ML Anthology

Unscented Transformation for Depth from Motion-Blur in Videos

Chandramouli Paramanand, A. N. Rajagopalan
CVPRW 2010 pp. 38-44
doi:10.1109/CVPRW.2010.5543835 /cvprw/2010/paramanand2010cvprw-unscented/

Abstract

In images and videos of a 3D scene, blur due to camera shake can be a source of depth information. Our objective is to find the shape of the scene from its motion-blurred observations without having to restore the original image. In this paper, we pose depth recovery as a recursive state estimation problem. We show that the relationship between the observation and the scale factor of the motion-blur kernel associated with the depth at a point is nonlinear and propose the use of the unscented Kalman filter for state estimation. The performance of the proposed method is evaluated on many examples.

CVPRW Semantic Scholar

Cite

Text

Paramanand and Rajagopalan. "Unscented Transformation for Depth from Motion-Blur in Videos." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2010. doi:10.1109/CVPRW.2010.5543835

Markdown

[Paramanand and Rajagopalan. "Unscented Transformation for Depth from Motion-Blur in Videos." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2010.](https://mlanthology.org/cvprw/2010/paramanand2010cvprw-unscented/) doi:10.1109/CVPRW.2010.5543835

BibTeX

@inproceedings{paramanand2010cvprw-unscented,
  title     = {{Unscented Transformation for Depth from Motion-Blur in Videos}},
  author    = {Paramanand, Chandramouli and Rajagopalan, A. N.},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2010},
  pages     = {38-44},
  doi       = {10.1109/CVPRW.2010.5543835},
  url       = {https://mlanthology.org/cvprw/2010/paramanand2010cvprw-unscented/}
}