I have a question about the resolution of super-resolved Sentinel-2 images.
Although it is said that the resolution of all bands is resampled to 10m after super-resolving, as the previous post mentioned (Sentinel-2 Super-Resolution (all bands at 10m): Snap plugin now available! - #21 by Kat.Hurst), I do not think all the bands are resampled to 10m. When I visualize the bands whose native resolution is 60m, the resampled resolution looks like 20m.
I assume that the pixel size is really 10m. You can use the range finder to measure the size of pixel. Or you can look into the Geo-Coding information panel (Menu: Analysis / Geo-Coding)
Probably the up-sampled images from 60m to 10m don’t look as sharp and crisp as the 10m images, but that due the the resampling. Even if they look like 20m, it is a nice result. @nicolas.brodu Maybe you add a few words?
Thank you for your replies.
I understand why not all the images are fine, but it is due to their original resolution.
Besides that, I did not know that I could check the size of pixels using the Geo-Coding. I am totally fine for the question.
I have another question about the super-resolved images.
Can I calculate some index, e.g., NDVI, by using them?
I am not sure that I can use them for that purpose as this method is similar to pan sharpening. If the super-resolved images are considered as same as pan-sharpened images, I think I cannot use them to calculate the indexes.
The bands one usually uses for NDVI (B4 and B8) are originally in 10 meters, so you can use the L1C.
And I think the bands do not change when applying the super-resolution plugin. You can check the values to verify this.
For NDVI an L2A product might be better, but I don’t know if the plugin can use L2A data.