Search
Publication Authors

Prof. Dr. Didier Stricker

Dr. Alain Pagani

Dr. Gerd Reis

Eric Thil

Keonna Cunningham

Dr. Oliver Wasenmüller

Dr. Muhammad Zeshan Afzal

Dr. Gabriele Bleser

Dr. Muhammad Jameel Nawaz Malik
Dr. Bruno Mirbach

Dr. Jason Raphael Rambach

Dr. Nadia Robertini

Dr. René Schuster

Dr. Bertram Taetz

Ahmed Aboukhadra

Sk Aziz Ali

Mhd Rashed Al Koutayni

Yuriy Anisimov

Jilliam Maria Diaz Barros

Ramy Battrawy
Hammad Butt

Mahdi Chamseddine

Steve Dias da Cruz
Fangwen Shu

Torben Fetzer

Ahmet Firintepe
Sophie Folawiyo

David Michael Fürst

Christiano Couto Gava

Tewodros Amberbir Habtegebrial
Simon Häring

Khurram Azeem Hashmi
Henri Hoyez

Jigyasa Singh Katrolia

Andreas Kölsch
Onorina Kovalenko

Stephan Krauß
Paul Lesur

Michael Lorenz

Dr. Markus Miezal

Mina Ameli

Nareg Minaskan Karabid

Mohammad Minouei

Pramod Murthy

Mathias Musahl

Peter Neigel

Manthan Pancholi
Mariia Podguzova

Praveen Nathan
Qinzhuan Qian
Rishav
Marcel Rogge
María Alejandra Sánchez Marín
Dr. Kripasindhu Sarkar

Alexander Schäfer

Pascal Schneider

Mohamed Selim

Tahira Shehzadi
Lukas Stefan Staecker

Yongzhi Su

Xiaoying Tan
Christian Witte

Yaxu Xie

Vemburaj Yadav

Dr. Vladislav Golyanik

Dr. Aditya Tewari

André Luiz Brandão
Publication Archive
New title
- ActivityPlus
- AlterEgo
- AR-Handbook
- ARVIDA
- Auroras
- AVILUSplus
- Be-greifen
- Body Analyzer
- CAPTURE
- COGNITO
- DAKARA
- Density
- DYNAMICS
- EASY-IMP
- Eyes Of Things
- iACT
- IMCVO
- IVMT
- LARA
- LiSA
- Marmorbild
- Micro-Dress
- Odysseus Studio
- On Eye
- OrcaM
- PAMAP
- PROWILAN
- ServiceFactory
- STREET3D
- SUDPLAN
- SwarmTrack
- TuBUs-Pro
- VIDETE
- VIDP
- VisIMon
- VISTRA
- You in 3D
Binarization-free OCR for historical documents using LSTM networks
Binarization-free OCR for historical documents using LSTM networks
Mohammad Reza Yousefi, Mohammad Reza Soheili, Thomas Breuel, Ehsanollah Kabir, Didier Stricker
International Conference on Document Analysis and Recognition (ICDAR-13), August 23-26, Nancy, France
- Abstract:
- A primary preprocessing block of almost any typical OCR system is binarization, through which it is intended to remove unwanted part of the input image, and only keep a binarized and cleaned-up version for further processing. The binarization step does not, however, always perform perfectly, and it can happen that binarization artifacts result in important information loss, by for instance breaking or deforming character shapes. In historical documents, due to a more dominant presence of noise and other sources of degradations, the performance of binarization methods usually deteriorates; as a result the performance of the recognition pipeline is hindered by such preprocessing phases. In this paper, we propose to skip the binarization step by directly training a 1D Long Short Term Memory (LSTM) network on gray-level text lines. We collect a large set of historical Fraktur documents, from publicly available online sources, and form train and test sets for performing experiments on both gray-level and binarized text lines. In order to observe the impact of resolution, the experiments are carried out on two identical sets of low and high resolutions. Overall, using gray-level text lines, the 1D LSTM network can reach 25% and 12.5% lower error rates on the low- and high-resolution sets, respectively, compared to the case of using binarization in the recognition pipeline.