David Doermann, Rajat Ahuja

A vast numbers of raster maps are available today without equivalent vector forms. Most existing GIS suites support vectorization of raster maps, but not automatic text-extraction. The objective of this project was to create a text extraction system for raster-scanned maps with an intuitive Graphical User Interface, supporting operator assistance for learning, performance evaluation and parameter tuning. Applications of such a system lie in the indexing and subsequent retrieval of map images based on query strings made-up of extracted text, and plugging-in text-extraction support to existing GIS suites.

Rajat Ahuja: rahuja@umiacs.umd.edu

(View All Project Publications)