Details
| Original language | English |
|---|---|
| Title of host publication | ACCV 2014 |
| Subtitle of host publication | Computer Vision -- ACCV 2014 |
| Pages | 692-707 |
| Number of pages | 16 |
| Volume | 9006 |
| Publication status | Published - 17 Apr 2015 |
| Event | 12th Asian Conference on Computer Vision, ACCV 2014 - Singapore, Singapore Duration: 1 Nov 2014 → 5 Nov 2014 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Publisher | Springer Verlag |
| ISSN (Print) | 0302-9743 |
Abstract
A wide variety of computer vision applications rely on superpixel or supervoxel algorithms as a preprocessing step. This underlines the overall importance that these algorithms have gained in the recent years. However, most methods show a lack of temporal consistency or fail in producing temporally stable segmentations. In this paper, we propose a novel, contour-based approach that generates temporally consistent superpixels for video content. It can be expressed in an expectationmaximization framework and utilizes an efficient label propagation built on backward optical flow in order to encourage the preservation of superpixel shapes and their spatial constellation over time. Using established benchmark suites, we show the superior performance of our approach compared to state of the art supervoxel and superpixel algorithms for video content.
ASJC Scopus subject areas
- Mathematics(all)
- Theoretical Computer Science
- Computer Science(all)
- General Computer Science
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
ACCV 2014: Computer Vision -- ACCV 2014. Vol. 9006 2015. p. 692-707 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Superpixels for video content using a contour-based EM optimization
AU - Reso, Matthias
AU - Jachalsky, Jörn
AU - Rosenhahn, Bodo
AU - Ostermann, Jörn
PY - 2015/4/17
Y1 - 2015/4/17
N2 - A wide variety of computer vision applications rely on superpixel or supervoxel algorithms as a preprocessing step. This underlines the overall importance that these algorithms have gained in the recent years. However, most methods show a lack of temporal consistency or fail in producing temporally stable segmentations. In this paper, we propose a novel, contour-based approach that generates temporally consistent superpixels for video content. It can be expressed in an expectationmaximization framework and utilizes an efficient label propagation built on backward optical flow in order to encourage the preservation of superpixel shapes and their spatial constellation over time. Using established benchmark suites, we show the superior performance of our approach compared to state of the art supervoxel and superpixel algorithms for video content.
AB - A wide variety of computer vision applications rely on superpixel or supervoxel algorithms as a preprocessing step. This underlines the overall importance that these algorithms have gained in the recent years. However, most methods show a lack of temporal consistency or fail in producing temporally stable segmentations. In this paper, we propose a novel, contour-based approach that generates temporally consistent superpixels for video content. It can be expressed in an expectationmaximization framework and utilizes an efficient label propagation built on backward optical flow in order to encourage the preservation of superpixel shapes and their spatial constellation over time. Using established benchmark suites, we show the superior performance of our approach compared to state of the art supervoxel and superpixel algorithms for video content.
UR - http://www.scopus.com/inward/record.url?scp=84983611848&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-16817-3_45
DO - 10.1007/978-3-319-16817-3_45
M3 - Conference contribution
AN - SCOPUS:84983611848
VL - 9006
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 692
EP - 707
BT - ACCV 2014
T2 - 12th Asian Conference on Computer Vision, ACCV 2014
Y2 - 1 November 2014 through 5 November 2014
ER -