Handwriting: Recognition, Development and Analysis


Byron Leite Dantas Bezerra, Cleber Zanchettin, Alejandro H. Toselli and Giuseppe Pirlo (Editors)
Department of Computer Engineering , University of Pernambuco, Recife, Brazil

Series: Computer Science, Technology and Applications
BISAC: COM014000

This book has the primary goal of presenting and discussing some recent advances and ongoing developments in the Handwritten Text Recognition (HTR) field, resulting from works done on different HTR-related topics for the achievement of more accurate and efficient recognition systems. Nowadays, there is an enormous worldwide interest in HTR systems, which is mostly driven by the emergence of new portable devices incorporating handwriting recognition functions. Others interests are the biometric identification systems employing handwritten signatures, as well as the requirements from cultural heritage institutions like historical archives and libraries in order to preserve their large collections of historical (handwritten) documents. The book is organized into two sections: the first one is mainly devoted to describing the current state-of-the-art applications in HTR and the last advances in some of the steps involved in HTR workflow (that is, preprocessing, feature extraction, recognition engines, etc.), whereas the second focuses more on some relevant HTR-related applications.

In more depth, the first part offers an overview of the current state-of-the-art applications of HTR technology and introduces the new challenges and research opportunities in the field. Besides, it provides a general discussion of currently ongoing approaches towards solving the underlying search problems on the basis of existing methods for HTR in terms of both accuracy and efficiency. In particular, there are chapters especially focused on image thresholding and enhancement, text image preprocessing techniques for historical handwritten documents and feature extraction methods for HTR. Likewise, in line with the breakout success of Deep Neural Networks (DNNs) in the field, a whole chapter is devoted to describing the designing of HTR systems based on DNNs. Finally, a chapter listing the most used benchmarking datasets for HTR is also included, providing detailed information about which types of HTR systems (on/offline) and features are commonly considered for each of them.

In the second part, several systems — also developed on the basis of the fundamental concepts and general approaches outlined in the first part — are described for several HTR-related applications. Presented in the corresponding chapters, these applications cover a wide spectrum of scenarios: mathematical formulae recognition, scripting language recognition, multimodal handwriting-speech recognition, hardware design for online HTR, student performance evaluation through handwriting analysis, performance evaluation methods, keyword spotting, and handwritten signature verification systems.

Last but not least, it is important to remark that to a large extent, this book is the result of works carried out by several researchers in the Handwritten Text Recognition field.
Therefore, it owes credit to these researchers that have directly contributed to their ideas, discussions and technical collaborations, and in general who, in one manner or another, have made it possible. (Imprint: Nova)



Table of Contents

PART I. Recognition and Development

Chapter 1. Handwriting Recognition: Overview, Challenges and Future Trends
Everton Barbosa Lacerda, Thiago Vinicius Machado de Souza, Cleber Zanchettin, Juliano Cícero Bitu Rabelo and Lara Dantas Coutinho (Federal University of Pernambuco, Recife, Brazil, and others)

Chapter 2. Thresholding
Edward Roe and Carlos Alexandre Barros de Mello (CESAR – Center for Advanced Studies of Recife, Recife, Brazil, and others)

Chapter 3. Historical Document Processing
Basilis Gatos, Georgios Louloudis, Nikolaos Stamatopoulos and Giorgos Sfikas (Computational Intelligence Laboratory, Institute of Informatics and Telecommunications, National Center for Scientific Research “Demokritos,” Agia Paraskevi, Greece)

Chapter 4. Wavelet Descriptors for Handwritten Text Recognition in Historical Documents
Leticia M. Seijas and Byron L. D. Bezerra (Department of Computing, Faculty of Exact and Natural Sciences, University of Buenos Aires, Buenos Aires, Argentina, and others)

Chapter 5. How to Design Deep Neural Networks for Handwriting Recognition
Théodore Bluche, Christopher Kermorvant and Hermann Ney (A2iA SAS, Paris, France, and others)

Chapter 6. Handwritten and Printed Image Datasets: A Review and Proposals for Automatic Building
Gearlles V. Ferreira, Felipe M. Gouveia, Byron L. D. Bezerra, Eduardo Muller, Cleber Zanchettin, and Alejandro Toselli (E-Comp, University of Pernambuco, Recife, Brazil, and others)

PART II. Analysis and Applications

Chapter 7. Mathematical Expression Recognition
Francisco Álvaro, Joan Andreu Sánchez and José Miguel Benedí (Pattern Recognition and Human Language Technologies Research Center, Polytechnic University of València, València, Spain, and others)

Chapter 8. Online Handwriting Recognition of Indian Scripts
Umapada Pal and Nilanjana Bhattacharya (Computer Vision and Pattern Recognition Unit, Indian Statistical Institute, Kolkata, India, and others)

Chapter 9. Historical Handwritten Document Analysis of Southeast Asian Palm Leaf Manuscripts
Made Windu Antara Kesiman, Jean-Christophe Burie, Jean-Marc Ogier, Gusti N. M. A. Wibawantara, and I Made G. Sunarya (Computer Science Image Interaction (L3i), University of La Rochelle, La Rochelle, France, and others)

Chapter 10. Using Speech and Handwriting in an Interactive Approach for Transcribing Historical Documents
Emilio Granell, Verónica Romero, and Carlos-D. Martínez-Hinarejos (PRHLT Research Center, Polytechnic University of València, Valencia – Spain)

Chapter 11. Handwritten Keyword Spotting the Query by Example (QbE) Case
Georgios Barlas, Konstantinos Zagoris and Ioannis Pratikakis (Department of Electrical and Computer Engineering, Democritus University of Thrace, Xanthi, Greece)

Chapter 12. Handwriting-Enabled E-Paper Based on Twisting-Ball Display
Yusuke Komazaki and Toru Torii (Graduate School of Frontair Sciences, The University of Tokyo, Tokyo, Japan)

Chapter 13. Speed and Legibility: Brazilian Students Performance in a Thematic Writing Task
Monique Herrera Cardoso and Simone Aparecida Capellini (Investigation Learning Disabilities Lab at the Speech and Hearing Sciences Department, São Paulo State University “Júlio de Mesquita Filho” – UNESP, São Paulo, Brazil)

Chapter 14. Datasets for Handwritten Signature Verification: A Survey and a New Dataset, the RPPDI-SigData
Victor Kléber Santos Leite Melo, Byron Leite Dantas Bezerra, Rebecca H. S. N. Do Nascimento, Gabriel Calazans Duarte de Moura, Giovanni L. L. de S. Martins, Giuseppe Pirlo and Donato (Polytechnic School, University of Pernambuco, Brazil, and others)

Chapter 15. Processing of Handwritten Online Signatures: An Overview and Future Trends
Alessandro Balestrucci, Donato Impedovo and Giuseppe Pirlo (Department of Computer Science, University of Bari, Bari, Italy)

Editor’s Contact Information


Additional Information

Students, professors, researchers, and engineers interested or working on the following topics:
• Handwriting Recognition Techniques
• Cursive Script Recognition
• Symbol, Equation, Sketch and Drawing Recognition
• Handwritten Document Processing and Understanding
• Computational Linguistics  and Information Integration
• Handwritten Databases and Digital Libraries
• Document Characterization
• Form Processing
• Word spotting
• Bank-Check Processing
• Historical Document Processing
• Writer Verification and Identification
• Performance Enhancement and System Evaluation

Additional information