ML Anthology

Robust Background-Subtraction for Person Detection in Thermal Imagery

James W. Davis, Vinay Sharma
CVPR 2004 pp. 128
doi:10.1109/CVPR.2004.431 /cvpr/2004/davis2004cvpr-robust/

Abstract

We present a new contour-based background-subtraction technique to detect people in widely varying thermal imagery. Statistical background-subtraction is first used to identify local regions-of-interest. Within each region, gradient information in the foreground and background are combined to form a contour saliency map. After thinning, an A* path-constrained search along watershed boundaries is used to complete any broken contour segments. Lastly, the contour image is flood-filled to produce silhouettes. Results are presented that demonstrate the robustness of the approach to detect people across a wide range of thermal imagery using a fixed set of parameters.

PDF CVPR Semantic Scholar

Cite

Text

Davis and Sharma. "Robust Background-Subtraction for Person Detection in Thermal Imagery." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2004. doi:10.1109/CVPR.2004.431

Markdown

[Davis and Sharma. "Robust Background-Subtraction for Person Detection in Thermal Imagery." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2004.](https://mlanthology.org/cvpr/2004/davis2004cvpr-robust/) doi:10.1109/CVPR.2004.431

BibTeX

@inproceedings{davis2004cvpr-robust,
  title     = {{Robust Background-Subtraction for Person Detection in Thermal Imagery}},
  author    = {Davis, James W. and Sharma, Vinay},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2004},
  pages     = {128},
  doi       = {10.1109/CVPR.2004.431},
  url       = {https://mlanthology.org/cvpr/2004/davis2004cvpr-robust/}
}