Dashboard link: https://acycliq.github.io/issplus/
Background image taken from:
\\basket.cortexlab.net\data\kenneth\iss\170315_161220KI_4-3\Output\background_boundaries.tif
-
Upscale the image so that the longest side is 65536px wide. That allows you to zoom 8 levels deep. (You will need these two numbers later on)
-
Now you can break-up the image into tiles. I used gdal2tiles for this, but its a bit tricky to install its dependencies. It is easier to install OSGeo4W which has gdal2tiles as a package (There is a post on the web for this, I think it was this one https://alastaira.wordpress.com/2011/07/11/maptiler-gdal2tiles-and-raster-resampling/)
-
gdal2tiles however will give tiles with different orientation (i think its called tms??). Download gdal2tilesG.py (from https://gist.github.com/jeffaudi/9da77abf254301652baa) and put it in the same folder as gdal2tiles.py that came with OSGeo4W.
-
Open an OSGeo4W shell
-
gdal2tiles needs explicit unit8, therefore you have to scale the bit depth. Do this with the following: (otherwise the tiles look fuzzy, blurry, messed-up). Run that in the OSGeo4W shell:
gdal_translate -ot Byte -scale "imageFilename_in.tif" out.tif -
You can now do the tiles. Run that in the OSGeo4W shell:
python C:\OSGeo4W64\bin\gdal2tilesG.py -p raster -z 0-8 -w none out.tif- You need to target gdal2tilesG.py not gdal2tiles.py
- switch
-z 0-8sets the zoom levels. 8 is the maximun and has to be the correct one, ie the one that corresponds to your image dimensions. In this case image is 65536px wide which allows up to 8 zoom levels. If you put something else instead of 8 the tiles are going to be wrong.
"# spacetx"