I have a selection of vintage aerial imagery from 1929 which features variable brightness, with edges of images generally overly dark. While I'm grateful that imagery from 1929 exists at all for my study area, this is annoying and detrimental to classification efforts. As most of the images overlap each other around their edges, I'd like to blend the pixels from area where the raster overlap in some way, ideally such that there are no sharp changes visible in transition from one exposure to another in the patched map.
I can patch the images together with r.patch, but then I get the artifacts of the edges. I also thought using r.mapcalc to blend pixels pixels 50/50 where maps overlap. But surely this has been done many times before and I can avoid inventing my own method. Do there exist any specialized tools for smart stitching of aerial imagery tiles together such that overlapping area a blended; or could someone suggest a good technique? -- Henry House House Agricultural Consultants _______________________________________________ grass-user mailing list grass-user@lists.osgeo.org https://lists.osgeo.org/mailman/listinfo/grass-user