Data and results: Variational Multi-Phase Segmentation using High-Dimensional Local Features
Data and results used in our paper presented at the IEEE Winter Conference on Applications of Computer Vision (WACV 2016) - see reference below.
CONTENTS:
- Outex_US_00000 texture segmentation test suite (converted to .png format)
- Crystal images (grains with 1,2 and 3 segments, CMS-GaAs two-phase crystal)
- Link to ICPR 2014 contest dataset
- All corresponding segmentations produced by our method PCA-MS
|- data/
|+ - Crystals ground truth crystal images (Figure 2)
|+ - ICPR2014 README.txt with link to the ICPR 2014 contest dataset
|+ - Outex: ground truth segments and texture mosaics
| from the Outex_US_00000 test suite converted from .ras to .png
|- results/
|+ - Crystals segmentations (.png) and boundary curves (.pdf)
|+ - ICPR2014 segmentation results used in Table 2
|+ + - Raw before TxtMerge post-processing
|+ + - TxtMerge after TxtMerge post-processing
|+ + - README.txt link to Prague website with benchmark results
|+ - Outex segmentation results used in Table 1
|+ + - README.txt instructions on how to run quantitative evaluation
|+ + - Clustering k-means clustering
|+ + - FSEG_ICPR2014 ICPR2014 version of FSEG (see link below)
|+ + - FSEGstar FSEG with adjusted spectral histograms (see paper)
|+ + - FSEGstar-TxtMerge FSEGstar without TxtMerge post-processing
The comparison with FSEG is based on the ICPR 2014 version.
QUANTITATIVE EVALUATION:
- ICPR 2014 benchmark: The Prague Texture Segmentation Datagenerator and Benchmark
- Outex_US_00000 test suite: results/Outex/README.txt
REFERENCE: Mevenkamp, N., and Berkels, B. Variational Multi-Phase Segmentation using High-Dimensional Local Features. Applications of Computer Vision (WACV), 2016 IEEE Winter Conference on, 2016, (accepted)
CONTACT:
Niklas Mevenkamp
Aachen Institute for Advanced Study in Computational Engineering Science
RWTH Aachen University
mevenkamp@aices.rwth-aachen.de
http://www.aices.rwth-aachen.de/people/mevenkamp