DOCSLIB.ORG

Deep Convolutional Neural Network for Recognition of Unified Multi-Language Handwritten Numerals

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Deep Convolutional Neural Network for Recognition of Unified Multi-Language Handwritten Numerals 2018 IEEE 2nd International Workshop on Arabic and Derived Script Analysis and Recognition (ASAR) Deep Convolutional Neural Network for Recognition of Unified Multi-Language Handwritten Numerals Jaafar Alghazo Ghazanfar Latif Loay Alzubaidi Computer Engineering Department, Computer Science Department, Computer Science Department, Prince Mohammad Bin Fahd University, Prince Mohammad Bin Fahd University, Prince Mohammad Bin Fahd University, Al Khobar, Saudi Arabia. Al Khobar, Saudi Arabia. Al Khobar, Saudi Arabia. [email protected] [email protected] [email protected] M. Muzzamal Naseer Yazan Alghazo College of Engineering, Humanities Department, Prince Australian National University, Mohammad Bin Fahd University, Al Canberra, Australia. Khobar, Saudi Arabia. [email protected] [email protected] Abstract— Deep learning systems have recently gained handwritten. The classification phase involves using well- importance as the architecture of choice in artificial intelligence developed classification techniques such as Multilayer (AI). Handwritten numeral recognition is essential for the Perceptron (MLP) and Support Vector Machines (SVM) among development of systems that can accurately recognize digits in many others. This phase involves training the Artificial different languages which is a challenging task due to variant Intelligence (AI) system and then recognizing the remaining writing styles. This is still an open area of research for developing portion of the dataset or new data. These algorithms can also an optimized Multilanguage writer independent technique for utilize either supervised or unsupervised learning while training numerals. In this paper, we propose a deep learning architecture the AI system on the dataset. It is observed that mostly research for the recognition of handwritten Multilanguage (mixed papers have only tackled numerals written in one language numerals belongs to multiple languages) numerals (Eastern Arabic, Persian, Devanagari, Urdu, Western Arabic). The overall The importance of recognizing handwritten numerals is accuracy of the combined Multilanguage database was 99.26% apparent due to the increase usage of handheld devices in both with a precision of 99.29% on average. The average accuracy of personal and professional settings. Banking industry relies each individual language was found to be 99.322%. Results heavily on handwritten numeral recognition especially for indicate that the proposed deep learning architecture produces numerals written on checks. Other industries also would benefit better results compared to methods suggested in the previous from handwritten numeral recognition algorithms in addition to literature. individuals with different handheld device applications developed that allow users to write numbers rather than input Keywords—Arabic Numberals; Mul-Language Numerals Recognition; Hand Written Numerals; Deep Convolutional Neural them using a keyboard. Networks In this research, a deep Convolutional Neural Network (CNN) architecture is developed for handwritten Eastern Arabic I. INTRODUCTION and Persian numerals while experiments are also performed on some other languages including Urdu, Devanagari, Western Due to the infinite number of variations of handwritten Arabic which in total, are spoken by approximately 1.86 billion numerals that are subject to many variable, the recognition of people. The rest of this paper is structured as follows: section 2 handwritten numeral is a complex problem that has been subject confers the literature review, section 3 presents the proposed of research. A large number of algorithms have been developed system, section 4 details the experimental datasets, section 5 to tackle this problem for both offline and online recognition. discusses the results of the proposed recognition system and Offline recognition refers to recognizing numerals written on section 6 concludes the paper. paper using a pen and online recognition refers to numerals written on handheld devices e.g. stylus. Handwritten numerals can differ even for the same writer when writing on paper or a II. LITERATURE REVIEW handheld device and can vary even in the same medium. The Though the concept of Deep Learning has been around