Introduction to Document Analysis and Recognition

Document Analysis and Recognition (DAR) aims at the automatic extraction of information presented on paper and initially addressed to human comprehension. The desired output of DAR systems is usually in a suitable symbolic representation that can subsequently be processed by computers. Over the centuries, paper documents have been the principal instrument to make permanent the progress of the humankind. Nowadays, most information is still recorded, stored, and distributed in paper format. The widespread use of computers for document editing, with the introduction of PCs and wordprocessors in the late 1980’s, had the effect of increasing, instead of reducing, the amount of information held on paper. Even if current technological trends seem to move towards a paperless world, some studies demonstrated that the use of paper as a media for information exchange is still increasing [1]. Moreover, there are still application domains where the paper persists to be the preferred media [2]. The most widely known applications of DAR are related to the processing of office documents (such as invoices, bank documents, business letters, and checks) and to the automatic mail sorting. With the current availability of inexpensive high-resolution scanning devices, combined with powerful computers, state-of-the-art OCR packages can solve simple recognition tasks for most users. Recent research directions are widening the use of the DAR techniques, significant examples are the processing of ancient/historical documents in digital libraries, the information extraction from “digital born” documents, such as PDF and HTML, and the analysis of natural images (acquired with mobile phones and digital cameras) containing textual information. The development of a DAR system requires the integration of several competences in computer science, among the others: image processing, pattern recognition, natural language processing, artificial intelligence, and database systems. DAR applications are particularly suitable for the incorporation of