File size: 1,653 Bytes
b6779d1 3dab24f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 |
---
license: unknown
---
# Description
Jasper Ridge is a popular hyperspectral data used in [[enviTutorials](http://www.cossa.csiro.au/hswww/Overview.htm), [SS-NMF](http://arxiv.org/abs/1403.4682), [DgS-NMF](http://www.sciencedirect.com/science/article/pii/S0924271613002761), [RRLbS](http://arxiv.org/abs/1409.0685), [L1-CENMF](http://arxiv.org/abs/1305.7311)]. There are 512 x 614 pixels in it. Each pixel is recorded at 224 channels ranging from 380 nm to 2500 nm. The spectral resolution is up to 9.46nm. Since this hyperspectral image is too complex to get the ground truth, we consider a subimage of 100 x 100 pixels. The first pixel starts from the (105,269)-th pixel in the original image. After removing the channels 1--3, 108--112, 154--166 and 220--224 (due to dense water vapor and atmospheric effects), we remain 198 channels (this is a common preprocess for HU analyses). There are four endmembers latent in this data: "#1 Road", "#2 Soil", "#3 Water" and "#4 Tree".
# Quick look
<figure>
<img src= "assets/1908991_orig.jpg" alt="Jasper Ridge" width="500" />
<figcaption>Jasper Ridge and its ground truth (GT:abundances and GT:endmembers).</figcaption>
</figure>
# Credits
Dataset originally collected by Feiyun Zhu and originally available at: http://www.escience.cn/people/feiyunZHU/Dataset_GT.html
To use this dataset, cite the associated paper:
```
@misc{zhu2017hyperspectral,
title={Hyperspectral Unmixing: Ground Truth Labeling, Datasets, Benchmark Performances and Survey},
author={Feiyun Zhu},
year={2017},
eprint={1708.05125},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
``` |