Markerless landmark localization on body shape scans by non-rigid model fitting

We present a method of localizing anatomical landmarks on the human body by fitting a template mesh model to individual scans. The template is created from a corpus of body scans, and then registered onto individual scans using a non-rigid variant of iterative closest point (ICP) techniques. The proposed algorithm is robust against such defects as incomplete body parts and presence of background objects, and capable of generating a high-resolution mesh model without large geometric distortion. To evaluate the performance of the proposed system, 1) the template models of Japanese female and male were created from the AIST/HQL Anthropometric Database 2003, 2) the average error in the estimated landmark positions were examined using a different dataset, and 3) the accuracy and reliability of the landmark positions were evaluated according to the criteria specified in ISO 20685. The experiments confirm that the accuracy depends on landmarks, and fifty-six landmarks out of sixty-four can be localized within allowable error variations.

[1]  Marc Alexa,et al.  As-rigid-as-possible surface modeling , 2007, Symposium on Geometry Processing.

[2]  Berthold K. P. Horn,et al.  Closed-form solution of absolute orientation using unit quaternions , 1987 .

[3]  Andrew E. Johnson,et al.  Spin-Images: A Representation for 3-D Surface Matching , 1997 .

[4]  Sami Romdhani,et al.  Optimal Step Nonrigid ICP Algorithms for Surface Registration , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[5]  Michael J. Black,et al.  Evaluating the automated alignment of 3D human body scans , 2011 .

[6]  Tamal K. Dey,et al.  Eurographics Symposium on Point-based Graphics (2005) Normal Estimation for Point Clouds: a Comparison Study for a Voronoi Based Method , 2022 .

[7]  Mitsunori Tada,et al.  Dhaiba: Functional Human Models to Represent Variation of Shape, Motion and Subjective Assessment , 2006 .

[8]  M. Floater.  Mean value coordinates , 2003, Computer Aided Geometric Design.

[9]  Paul J. Besl,et al.  A Method for Registration of 3-D Shapes , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  Chang Shu,et al.  Automatic Locating of Anthropometric Landmarks on 3D Human Models , 2006, Third International Symposium on 3D Data Processing, Visualization, and Transmission (3DPVT'06).

[11]  Hans-Peter Seidel,et al.  Animation cartography—intrinsic reconstruction of shape and motion , 2012, TOGS.

[12]  Marc Levoy,et al.  Efficient variants of the ICP algorithm , 2001, Proceedings Third International Conference on 3-D Digital Imaging and Modeling.

[13]  Huamin Wang,et al.  Modeling deformable objects from a single depth camera , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[14]  Yücel Yemez,et al.  Minimum-Distortion Isometric Shape Correspondence Using EM Algorithm , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[15]  Leonidas J. Guibas,et al.  Non-Rigid Registration Under Isometric Deformations , 2008 .

[16]  Sebastian Thrun,et al.  The Correlated Correspondence Algorithm for Unsupervised Registration of Nonrigid Surfaces , 2004, NIPS.

[17]  David M Rempel,et al.  A three-dimensional anthropometric solid model of the hand based on landmark measurements , 2008, Ergonomics.

[18]  Zoran Popovic,et al.  The space of human body shapes: reconstruction and parameterization from range scans , 2003, ACM Trans. Graph..