Detail publikace

Determination of a Person’s Height from Images Using a Known Object Size

KŘÍŽ, P. ŘÍHA, K. DUTTA, M. SINGH, A. NECHITI, M.

Originální název

Determination of a Person’s Height from Images Using a Known Object Size

Anglický název

Determination of a Person’s Height from Images Using a Known Object Size

Jazyk

en

Originální abstrakt

This paper describes an image processing method for an automatic determination of a person’s height from image captured by a single camera. It is supposed that at least one object with a known size which defines the metric calibration is present in the image. Our global algorithm consists of several image processing techniques such as camera calibration, edge line detection, Hough transform, vanishing points detection and Histograms of oriented gradients for human detection. Original methodology was developed and tested in C++ as a standalone process which is applicable only on static frames (not optimized for video). The accuracy of the method is about 2.25 ± 2.79 cm, preliminary results were obtained using a database which contained 19 images of standing persons inside a room.

Anglický abstrakt

This paper describes an image processing method for an automatic determination of a person’s height from image captured by a single camera. It is supposed that at least one object with a known size which defines the metric calibration is present in the image. Our global algorithm consists of several image processing techniques such as camera calibration, edge line detection, Hough transform, vanishing points detection and Histograms of oriented gradients for human detection. Original methodology was developed and tested in C++ as a standalone process which is applicable only on static frames (not optimized for video). The accuracy of the method is about 2.25 ± 2.79 cm, preliminary results were obtained using a database which contained 19 images of standing persons inside a room.

Dokumenty

BibTex


@inproceedings{BUT148930,
  author="Petr {Kříž} and Kamil {Říha} and Malay Kishore {Dutta} and Anushikha {Singh} and Marius {Nechiti}",
  title="Determination of a Person’s Height from Images Using a Known Object Size",
  annote="This paper describes an image processing method for an automatic determination of a person’s height from image captured by a single camera. It is supposed that at least one object with a known size which defines the metric calibration is present in the image. Our global algorithm consists of several image processing techniques such as camera calibration, edge line detection, Hough transform, vanishing points detection and Histograms of oriented gradients for human detection. Original methodology was developed and tested in C++ as a standalone process which is applicable only on static frames (not optimized for video). The accuracy of the method is about 2.25 ± 2.79 cm, preliminary results were obtained using a database which contained 19 images of standing persons inside a room.",
  address="Faculty of Electrical Engineering and Communacation, Brno University of Technology",
  booktitle="Proceedings of the 2018 41st International Conference on Telecommunications and Signal Processing",
  chapter="148930",
  howpublished="online",
  institution="Faculty of Electrical Engineering and Communacation, Brno University of Technology",
  year="2018",
  month="july",
  pages="384--387",
  publisher="Faculty of Electrical Engineering and Communacation, Brno University of Technology",
  type="conference paper"
}