Measurement of human stature from surveillance camera based on projective geometry

Takuya Inoue, Hideo Saito, Makoto Kimura

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

In this paper, we propose a novel measurement method of human stature by single view surveillance camera with minimum and sufficient effort of camera calibration. In this method, we calibrate the camera only towards the vertical direction to the ground by using multiple poles with known length which are placed at the arbitrary positions. Even though the horizontal positions of the poles are unknown, we can obtain the geometrical information which is sufficient to estimate the vertical length of the human according to the length of the poles. In the estimation of the human stature, we compute the homography between the plane along with the tips of the poles (pole-top plane) and the ground plane. Using this homography, we can determine any image point that has the same height from the ground, then we can determine the point of the pole length height in the human region. By comparing with the point and the head top point of the human, we can compute the stature with real scale. We have experimentally confirmed that proposed method can be more accurate and easier measurement than the measurement by the traditional camera calibration method using a chess board.

Original languageEnglish
Title of host publicationProceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011
Pages508-511
Number of pages4
Publication statusPublished - 2011
Event12th IAPR Conference on Machine Vision Applications, MVA 2011 - Nara, Japan
Duration: 2011 Jun 132011 Jun 15

Other

Other12th IAPR Conference on Machine Vision Applications, MVA 2011
CountryJapan
CityNara
Period11/6/1311/6/15

Fingerprint

Poles
Cameras
Geometry
Calibration

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition

Cite this

Inoue, T., Saito, H., & Kimura, M. (2011). Measurement of human stature from surveillance camera based on projective geometry. In Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011 (pp. 508-511)

Measurement of human stature from surveillance camera based on projective geometry. / Inoue, Takuya; Saito, Hideo; Kimura, Makoto.

Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011. 2011. p. 508-511.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Inoue, T, Saito, H & Kimura, M 2011, Measurement of human stature from surveillance camera based on projective geometry. in Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011. pp. 508-511, 12th IAPR Conference on Machine Vision Applications, MVA 2011, Nara, Japan, 11/6/13.
Inoue T, Saito H, Kimura M. Measurement of human stature from surveillance camera based on projective geometry. In Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011. 2011. p. 508-511
Inoue, Takuya ; Saito, Hideo ; Kimura, Makoto. / Measurement of human stature from surveillance camera based on projective geometry. Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011. 2011. pp. 508-511
@inproceedings{39cbf424e53144f2b45b1c69a86e72c0,
title = "Measurement of human stature from surveillance camera based on projective geometry",
abstract = "In this paper, we propose a novel measurement method of human stature by single view surveillance camera with minimum and sufficient effort of camera calibration. In this method, we calibrate the camera only towards the vertical direction to the ground by using multiple poles with known length which are placed at the arbitrary positions. Even though the horizontal positions of the poles are unknown, we can obtain the geometrical information which is sufficient to estimate the vertical length of the human according to the length of the poles. In the estimation of the human stature, we compute the homography between the plane along with the tips of the poles (pole-top plane) and the ground plane. Using this homography, we can determine any image point that has the same height from the ground, then we can determine the point of the pole length height in the human region. By comparing with the point and the head top point of the human, we can compute the stature with real scale. We have experimentally confirmed that proposed method can be more accurate and easier measurement than the measurement by the traditional camera calibration method using a chess board.",
author = "Takuya Inoue and Hideo Saito and Makoto Kimura",
year = "2011",
language = "English",
isbn = "9784901122115",
pages = "508--511",
booktitle = "Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011",

}

TY - GEN

T1 - Measurement of human stature from surveillance camera based on projective geometry

AU - Inoue, Takuya

AU - Saito, Hideo

AU - Kimura, Makoto

PY - 2011

Y1 - 2011

N2 - In this paper, we propose a novel measurement method of human stature by single view surveillance camera with minimum and sufficient effort of camera calibration. In this method, we calibrate the camera only towards the vertical direction to the ground by using multiple poles with known length which are placed at the arbitrary positions. Even though the horizontal positions of the poles are unknown, we can obtain the geometrical information which is sufficient to estimate the vertical length of the human according to the length of the poles. In the estimation of the human stature, we compute the homography between the plane along with the tips of the poles (pole-top plane) and the ground plane. Using this homography, we can determine any image point that has the same height from the ground, then we can determine the point of the pole length height in the human region. By comparing with the point and the head top point of the human, we can compute the stature with real scale. We have experimentally confirmed that proposed method can be more accurate and easier measurement than the measurement by the traditional camera calibration method using a chess board.

AB - In this paper, we propose a novel measurement method of human stature by single view surveillance camera with minimum and sufficient effort of camera calibration. In this method, we calibrate the camera only towards the vertical direction to the ground by using multiple poles with known length which are placed at the arbitrary positions. Even though the horizontal positions of the poles are unknown, we can obtain the geometrical information which is sufficient to estimate the vertical length of the human according to the length of the poles. In the estimation of the human stature, we compute the homography between the plane along with the tips of the poles (pole-top plane) and the ground plane. Using this homography, we can determine any image point that has the same height from the ground, then we can determine the point of the pole length height in the human region. By comparing with the point and the head top point of the human, we can compute the stature with real scale. We have experimentally confirmed that proposed method can be more accurate and easier measurement than the measurement by the traditional camera calibration method using a chess board.

UR - http://www.scopus.com/inward/record.url?scp=84872578422&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872578422&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9784901122115

SP - 508

EP - 511

BT - Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011

ER -