Automatic Text Extraction from Arabic Newspapers

Abstract

A system for extracting the textual information from document images with complex layouts is presented. It is based on both layout analysis and text localization techniques. Layout analysis is first applied to segment the page in text and non-text blocks and then text localization is used to detect text that may be embedded inside images, charts, diagrams, tables etc. Detailed experiments on scanned Arabic newspapers showed that combining layout analysis and text localization methods can lead to improved page segmentation and text extraction results.