languageEnglish
Pages (from-to)357-363+371
JournalJisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics
Volume24
Issue number3
StatePublished - Mar 2012
Externally publishedYes

Keywords

  • 3D motion tracking
  • Linear blend skinning
  • Snake
  • Visual hull

Fingerprint

Dive into the research topics of '3D human body shape and motion tracking by LBS and snake'. Together they form a unique fingerprint.

Cite this