I need to transform a large number of Esri ASCII grid files to polygonized WKT format, e.g. as CSV. Currently, I use GeoTools with the ArcGrid reader and a polygon extraction process. For parallelism this function is executed via Spark. Still converting 30-40 GB of the raster data is very slow. Do you see any possibilities to improve performance?

Would you recommend GDAL / OGR over GeoTools for this purpose? Is there another library which is faster?


Also as these are many small files I want to point out that it is related to https://stackoverflow.com/questions/44322914/spark-read-zip-file-which-contain-multiple-files-with-cpu-intensive-parsing where I try to parse the zip files in parallel.

1 Answer 1


Geotools includes code in the unsupported process raster module.

The tricky thing is to concatenate the right configuration files as otherwise some modules are not registered and thus not available at runtime.

Then a polygon extraction process serves fine.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.