109

urn:lsid:arphahub.com:pub:3dc5f44e-8666-58db-bc76-a455210e8891

JUCS - Journal of Universal Computer Science

jucs

0948-695X 0948-6968

Journal of Universal Computer Science

10.3217/jucs-014-02-0266

28944

Research Article

H.3.3 - Information Search and Retrieval

Detailing a Quantitative Method for Assessing Algorithms to Remove Back-to-Front Interference in Documents

Lins

Rafael Dueire

rdl@ufpe.br 1 João Marcelo Monte Da Silva

2 Martins

Fernando Mário Junqueira

Federal University of Pernambuco, Recife, Brazil

Federal University of Pernambuco

Recife

Brazil 2

Universidade Federal de Pernambuco, Recife, Brazil

Universidade Federal de Pernambuco

Recife

Brazil 3

Universidade do Minho, Braga, Portugal

Universidade do Minho

Braga

Portugal

Corresponding author: Rafael Dueire Lins (rdl@ufpe.br).

Academic editor:

2008

28 01 2008

14 2 266 283 B0C152BC-A132-5537-BEFA-1056C6731CC1 7000089

Rafael Dueire Lins, João Marcelo Monte Da Silva, Fernando Mário Junqueira Martins

This article is freely available under the J.UCS Open Content License.

Abstract

Documents written on both sides on translucent paper make visible the ink from one side on the other. This artefact is called "back-to-front interference", "bleeding" or "show-through". The direct binarization of documents with such interference yields unreadable documents. The literature presents several algorithms for suitably removing such artefact. This paper presents a quantitative method to assess algorithms to remove back-to-front interference.