Datasets
Standard Dataset
TJ_CS_Hyperspectral
- Citation Author(s):
-
XuebinTang
National University of Defense Technology - Submitted by:
- Xuebin Tang
- Last updated:
- Thu, 03/07/2024 - 09:59
- DOI:
- 10.21227/d70b-r620
- Data Format:
- Research Article Link:
- License:
126 Views- Categories:
- Keywords:
Abstract
We employed Hexacopter unmanned aerial vehicle (UAV) equipped with the SPECIM FX17E hyperspectral camera to implement ultra-low-altitude flight aerial photography missions with atmospheric correction processing. We collected three hyperspectral images and combined them into three data pairs, which exhibit varying degrees of spectral shift. Among them, a hyperspectral image including six types of ground objects was collected in Changsha at 4:00 pm on September 27, 2021, with sunny weather and a flight altitude of 30m, named CSSunny. Besides, we captured two hyperspectral images in Tianjin, both of them contain twelve types of ground objects and the flight altitude was 60m. One of them was shot at 3:00 pm on March 21, 2023, with sunny weather, and the other was shot at 3:30 pm on March 22, 2023, with cloudy weather, named TJSunny and TJCloudy respectively. The spatial resolution of three images in the self-collected data pairs is 482 × 640, and the spectral wavelength of them ranges from 935nm to 1720nm with 224 bands.
We combine three self-collected images into three data pairs, namely TJSunny-TJCloudy, TJSunny-CSSunny, and TJCloudy-CSSunny. In particular, the categories of the HSI images in TJSunny-CSSunny and TJCloudy-CSSunny scenes are not the same, and we select six classes shared for experiments. We recorded the number of samples in each class of self-collected data pairs at the end of the Readme file. All images and descriptions about the dataset are also in the Readme file. The labeling process of our datasets is completed on the Matlab platform.
Dataset Files
- Data of ChangSha under sunny weather data_CSSunny.mat (224.66 MB)
- Data of TianJin under cloudy weather data_TJCloudy.mat (70.42 MB)
- Data of TianJin under sunny weather data_TJSunny.mat (66.47 MB)
- ground-truth of ChangSha under sunny weather (6 classes) label_CSSunny_6.mat (2.62 kB)
- ground-truth of TianJin under cloudy weather (6 classes) label_TJCloudy_6.mat (2.29 kB)
- ground-truth of TianJin under cloudy weather (12 classes) label_TJCloudy_12.mat (4.28 kB)
- ground-truth of TianJin under sunny weather (6 classes) label_TJSunny_6.mat (2.74 kB)
- ground-truth of TianJin under sunny weather (12 classes) label_TJSunny_12.mat (4.49 kB)
- All images and descriptions related to our datasets. Scripts.zip (4.28 MB)
Documentation
| Attachment | Size |
|---|---|
 Description of our datasets. | 527.62 KB |
Comments
We sincerely hope that these datasets can be valuable for subsequent studies in the hyperspectral community.