International Science Community Association Research Journal of Recent Sciences 2277 - 2502 3 11 2014 November 2 Offline Urdu Numeral Recognition Using Non-Negative Matrix Factorization 98 102 EN Amanati Sahar 1 Business Management Department, Allameh Tabataba'i University, Tehran, IRAN yazdandoust Majid 1 Department of Civil Eng., Tarbiat Modares University, Tehran- 14115-111, IRAN panah Alikomak 2 Department of Civil Engineering, Parsian University, Qazvin- 14115-111, IRAN Reza Sadeghzadegan 3 Khorasani Homa 1 Faculty of Psychology, Roudehen Branch, Islamic Azad University, Roudehen, IRAN Hayedeh Saberi 2 Bagheri Nasrin 3 Beigloo Leila 1 East Azarbaijan Science and Research Branch, Islamic Azad University,Tabriz, IRAN Bohloli Nader 2 Department of Management, Bonab Branch, Islamic Azad University, Bonab, IRAN Vahedi Majid 3 Department of Management, Payam Noor University, PO BOX 19395-4697 Tehran, IRAN Abid Muhammad 1 GIFT University Gujranwala, PAKISTAN Rehman MohsinAbdul 2 Sajjad HassanyPazoky 1 Department of Geomatics, Faculty of Engineering, University of Tehran, Tehran, IRAN Mohammad Saadatseresht 2 Chehreghan Alireza 3 Michalovic Roman 1 Faculty of Special Engineering, 897, Rosina, SLOVAKIA Mikomix 2 Družinská 3 M.N. Hafiza 1 School of Fundamental Science, Universiti Malaysia Terengganu, 21030 Kuala Terengganu,Terengganu, MALAYSIA Isa M.I.N. 2 Center of Corporate Communication and Image Development, Universiti Malaysia Terengganu, 21030 Kuala Terengganu,Terengganu, MALAYSIA N.V. Biradar 1 Department of Zoology, Karnatak Science College, Dharwad, Karnataka, INDIA Sindagi AmbarishS. 2 Department of Microbiology, Karnatak Science College, Dharwad, Karnataka, INDIA Bellad A.S 3 Department of Botany, St. Joseph’s Post Graduate and Research Centre, Bangalore, Karnataka, INDIA Reddy Jayarama 4 Department of life sciences, Jain University, Bangalore, Karnataka, INDIA Navalur Ravi 5 Department of Microbiology, Karnatak University, Dharwad, Karnataka, INDIA Naykar Shivaraj 6 Department of Microbiology, University of Gondar, ETHIOPIA MathewsP. Raj 7 Sadashiv S.O 8 Chandrashekhar Unakal 9 Khadijeh KhodabakhshiParijan 1 DOS in B.N. Bahadur Institute of Management Sciences, Manasagangotri, University of Mysore, Mysore, Karnataka, INDIA Naderian Arash 2 Department of Accounting, Aliabad Katoul Branch, Islamic Azad University, Aliabad Katoul, IRAN Hayati Manouchehr 1 Islamic Azad University, UAE Branch, IRAN Vahedi Elham 2 Department of Executive Master of Business Administration, Alborz, IRAN Tavani SaeedNajafi 3 Department of Management, Institute of Management and Planning Studies, IRAN Department of Management and Global Economy, Instroct Scientific Applied Training Center, Tehran, IRAN Shirian Abbas 4 Department of Management, Alborz, IRAN Ebrahimi MohammadReza 5 Siji S. 1 Department of Chemistry, Sree Narayana College, Kannur-670007, Kerala, INDIA Dhanitha M.A. 2 Janardanan C. 3 Jana Müllerová 1 Department of Fire Engineering, Faculty of Special Engineering, University of Zilina, 1. mája 32, 010 08 Žilina, SLOVAKIA Talebi P 1 English Department, Mysore University, INDIA Memarian Arman 1 Islamic Azad University, Zanjan Branch, IRAN Ghader Bagheri 2 Uddin Shahab 1 Department of Computer Science, Federal Urdu University of Arts, Sciences and Technology, Karachi, PAKISTAN Sarim Muhammad 2 Shaikh AbdulBasit 3 Raffat SheikhKashif 4 > Research Article 2013 12 1 2014 November 2 By the rapid change and advancement in technology a need for processing and preserving many texts had been felt. These texts are either in hard copies or in handwritten form. Hand-written numerals, written in various languages and scripts, are an integral part of these texts. Several efforts have been made to recognize numerals and a variety of Optical Character Recognition (OCR) systems have been successfully implemented and marketed. Urdu numerals, as opposed to English numerals, are different due to their style and format of writing. Various methods have been proposed but majority of them only address computer typed numerals in different forms and sizes. Therefore we need to develop new and enhance existing handwritten Urdu numerals recognition systems due to their wide scale use and application in many fields. This research addresses the problem of handwritten offline numerals. A novel approach of Non-negative Matrix Factorization (NMF) for Urdu handwritten character recognition has been proposed in this research. Copyright@ International Science Community Association