for algorithm consisting of pre-processing, feature extraction and quite some time, yet the utilization of Deep learning to its full the final phase of classification is used for both cases. In the potential was not achieved until recently with advances in preprocessing phase, the image is extracted whether in the technology, computing power and computing resources. Deep spatio-temporal or spatio-luminance representation, noise learning algorithms have achieved new records in different reduction, centering the image, etc. In the feature extraction fields such as image recognition and speech recognition among phase, researchers proposed different methods for feature others [1-2]. The concept of Deep Learning relies on what is extraction that can be used for in final step of classification of known as Convolutional Neural Networks (ConvNets or CNN). XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE 978-1-5386-1459-4/18/$31.00 ©2018 IEEE 90 Authorized licensed use limited to: Prince Mohammad Bin Fahd University. Downloaded on August 13,2020 at 11:49:58 UTC from IEEE Xplore. Restrictions apply. 2018 IEEE 2nd International Workshop on Arabic and Derived Script Analysis and Recognition (ASAR) LeNet 5 developed by LeCun, Y. et al. was one of the first are used in [20] for improving handwritten Chinese text convolutional networks used in the 1990s [3]. In 2012, recognition. Deep Neural Network is used as part of an Krizhevsky, A. et al. [4] introduced a newer CNN that was wider algorithm implemented in [21] for SAR target configuration and deeper than LeNet and subsequently won the ImageNet- recognition. Fisher Vector (FV) and CNN are used in [22] for 2012 Large Scale Visual Recognition Challenge (ILSVRC). visual recognition of the ancient inscriptions. This is considered by many as the initial point of the subsequent Since Deep learning algorithm handwritten numeral interest in Deep Learning. The accuracy rate achieved was recognition rate will be compared to widely studied Artificial 84.7%. The Architecture of [4] was modified into a new Intelligence Algorithms, a few references will be mentioned architecture proposed by M. D. Zeiler and R. Fergus in [5]. This here in the literature review. Different novel feature extraction new architecture ended up winning the ILSVRC 2013. It methods were developed and used in [23-25] for numeral and achieved an accuracy rate of 88.3%. An Inception module was character recognition. Different classifier have also been used to developed that shrinks the number of parameters in the network test the proposed algorithms. and was the contribution of the CNN developed by Szegedy, C. et al. and ended up winning the ILSVRC 2014 [6]. The accuracy rate achieved was 93.33%. In the same year and in the same III. PROPOSED SYSTEM competition Simonyan, K., & Zisserman, A. introduced a CNN proving that the number of layers in a deep network affects the A. Preprocessing performance [7]. They showed that results are improved by Numerals in Arabic, Urdu and Persian are written from left increasing the depth to 16-19 weighted layers. Their network to right. Their digits are mostly similar except for the numbers ended up winning second place in the 2014 ILSVRC. He, K. et (4) four, (5) five and (6) six. In this paper, we target the al. in [8] introduced residual network CNN and won the recognition of handwritten digits of these languages in addition ILSVRC 2015. The error rate achieved was 3.6%. Huang et al. to Western Arabic (English) and Devanagari which are all [9] introduced the densely connected CNN which exhibited shown in table 1. improvement over some previous architectures. Alghazo et al. also proposed an efficient geometric feature based numeral TABLE I. REPRESENTATION NUMERALS IN EASTERN ARABIC, recognition system for Arabic and Persian numerals [10]. Yet WESTERN ARABIC, PERSIAN, URDU AND DEVANAGARI the CNN introduced in [8] is still the most popular and preferred Western Eastern CNN architecture for practical applications. Persian Urdu Devanagari Arabic Arabic ० ٠ ٠ A number of papers have already been completed on digit or 0 १ ١ ١ character recognition using some types of Deep Learning 1 Architectures. For example, Singh et al. [11] compared the use २ ٢ ٢ of the deep Convolutional Neural Network and the fully 2 ३ ٣ ٣ connected Feed-forward Neural Network for the recognition of 3 the Devangari handwritten characters. They reported a 98.11% ४ ۴ ٤ 4 recognition rate. In [12], Deep-Learning Feed