Convert pixel coordinates to latitude, longitude with the Sentinel-1 image

[NS-Nik]

Question
Hi!

I am trying to convert pixel coordinates to the geographical latitude, longitude. I am following steps, previously describe in this discussion: #51

However, I cannot retrieve information about CRS, coordinates of upper left corner and resolution.
I use rasterio:

src = rasterio.open(image)     
transform = src.transform 

But it seems like the standard GeoTIFF Sentinel-1 format does not contain the necessary information and transformation matrix looks like this:

print("Affine Transform:", transform)
Affine Transform:
 | 1.00, 0.00, 0.00|
| 0.00, 1.00, 0.00|
| 0.00, 0.00, 1.00|

Any ideas how to solve this problem?

Thank you a lot in advance for your help!

[mlubej]

Hi @NS-Nik!

Do you have gdalinfo in your command-line tool? Could you post the output for the tiff in question? Or share the file via some service, if it isn't too large.

[NS-Nik]

HI @mlubej!
Thank you for your reply!

The image size is quite big (~1.7 Gb), so here is the output from the gdalinfo:

metadata.txt

[mlubej]

Hi @NS-Nik!

Thanks for the command output. I wanted to check if the CRS info is actually available in the tif. Based on what you sent, it seems to be present, you just might not be accessing it correctly.

You mention that you are interested in the following:

However, I cannot retrieve information about CRS, coordinates of upper left corner and resolution.

I believe you should be able to extract the above by using other attributes rather than the transform, such as:

import rasterio

with rasterio.open(path_to_tif) as src:
    x1, y1, x2, y2 = src.bounds
    height = src.height
    width = src.width
    crs = src.crs.to_epsg()

res_x = (x2 - x1) // width
res_y = (y2 - y1) // height

This should work for the tiff based on what you sent me. You can then check if these values are also what you expect to get based on the gdalinfo output.

Hope this works!

[NS-Nik]

@mlubej Many thanks for your reply!

I still a little confused. Based on the information from the metadata (and from src.crs.to_epsg() command) , the srs is 4326 which is WGS 84.

However, the corner coordinates from the src.bounds do not look like they in the WGS 84.

Maybe you can take a look at the geotiff image?
https://drive.google.com/file/d/18MtufFSqfqm4F8xDG_FfJFwE3ewhShHh/view?usp=sharing

[mlubej]

Hmm, you are right @NS-Nik, the bounds or the CRS don't seem to match and there is some issue with the tif.

Is this the original tif, or did you get it somewhere else, or produced it yourself?

[NS-Nik]

@mlubej
This is the satellite image and it went through a series of processing steps after I downloaded it.
I have tried to do the same but with the different image (without implementing any additional processing steps).
Here is the metadata from this image:
meta.txt

The difference from the previous one - the information appears on the ground control points.

However, the problem stays the same: the Corner Coordinates are still not in the WGS 84 format.

[mlubej]

It seems that one or more of your processing steps after downloading are not done correctly, as the transformation information gets lost.

I would suggest you take a look into each of the processing steps after downloading and check when the data gets corrupted.