libtholoura is a text-image distortion library, useful for sites that want to ensure that a human (and not a bot) is viewing this page.
It is optimized to produce images of strings distorted in a way that made impossible for an OCR to read, but humans do.
The library takes as input any (ASCII preferably) string and is able to dynamically produce PNG indexed images (with 256 colors, interlaced) with the distorted image. In order to check this library a test program is provided.
Another feature of this library is that the distortion filters are actually plugins (i.e. dynamic libraries itself). It is easy to add or remove plugins on demand without the need to recompile the library itself.
As an extra value, it is possible to (easily) create new plugins since the developer does not need to deal with things like "how to produce an PNG" or "how to render text"; the only thing is needed is to be able to deal with a fixed size table (and a structure defining this table).
Installation:
Installation should be fairly easy. First become root (if you want to compile it somewhere system wide - check the configuration section above only if you want to change this default behavior).
Then type "make install", make sure that the paths are correct and press [RETURN] to start the installation.
Product's homepage
Find us on Google+
