Skip to Main Content
Mobile Menu

Image Pre-Processing Routines For Gamera

Masters Thesis, Boise State University, December 2007

By Uma Kompella

Abstract:

A set of image pre-processing algorithms is implemented within the Gamera framework. The implementation, which is a significant addition, is carried out by extending the functionality of Gamera, an existing open-source framework for conducting research and building applications for document analysis. Research in the field of document processing can benefit from a common software library that can be used by academic, business, and government domains. The implemented and integrated algorithms fall in different categories of image pre-processing namely global and local binarization, noise reduction, skew estimation, and segmentation. This project integrates some of the proposed algorithms into Gamera as plugins and some as part of a toolkit. The plugins are integrated into Gamera’s build and are distributed with Gamera’s source tree. A toolkit is a way to distribute code, based on Python’s generic package that uses Gamera but is not included in the Gamera source tree.