Efficient Change Detection for Very Large Motion Blurred Images
Abstract
In this paper, we address the challenging problem of registration and change detection in very large motion blurred images. The unreasonable demand that this task puts on computational and memory resources precludes the possibility of any direct attempt at solving this problem. We address this issue by observing the fact that the camera motion experienced by a sufficiently large sub-image is approximately the same as that of the entire image itself. We devise an algorithm for judicious sub-image selection so that the camera motion can be deciphered correctly, irrespective of the presence or absence of occluder. We follow a reblur-difference framework to detect changes as this is an artifact-free pipeline unlike the traditional deblur-difference approach. We demonstrate the results of our algorithm on both synthetic and real data.
Cite
Text
Rengarajan et al. "Efficient Change Detection for Very Large Motion Blurred Images." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2014. doi:10.1109/CVPRW.2014.55Markdown
[Rengarajan et al. "Efficient Change Detection for Very Large Motion Blurred Images." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2014.](https://mlanthology.org/cvprw/2014/rengarajan2014cvprw-efficient/) doi:10.1109/CVPRW.2014.55BibTeX
@inproceedings{rengarajan2014cvprw-efficient,
title = {{Efficient Change Detection for Very Large Motion Blurred Images}},
author = {Rengarajan, Vijay and Punnappurath, Abhijith and Rajagopalan, A. N. and Seetharaman, Guna},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2014},
pages = {315-322},
doi = {10.1109/CVPRW.2014.55},
url = {https://mlanthology.org/cvprw/2014/rengarajan2014cvprw-efficient/}
}