Really Quick Shift: Image Segmentation on a GPU

Abstract

The paper presents an exact GPU implementation of the quick shift image segmentation algorithm. Variants of the implementation which use global memory and texture caching are presented, and the paper shows that a method backed by texture caching can produce a 10-50X speedup for practical images, making computation of super-pixels possible at 5-10Hz on modest sized (256x256) images.

Cite

Text

Fulkerson and Soatto. "Really Quick Shift: Image Segmentation on a GPU." European Conference on Computer Vision Workshops, 2010. doi:10.1007/978-3-642-35740-4_27

Markdown

[Fulkerson and Soatto. "Really Quick Shift: Image Segmentation on a GPU." European Conference on Computer Vision Workshops, 2010.](https://mlanthology.org/eccvw/2010/fulkerson2010eccvw-really/) doi:10.1007/978-3-642-35740-4_27

BibTeX

@inproceedings{fulkerson2010eccvw-really,
  title     = {{Really Quick Shift: Image Segmentation on a GPU}},
  author    = {Fulkerson, Brian and Soatto, Stefano},
  booktitle = {European Conference on Computer Vision Workshops},
  year      = {2010},
  pages     = {350-358},
  doi       = {10.1007/978-3-642-35740-4_27},
  url       = {https://mlanthology.org/eccvw/2010/fulkerson2010eccvw-really/}
}