Re: [INDOLOGY] Aṅgavijjā

Matthew Kapstein mkapstei at uchicago.edu
Mon Sep 4 07:53:11 UTC 2017


Dear David,

I believe that you are referring to the SanskritOCR developed by Oliver Hellwig:
http://www.indsenz.com/int/index.php

I have used this and it is quite accurate when the source -- a scanned pdf of a devanagari text -- is
neat and clean.  But even when it is "quite accurate" the results must be checked manually and, in
the case of large texts, this is of course tedious.

Hellwig's work has been remarkable, however -- it was not many years ago that I was told by an individual
prominently involved in the digitalization of South Asia texts that devanagari OCR was well nigh impossible
owing to the manner in which the characters are connected by a continuous upper line.

best,
Matthew

Matthew Kapstein
Directeur d'études,
Ecole Pratique des Hautes Etudes

Numata Visiting Professor of Buddhist Studies,
The University of Chicago
________________________________



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://list.indology.info/pipermail/indology/attachments/20170904/4de25981/attachment.htm>


More information about the INDOLOGY mailing list