Example-based pose estimation in monocular images using compact fourier descriptors

Abstract

Automatically estimating human poses from visual input is useful but challenging due to variations in image space and the high dimensionality of the pose space. In this paper, we assume that a human silhouette can be extracted from monocular visual input. We compare the recovery performance of Fourier descriptors with a number of coefficients between 8 and 128, and two different sampling methods. An examplebased approach is taken to recover upper body poses from the descriptors. We test the robustness of our approach by investigating how shape deformations due to changes in body dimensions, viewpoint and noise affect the recovery of the pose. The average error per joint is approximately 16-17° for equidistant sampling and slightly higher for extreme point sampling. Increasing the number of descriptors does not have any influence on the performance. Noise and small changes in viewpoint have only a very small effect on the recovery performance but we obtain higher error scores when recovering poses using silhouettes from a person with different body dimensions.
Original languageUndefined
Place of PublicationEnschede
PublisherINF-Centre for Telematics and Information Techn.
Number of pages19
StatePublished - 2005

Publication series

NameTR-CTIT-05, ISSN 1381-3625
PublisherUniversity of Twente, Centre for Telematics and Information Technology
No.49

Fingerprint

Sampling
Recovery

Keywords

  • IR-53979
  • EWI-1881
  • EC Grant Agreement nr.: FP6/506811
  • METIS-227388

Cite this

Poppe, R. W., & Poel, M. (2005). Example-based pose estimation in monocular images using compact fourier descriptors. (TR-CTIT-05, ISSN 1381-3625; No. 49). Enschede: INF-Centre for Telematics and Information Techn.

Poppe, Ronald Walter; Poel, Mannes / Example-based pose estimation in monocular images using compact fourier descriptors.

Enschede : INF-Centre for Telematics and Information Techn., 2005. 19 p. (TR-CTIT-05, ISSN 1381-3625; No. 49).

Research output: ProfessionalReport

@book{5485aed7606c4a4caab7854878b421b0,
title = "Example-based pose estimation in monocular images using compact fourier descriptors",
abstract = "Automatically estimating human poses from visual input is useful but challenging due to variations in image space and the high dimensionality of the pose space. In this paper, we assume that a human silhouette can be extracted from monocular visual input. We compare the recovery performance of Fourier descriptors with a number of coefficients between 8 and 128, and two different sampling methods. An examplebased approach is taken to recover upper body poses from the descriptors. We test the robustness of our approach by investigating how shape deformations due to changes in body dimensions, viewpoint and noise affect the recovery of the pose. The average error per joint is approximately 16-17° for equidistant sampling and slightly higher for extreme point sampling. Increasing the number of descriptors does not have any influence on the performance. Noise and small changes in viewpoint have only a very small effect on the recovery performance but we obtain higher error scores when recovering poses using silhouettes from a person with different body dimensions.",
keywords = "IR-53979, EWI-1881, EC Grant Agreement nr.: FP6/506811, METIS-227388",
author = "Poppe, {Ronald Walter} and Mannes Poel",
note = "Imported from HMI",
year = "2005",
series = "TR-CTIT-05, ISSN 1381-3625",
publisher = "INF-Centre for Telematics and Information Techn.",
number = "49",

}

Poppe, RW & Poel, M 2005, Example-based pose estimation in monocular images using compact fourier descriptors. TR-CTIT-05, ISSN 1381-3625, no. 49, INF-Centre for Telematics and Information Techn., Enschede.

Example-based pose estimation in monocular images using compact fourier descriptors. / Poppe, Ronald Walter; Poel, Mannes.

Enschede : INF-Centre for Telematics and Information Techn., 2005. 19 p. (TR-CTIT-05, ISSN 1381-3625; No. 49).

Research output: ProfessionalReport

TY - BOOK

T1 - Example-based pose estimation in monocular images using compact fourier descriptors

AU - Poppe,Ronald Walter

AU - Poel,Mannes

N1 - Imported from HMI

PY - 2005

Y1 - 2005

N2 - Automatically estimating human poses from visual input is useful but challenging due to variations in image space and the high dimensionality of the pose space. In this paper, we assume that a human silhouette can be extracted from monocular visual input. We compare the recovery performance of Fourier descriptors with a number of coefficients between 8 and 128, and two different sampling methods. An examplebased approach is taken to recover upper body poses from the descriptors. We test the robustness of our approach by investigating how shape deformations due to changes in body dimensions, viewpoint and noise affect the recovery of the pose. The average error per joint is approximately 16-17° for equidistant sampling and slightly higher for extreme point sampling. Increasing the number of descriptors does not have any influence on the performance. Noise and small changes in viewpoint have only a very small effect on the recovery performance but we obtain higher error scores when recovering poses using silhouettes from a person with different body dimensions.

AB - Automatically estimating human poses from visual input is useful but challenging due to variations in image space and the high dimensionality of the pose space. In this paper, we assume that a human silhouette can be extracted from monocular visual input. We compare the recovery performance of Fourier descriptors with a number of coefficients between 8 and 128, and two different sampling methods. An examplebased approach is taken to recover upper body poses from the descriptors. We test the robustness of our approach by investigating how shape deformations due to changes in body dimensions, viewpoint and noise affect the recovery of the pose. The average error per joint is approximately 16-17° for equidistant sampling and slightly higher for extreme point sampling. Increasing the number of descriptors does not have any influence on the performance. Noise and small changes in viewpoint have only a very small effect on the recovery performance but we obtain higher error scores when recovering poses using silhouettes from a person with different body dimensions.

KW - IR-53979

KW - EWI-1881

KW - EC Grant Agreement nr.: FP6/506811

KW - METIS-227388

M3 - Report

T3 - TR-CTIT-05, ISSN 1381-3625

BT - Example-based pose estimation in monocular images using compact fourier descriptors

PB - INF-Centre for Telematics and Information Techn.

ER -

Poppe RW, Poel M. Example-based pose estimation in monocular images using compact fourier descriptors. Enschede: INF-Centre for Telematics and Information Techn., 2005. 19 p. (TR-CTIT-05, ISSN 1381-3625; 49).