Guide to OCR for Arabic Scripts (eBook)

Volker Märgner is Academic Director of the Institute for Communications Technology (IfN) at Technische Universität Braunschweig, Germany. He has over 30 years research experience in image processing, pattern recognition, and handwriting recognition. He developed the IfN/ENIT-database of Arabic handwritten names which is the reference for Arabic handwritten word recognition systems and organized competitions both together with Haikal El Abed.Haikal El Abed is a Senior Research Engineer at the Institute for Communications Technology (IfN) at Technische Universität Braunschweig, Germany. He has more than 10 years research experience in pattern recognition and Arabic text recognition, on-line and off-line. He organizes competitions and works on the collection of databases.

Part I: Pre-ProcessingAn Assessment of Arabic Handwriting Recognition TechnologySargur N. Srihari and Gregory BallLayout Analysis of Arabic Script DocumentsSyed Saqib Bukhari, Faisal Shafait and Thomas M. BreuelA Multi-Stage Approach to Arabic Document AnalysisEugene Borovikov and Ilya ZavorinPre-Processing Issues in Arabic OCRZhixin Shi, Srirangaraj Setlur and Venu GovindarajuSegmentation of Ancient Arabic DocumentsAbdel Belaïd and Nazih OuwayedFeatures for HMM-Based Arabic Handwritten Word Recognition SystemsLaurence Likforman-Sulem, Ramy Al Hajj Mohammad, Chafic Mokbel, Fares Menasri, Anne-Laure Bianne-Bernard and Christopher KermorvantPart II: RecognitionPrinted Arabic Text RecognitionIrfan Ahmed, Sabri A. Mahmoud and Mohammed Tanvir ParvezHandwritten Arabic Word Recognition Using the IFN/ENIT-DatabaseMario Pechwitz, Haikal El Abed and Volker MärgnerRWTH OCR: A Large Vocabulary Optical Character Recognition System for Arabic ScriptsPhilippe Dreuw, David Rybach, Georg Heigold and Hermann NeyArabic Handwriting Recognition using Bernoulli HMMsIhab Alkhoury, Adrià Giménez and Alfons JuanHandwritten Farsi Words Recognition Using Hidden Markov ModelsPuntis Jifroodian and Ching Y. SuenOffline Arabic Handwriting Recognition with Multidimensional Recurrent Neural NetworksAlex GravesApplication of Fractal Theory in Farsi/Arabic Document AnalysisSaeed MozaffariMulti-Stream Markov Models for Arabic Handwriting RecognitionYousri Kessentini, Thierry Paquet and AbdelMajid Ben HamadouTowards Distributed Cursive Writing OCR Systems based on the Combination of Complementary ApproachesMaher Khemakhem and Abdelfettah BelghithPart III: EvaluationData Collection and Annotation for Arabic Document AnalysisIlya Zavorin and Eugene BorovikovArabic Handwriting Recognition CompetitionsVolker Märgner and Haikal El AbedBenchmarking Strategy for Arabic Screen Rendered Word RecognitionFouad Slimane, Slim Kanoun, Jean Hennebert, Rolf Ingold, Adel M. Alimi and Jean HennebertPart IV: ApplicationsA Robust Word Spotting System for Historical Arabic ManuscriptsMohamed Cheriet and Reza Farrahi MoghaddamArabic Text recognition using a Script-Independent Methodology: A Unified HMM-based Approach for Machine-print and Handwritten TextPremkumar Natarajan, Rohit Prasad, Huaigu Cao, Krishna Subramanian, Shirin Saleem, David Belanger, Shiv Vitaladevuni, Matin Kamali and Ehry MacRostieArabic Handwriting Recognition Using VDHMM and Over-SegmentationAmlan Kundu and Tom HinesOnline Arabic Databases and ApplicationsHoucine Boubaker, Abdelkarim Elbaati, Najiba Tagougui, Haikal El Abed, Monji Kherallah and Adel M. AlimiOnline Arabic Handwritten Words Recognition Based on HMM and Combination of Online and Offline FeaturesSherif Abdelazeem, Hesham M. Eraqi and Hany Ahmed