hello
I have a project with Landsat data time series. but in some images there are clouds and I must detect and mask them. is there an index or something like that to detect clouds?
Amir Tahooni, you can use the Quality Assessment (QA) bands from Level 1 or Level 2 Landsat data to identify and mask pixels that are affected by clouds or shadows. I would suggest you to order and download the pixel_qa band of each L2 Landsat scene from https://espa.cr.usgs.gov/ by uploading the Landsat Product IDs. Secondly, use the link attached below to download the Toolbox that will help you to extract the pixel_qa bands. The Level 2 pixel_qa band when extracted contains pixel information about clouds, shadows, etc. Finally use the pixels that are already classified as clouds as a mask.
Also follow the link of an existing discussion on the topic
- https://www.usgs.gov/land-resources/nli/landsat/landsat-quality-assessment-tools?qt-science_support_page_related_con=4#qt-science_support_page_related_con
- https://www.researchgate.net/post/How_can_I_remove_cloud_from_landsat_satellite_image
Hi Amir,
maybe you can extract some information from this paper: Article Total and partial cloud amount detection during summer 2005 ...
Best regards
Henner
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing