GF1MS-WHU and GF2MS-WHU

Citation Author(s):
Zhu
Shaocong
Li
Zhiwei
Shen
Huanfeng
Submitted by:
Zhu Shaocong
Last updated:
Mon, 07/08/2024 - 15:58
DOI:
10.21227/8vc2-yf66
License:
Creative Commons Attribution
24 Views
Categories:
Image Processing
Keywords:
Cloud Detection
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

Two novel datasets GF1MS-WHU and GF2MS-WHU are introduced for cloud detection. The GF1MS-WHU dataset consists of 141 unlabeled and 33 well-annotated 8-m Gaofen-1 multispectral (GF1-MS) images. Furthermore, the GF2MS-WHU dataset includes 163 unlabeled and 29 well-annotated 4-m Gaofen-2 multispectral (GF2-MS) images. Based on the labeled images in the two datasets, a total of 10428 and 21917 fully labeled image patches are available.

The acquisition of GF-1 images spanned from June 2014 to December 2020 and encompassed four MS bands in both PMS sensors.

The acquisition of GF-2 images spanned from June 2014 to October 2020 and encompassed four MS bands in both PMS sensors.

Instructions: 

tiff is the original image file

_Mask.tif is its corresponding mask file

TrainList.txt is the training data list

TestList.txt is the test data list

Dataset Files

LOGIN TO ACCESS DATASET FILES

 

QUESTIONS?