forward ۵ ५ ٥ Backpropagation Neural Network (DFBNN) along with 5 ۶ ६ ٦ Extreme Learning Machine (ELM) were used in handwritten 6 Numeral Recognition. The methods were applied to datasets in ७ ٧ ٧ Thai, Bangla and Devangari. DRBNN outperformed ELM 7 ८ ٨ ٨ slightly. In DFBNN, the accuracy rates were 98.4% for Thai, 8 95.84% for Bangla, and 78.4% for Devangari. Deep Belief ९ ٩ ٩ 9 Networks (DBN), A Deep learning Structure, was used in [13] for Bangla Handwritten Numeral Recognition. The recognition accuracy was 90.27%. DBN was also used in [14] for high- Handwriting styles are different from one person to another resolution Synthetic Aperture Radar (SAR) image classification. and even the handwriting of the same person can be different Alghazo proposed an online numerals recognition system using according to several variables. For example, more than 52 structural
Load more

Recommended publications
  • The Number-System of Algebra : Treated Theoretically and Historically
    - ; THE NUMBER-SYSTEM OF ALGEBRA TKEATED THEORETICALLY AND HISTORICALLY BY HENRY B. FINE, PH.D. PROFESSOR OF MATHEMATICS IN PRINCETON UNIVERSITY SECOND EDITION, WITH CORRECTIONS BOSTON, U.S.A. D. C. HEATH & CO., PUBLISHERS 1907 COPYRIGHT, 1890, BY HENRY B. FINE. 66 f 6 T PREFACE. THE theoretical part of this little book is an elementary of exposition the nature of the number concept, of the posi- tive integer, and of the four artificial forms of number which, with the positive integer, constitute the "number- " system of algebra, viz. the negative, the fraction, the irra- tional, and the imaginary. The discussion of the artificial numbers follows, in general, the same lines as my pam- phlet : On the Forms of Number arising in Common Algebra, but it is much more exhaustive and thorough- going. The point of view is the one first suggested by Peacock and Gregory, and accepted by mathematicians gen- erally since the discovery of quaternions and the Ausdeh- nungslehre of Grassmann, that algebra is completely defined formally by the laws of combination to which its funda- mental operations are subject; that, speaking generally, these laws alone define the operations, and the operations the various artificial numbers, as their formal or symbolic results. This doctrine was fully developed for the neg the fraction, and the imaginary by Hankel, in his Complexe Zahlensystemen, in 1867, and made complete by Cantor's beautiful theory of the irrational in 1871, but it has not as yet received adequate treatment in English. this kind is Any large degree of originality in work of from a naturally out of the question.
    [Show full text]
  • A Review on Handwritten Devnagari Numerals Recognition Using Support Vector Machine
    Proceedings of VESCOMM-2016 4th NATIONAL CONFERENCE ON “RECENT TRENDES IN VLSI, EMBEDED SYSTEM, SIGNAL PROCESSING AND COMMUNICATION In Association with “The Institution of Engineers (India)” and IJRPET February 12th, 2016 Paper ID: VESCOMM03 A REVIEW ON HANDWRITTEN DEVNAGARI NUMERALS RECOGNITION USING SUPPORT VECTOR MACHINE Rupali Vitthalrao Suryawanshi Department of Electronics Engineering, Bharatratna Indira Gandhi College, Solapur, India, 413006 [email protected] Abstract— Handwriting has continued to persist as a mean of SVM stands for Support Vector Machine and this method of communication and recording information in day-to-day life even recognition is gaining immense popularity now-a-days. SVM with the invention of new technologies. Natural handwriting is classifier gives better results as compared to the other one of the easiest ways of information exchange between a human classifiers for the handwritten numeral recognition of Kannada and a computer. Handwriting recognition has attracted many script in [3]. In [6], SVM approach is used to recognize researchers across the world since many years. Recognition of strokes in Telugu script. The set of strokes are segmented into online handwritten Devanagari numerals is a goal of many subsets based on the relative position of the stroke in a research efforts in the pattern recognition field The main goal of character. An SVM based classifier is built for recognition of the work is the recognition of online hand written Devanagari strokes in each subset. A rule based approach is used to numerals using support vector machine. In the data collection recognize a character from the sequence of stroke classes phase, co-ordinate points of the input handwritten numeral are given by the stroke classifier.
