ML Anthology

Motion-Based Background Subtraction Using Adaptive Kernel Density Estimation

Anurag Mittal, Nikos Paragios
CVPR 2004 pp. 302-309
doi:10.1109/CVPR.2004.164 /cvpr/2004/mittal2004cvpr-motion/

Abstract

Background modeling is an important component of many vision systems. Existing work in the area has mostly addressed scenes that consist of static or quasi-static structures. When the scene exhibits a persistent dynamic behavior in time, such an assumption is violated and detection performance deteriorates. In this paper, we propose a new method for the modeling and subtraction of such scenes. Towards the modeling of the dynamic characteristics, optical flow is computed and utilized as a feature in a higher dimensional space. Inherent ambiguities in the computation of features are addressed by using a data-dependent bandwidth for density estimation using kernels. Extensive experiments demonstrate the utility and performance of the proposed approach.

CVPR Semantic Scholar

Cite

Text

Mittal and Paragios. "Motion-Based Background Subtraction Using Adaptive Kernel Density Estimation." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2004. doi:10.1109/CVPR.2004.164

Markdown

[Mittal and Paragios. "Motion-Based Background Subtraction Using Adaptive Kernel Density Estimation." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2004.](https://mlanthology.org/cvpr/2004/mittal2004cvpr-motion/) doi:10.1109/CVPR.2004.164

BibTeX

@inproceedings{mittal2004cvpr-motion,
  title     = {{Motion-Based Background Subtraction Using Adaptive Kernel Density Estimation}},
  author    = {Mittal, Anurag and Paragios, Nikos},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2004},
  pages     = {302-309},
  doi       = {10.1109/CVPR.2004.164},
  url       = {https://mlanthology.org/cvpr/2004/mittal2004cvpr-motion/}
}