gdal2mbtiles helps you generate web mapping tiles that can be shown through a browser-based mapping library on your website.

GDAL-readable files are images that are georeference, that means that they are positioned and projected on to the world. In order to display a dynamic map on the web, you don't want to serve the whole image at once, so it must be sliced into tiles that are hosted by a tile server.

The MBTiles file format was developed by MapBox to make tile storage easier. You can upload the final file to their service, or run your own tile server. MapBox provides one called TileStream.

You can get a copy of the source by using:

$ git clone https://github.com/ecometrica/gdal2mbtiles.git

Note that this program requires Python 2.7 or higher.

We rely on GDAL to read georeferenced datasets. However, it is not available on PyPi.

Under Debian or Ubuntu, run the following to install it:

$ sudo apt-get install python-gdal

We also rely on VIPS to do fast image processing. It's also not available on PyPi.

Under Debian or Ubuntu, run the following to install it:

$ sudo apt-get install python-vipscc

If you are using a virtualenv, you will need to symlink Python library in the right place. Under Debian or Ubuntu, assuming Python 2.7, run the following:

$ ln -s /usr/lib/python2.7/dist-packages/vipsCC $VIRTUAL_ENV/lib/python2.7/site-packages/

You'll also need a few other libraries to deal with large TIFF files and to optimize the resulting PNG tiles.

Under Debian or Ubuntu, run the following to install them:

$ sudo apt-get install libtiff5 optipng pngquant

Please check our issue tracker for known bugs and feature requests.

We accept pull requests for fixes and new features.

Maxime Dupuis and Simon Law wrote this program, with the generous support of Ecometrica.