Real-time human motion estimation using biomechanical models and non-linear state-space filters

Pietro Cerveri, M. Rabuffetti, A. Pedotti, G. Ferrigno

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

In the field of sports biomechanics and rehabilitation engineering, the possibility of computing, in real time, the angular displacements and derivatives of human joints, from a video of motion sequences, represents an appealing goal. In particular, applications of biofeedback protocols in rehabilitation can benefit from this capability. The focus of the investigation was concerned with the application of biomechanical models, comprising of a kinematic chain and surface envelopes, and state-space filters, to the computation, in real time and with high accuracy, of the angular data and derivatives. By minimising the distances, measured with TV cameras, between the 2D marker projections and the corresponding back-projected markers located on the mannequin, the configuration of the biomechanical model was automatically updated. The use of state-space estimation allowed the computation of smooth derivatives of the orientation data. Owing to the non-linearity of the functions involved, the derivatives of the observation model were obtained through a multidimensional extension of Stirling's interpolation formula. Proper algorithms were developed to cope with the model calibration, initialisation and data labelling. Extensive experiments on real and simulated motions proved the reliability (maximum angular error less than 1°, maximum point reconstruction less than 1 mm) of the developed system, which is robust to false matching caused by marker occlusions. Moreover, orientation artifacts due to skin motion can be reduced by a factor of 50%.

Original languageEnglish
Pages (from-to)109-123
Number of pages15
JournalMedical and Biological Engineering and Computing
Volume41
Issue number2
DOIs
Publication statusPublished - Mar 2003

Fingerprint

Nonlinear Dynamics
Motion estimation
Derivatives
Biomechanical Phenomena
Rehabilitation
Patient rehabilitation
Manikins
Sports medicine
Biofeedback
Artifacts
Calibration
Sports
Joints
Observation
Labeling
Interpolation
Skin
Kinematics
Cameras
Experiments

Keywords

  • Biomechanical models
  • Extended Kalman filters
  • Human motion estimation
  • Motion tracking
  • Multi-camera systems
  • Rehabilitation

ASJC Scopus subject areas

  • Biomedical Engineering
  • Health Informatics
  • Health Information Management
  • Computer Science Applications
  • Computational Theory and Mathematics

Cite this

Real-time human motion estimation using biomechanical models and non-linear state-space filters. / Cerveri, Pietro; Rabuffetti, M.; Pedotti, A.; Ferrigno, G.

In: Medical and Biological Engineering and Computing, Vol. 41, No. 2, 03.2003, p. 109-123.

Research output: Contribution to journalArticle

@article{87a16f9e3a7e43d3b54dc2c12dace800,
title = "Real-time human motion estimation using biomechanical models and non-linear state-space filters",
abstract = "In the field of sports biomechanics and rehabilitation engineering, the possibility of computing, in real time, the angular displacements and derivatives of human joints, from a video of motion sequences, represents an appealing goal. In particular, applications of biofeedback protocols in rehabilitation can benefit from this capability. The focus of the investigation was concerned with the application of biomechanical models, comprising of a kinematic chain and surface envelopes, and state-space filters, to the computation, in real time and with high accuracy, of the angular data and derivatives. By minimising the distances, measured with TV cameras, between the 2D marker projections and the corresponding back-projected markers located on the mannequin, the configuration of the biomechanical model was automatically updated. The use of state-space estimation allowed the computation of smooth derivatives of the orientation data. Owing to the non-linearity of the functions involved, the derivatives of the observation model were obtained through a multidimensional extension of Stirling's interpolation formula. Proper algorithms were developed to cope with the model calibration, initialisation and data labelling. Extensive experiments on real and simulated motions proved the reliability (maximum angular error less than 1°, maximum point reconstruction less than 1 mm) of the developed system, which is robust to false matching caused by marker occlusions. Moreover, orientation artifacts due to skin motion can be reduced by a factor of 50{\%}.",
keywords = "Biomechanical models, Extended Kalman filters, Human motion estimation, Motion tracking, Multi-camera systems, Rehabilitation",
author = "Pietro Cerveri and M. Rabuffetti and A. Pedotti and G. Ferrigno",
year = "2003",
month = "3",
doi = "10.1007/BF02344878",
language = "English",
volume = "41",
pages = "109--123",
journal = "Medical and Biological Engineering and Computing",
issn = "0140-0118",
publisher = "Springer Verlag",
number = "2",

}

TY - JOUR

T1 - Real-time human motion estimation using biomechanical models and non-linear state-space filters

AU - Cerveri, Pietro

AU - Rabuffetti, M.

AU - Pedotti, A.

AU - Ferrigno, G.

PY - 2003/3

Y1 - 2003/3

N2 - In the field of sports biomechanics and rehabilitation engineering, the possibility of computing, in real time, the angular displacements and derivatives of human joints, from a video of motion sequences, represents an appealing goal. In particular, applications of biofeedback protocols in rehabilitation can benefit from this capability. The focus of the investigation was concerned with the application of biomechanical models, comprising of a kinematic chain and surface envelopes, and state-space filters, to the computation, in real time and with high accuracy, of the angular data and derivatives. By minimising the distances, measured with TV cameras, between the 2D marker projections and the corresponding back-projected markers located on the mannequin, the configuration of the biomechanical model was automatically updated. The use of state-space estimation allowed the computation of smooth derivatives of the orientation data. Owing to the non-linearity of the functions involved, the derivatives of the observation model were obtained through a multidimensional extension of Stirling's interpolation formula. Proper algorithms were developed to cope with the model calibration, initialisation and data labelling. Extensive experiments on real and simulated motions proved the reliability (maximum angular error less than 1°, maximum point reconstruction less than 1 mm) of the developed system, which is robust to false matching caused by marker occlusions. Moreover, orientation artifacts due to skin motion can be reduced by a factor of 50%.

AB - In the field of sports biomechanics and rehabilitation engineering, the possibility of computing, in real time, the angular displacements and derivatives of human joints, from a video of motion sequences, represents an appealing goal. In particular, applications of biofeedback protocols in rehabilitation can benefit from this capability. The focus of the investigation was concerned with the application of biomechanical models, comprising of a kinematic chain and surface envelopes, and state-space filters, to the computation, in real time and with high accuracy, of the angular data and derivatives. By minimising the distances, measured with TV cameras, between the 2D marker projections and the corresponding back-projected markers located on the mannequin, the configuration of the biomechanical model was automatically updated. The use of state-space estimation allowed the computation of smooth derivatives of the orientation data. Owing to the non-linearity of the functions involved, the derivatives of the observation model were obtained through a multidimensional extension of Stirling's interpolation formula. Proper algorithms were developed to cope with the model calibration, initialisation and data labelling. Extensive experiments on real and simulated motions proved the reliability (maximum angular error less than 1°, maximum point reconstruction less than 1 mm) of the developed system, which is robust to false matching caused by marker occlusions. Moreover, orientation artifacts due to skin motion can be reduced by a factor of 50%.

KW - Biomechanical models

KW - Extended Kalman filters

KW - Human motion estimation

KW - Motion tracking

KW - Multi-camera systems

KW - Rehabilitation

UR - http://www.scopus.com/inward/record.url?scp=12444328760&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=12444328760&partnerID=8YFLogxK

U2 - 10.1007/BF02344878

DO - 10.1007/BF02344878

M3 - Article

C2 - 12691430

AN - SCOPUS:12444328760

VL - 41

SP - 109

EP - 123

JO - Medical and Biological Engineering and Computing

JF - Medical and Biological Engineering and Computing

SN - 0140-0118

IS - 2

ER -