The Emille project (http://www.emille.lancs.ac.uk/about.php) produced a POS tagger and a tagged corpus of Urdu. There is also a page off the ParGram project (http://www2.parc.com/isl/groups/nltt/pargram/) for Urdu, but at the moment it seems to be broken.
As for Persian, Karine Megerdoomian has a web page (http://www.zoorna.org/publications.html) listing a paper of hers entitled "Developing a Persian part-of-speech tagger." --
Mike Maxwell
What good is a universe without somebody around to look at it?
--Robert Dicke, Princeton physicist