An OCR system for Greek printed early books based on computational geometry algorithms

Abstract

In this paper we propose a novel OCR system for Greek printed early books, combining image preprocessing with computational geometry technologies. Our aim was to carry out OCR digitalisation of a large collection of digitised, Greek early printed books dated from the late 15th century to the mid-18th century. The proposed method is based on: (i) image preprocessing using image binarisation and enhancement; (ii) the creation of a convex polygon that represents the feature extraction of each fount; and (iii) training and identification procedures based on algorithms of intersecting convex polygons. The major advantage of this method is that it can control the authentication of a published document image or its partial modification to a reliable degree. In this way the proposed system decision for the classification of a candidate letter is based on smart geometric practice. Experimental results have proved the efficiency of the proposed approach. © 2010 IADIS.