Cropping with GDAL python outputs a black section in the image

Question

I am trying to crop an image using GDAL python. I have tried using both gdal.Warp and gdal.Translate, and both keep bringing the same result.

I have an image that I have orthorectified using gdal.Warp called imortho. I want to select a small window from this image:

window = [-76.845, 39.054, -76.805, 39.012]

This is what I have tried:

gdal.Translate(outfile, imortho, projWin = window)

The size of the square is accurate, but almost half of the image has no data. The original image has data on the entirety of the window (and beyond).

When I try to do the same, specifying the projWinSRS option the result is the same. When I try to use the gdal.Warp function instead:

gdal.Warp(outfile, imortho, outputBounds = window)

The result is similar (even though the nodata region is in a different place). See the image:

Is this a bug or is there anything I'm not doing correctly to crop the data?

Answer 1

Look at this code:

• Use a irregular geometry

clip = gdal.Warp(destNameOrDestDS = outputpath, 
                 srcDSOrSrcDSTab  = intputpath, 
                 cutlineDSName    = shapefile or geojson, 
                 cropToCutline    = True,
                 copyMetadata     = True,
                 dstNodata        = 0)      # Select the nodata value
clip = None

• Use the image extent (bounds)

clip =  gdal.Warp(destNameOrDestDS = outputpath, 
                  srcDSOrSrcDSTab  = inputpath,
                  outputBounds     = (minx, miny, maxx, maxy),
                  cropToCutline    = True,
                  copyMetadata     = True)

clip = None