DL Architecture for Indic Scripts

Suryaprakash Kompalli, Srirangaraj Setlur, Venugopal Govindaraju

Abstract: In this study, we outline computational issues in the design of a Digital Library (DL) for Indic languages. The complicated character structure of Indic scripts entails novel OCR analysis techniques and user interface (UI) designs. This paper describes a multi-tier software architecture, which provides text and image processing tools as independent, reusable entities. Techniques for measuring and evaluating different stages of an Indic script recognition engine are outlined.