3D human model and joint parameter estimation from monocular image

Minglei Tong, Yuncai Liu, Thomas S. Huang

Research output: Contribution to journalArticle

Abstract

In this paper we present a novel class of human model described by convolution surface attached to articulated kinematics skeletons. The human pose can be estimated from silhouette in monocular images. The contribution of this paper consists of three points: First, human model of convolution surface is presented and its shape is deformable when changing polynomial parameters and radius parameters. Second, convolution surface and curve correspondence theorem is presented to give a map between 3D pose and 2D contour. Third, we model the human silhouette with convolution curve in order to estimate joint parameters from monocular images and we also give an effective constraint function. Evaluation of this approach is performed on some video frames about a walking man. The experiment result shows that our method works well without self-occlusion.

Original languageEnglish (US)
Pages (from-to)797-805
Number of pages9
JournalPattern Recognition Letters
Volume28
Issue number7
DOIs
StatePublished - May 1 2007

Fingerprint

Convolution
Parameter estimation
Kinematics
Polynomials
Experiments

Keywords

  • 2D images
  • 3D human model
  • Convolution surface
  • Model initialization
  • Motion estimation

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

3D human model and joint parameter estimation from monocular image. / Tong, Minglei; Liu, Yuncai; Huang, Thomas S.

In: Pattern Recognition Letters, Vol. 28, No. 7, 01.05.2007, p. 797-805.

Research output: Contribution to journalArticle

@article{3e9f7cc22e75465c83144e0d3b1da5b9,
title = "3D human model and joint parameter estimation from monocular image",
abstract = "In this paper we present a novel class of human model described by convolution surface attached to articulated kinematics skeletons. The human pose can be estimated from silhouette in monocular images. The contribution of this paper consists of three points: First, human model of convolution surface is presented and its shape is deformable when changing polynomial parameters and radius parameters. Second, convolution surface and curve correspondence theorem is presented to give a map between 3D pose and 2D contour. Third, we model the human silhouette with convolution curve in order to estimate joint parameters from monocular images and we also give an effective constraint function. Evaluation of this approach is performed on some video frames about a walking man. The experiment result shows that our method works well without self-occlusion.",
keywords = "2D images, 3D human model, Convolution surface, Model initialization, Motion estimation",
author = "Minglei Tong and Yuncai Liu and Huang, {Thomas S.}",
year = "2007",
month = "5",
day = "1",
doi = "10.1016/j.patrec.2006.11.011",
language = "English (US)",
volume = "28",
pages = "797--805",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",
number = "7",

}

TY - JOUR

T1 - 3D human model and joint parameter estimation from monocular image

AU - Tong, Minglei

AU - Liu, Yuncai

AU - Huang, Thomas S.

PY - 2007/5/1

Y1 - 2007/5/1

N2 - In this paper we present a novel class of human model described by convolution surface attached to articulated kinematics skeletons. The human pose can be estimated from silhouette in monocular images. The contribution of this paper consists of three points: First, human model of convolution surface is presented and its shape is deformable when changing polynomial parameters and radius parameters. Second, convolution surface and curve correspondence theorem is presented to give a map between 3D pose and 2D contour. Third, we model the human silhouette with convolution curve in order to estimate joint parameters from monocular images and we also give an effective constraint function. Evaluation of this approach is performed on some video frames about a walking man. The experiment result shows that our method works well without self-occlusion.

AB - In this paper we present a novel class of human model described by convolution surface attached to articulated kinematics skeletons. The human pose can be estimated from silhouette in monocular images. The contribution of this paper consists of three points: First, human model of convolution surface is presented and its shape is deformable when changing polynomial parameters and radius parameters. Second, convolution surface and curve correspondence theorem is presented to give a map between 3D pose and 2D contour. Third, we model the human silhouette with convolution curve in order to estimate joint parameters from monocular images and we also give an effective constraint function. Evaluation of this approach is performed on some video frames about a walking man. The experiment result shows that our method works well without self-occlusion.

KW - 2D images

KW - 3D human model

KW - Convolution surface

KW - Model initialization

KW - Motion estimation

UR - http://www.scopus.com/inward/record.url?scp=33847067869&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33847067869&partnerID=8YFLogxK

U2 - 10.1016/j.patrec.2006.11.011

DO - 10.1016/j.patrec.2006.11.011

M3 - Article

AN - SCOPUS:33847067869

VL - 28

SP - 797

EP - 805

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

IS - 7

ER -