    [Show full text]
  • Elements of South-Indian Palaeography, from the Fourth To
    This is a reproduction of a library book that was digitized by Google as part of an ongoing effort to preserve the information in books and make it universally accessible. https://books.google.com ELEMENTS SOUTH-INDIAN PALfi3&BAPBY FROM THE FOURTH TO THE SEVENTEENTH CENTURY A. D. BEIN1 AN INTRODUCTION TO ?TIK STUDY OF SOUTH-INDIAN INSCRIPTIONS AND MSS. BY A. C. BURNELL HON'. PH. O. OF TUE UNIVERSITY M. K. A, ri'VORE PIS I. A SOClfcTE MANGALORE \ BASEL MISSION BOOK & TRACT DEPOSITORY ft !<3 1874 19 Vi? TRUBNER & Co. 57 & 69 LUDOATE HILL' . ' \jj *£=ggs3|fg r DISTRIBUTION of S INDIAN alphabets up to 1550 a d. ELEMENTS OF SOUTH-INDIAN PALEOGRAPHY FROM THE FOURTH TO THE SEVENTEENTH CENTURY A. D. BEING AN INTRODUCTION TO THE STUDY OF SOUTH-INDIAN INSCRIPTIONS AND MSS. BY A. p. j^URNELL HON. PH. D. OF THE UNIVERSITY OF STRASSBUB.G; M. R. A. S.; MEMBKE DE LA S0CIETE ASIATIQUE, ETC. ETC. MANGALORE PRINTED BY STOLZ & HIRNER, BASEL MISSION PRESS 1874 LONDON TRtlBNER & Co. 57 & 59 LUDGATE HILL 3« w i d m « t als ^'ctdjcn kr §anltekcit fiir Mc i|jm bdic<jcnc JJoctorMvk ttcsc fetlings^kit auf rincm fejjcr mtfrckntcn Jfclk bet 1®4 INTRODUCTION. I trust that this elementary Sketch of South-Indian Palaeography may supply a want long felt by those who are desirous of investigating the real history of the peninsula of India. Trom the beginning of this century (when Buchanan executed the only archaeological survey that has ever been done in even a part of the South of India) up to the present time, a number of well meaning persons have gone about with much simplicity and faith collecting a mass of rubbish which they term traditions and accept as history.
    [Show full text]
  • Numbering Systems Developed by the Ancient Mesopotamians
    Emergent Culture 2011 August http://emergent-culture.com/2011/08/ Home About Contact RSS-Email Alerts Current Events Emergent Featured Global Crisis Know Your Culture Legend of 2012 Synchronicity August, 2011 Legend of 2012 Wednesday, August 31, 2011 11:43 - 4 Comments Cosmic Time Meets Earth Time: The Numbers of Supreme Wholeness and Reconciliation Revealed In the process of writing about the precessional cycle I fell down a rabbit hole of sorts and in the process of finding my way around I made what I think are 4 significant discoveries about cycles of time and the numbers that underlie and unify cosmic and earthly time . Discovery number 1: A painting by Salvador Dali. It turns that clocks are not as bad as we think them to be. The units of time that segment the day into hours, minutes and seconds are in fact reconciled by the units of time that compose the Meso American Calendrical system or MAC for short. It was a surprise to me because one of the world’s foremost authorities in calendrical science the late Dr.Jose Arguelles had vilified the numbers of Western timekeeping as a most grievious error . So much so that he attributed much of the worlds problems to the use of the 12 month calendar and the 24 hour, 60 minute, 60 second day, also known by its handy acronym 12-60 time. I never bought into his argument that the use of those time factors was at fault for our largely miserable human-planetary condition. But I was content to dismiss mechanized time as nothing more than a convenient tool to facilitate the activities of complex societies.
    [Show full text]
  • A History of Elementary Mathematics, with Hints on Methods of Teaching
    ;-NRLF I 1 UNIVERSITY OF CALIFORNIA PEFARTMENT OF CIVIL ENGINEERING BERKELEY, CALIFORNIA Engineering Library A HISTORY OF ELEMENTARY MATHEMATICS THE MACMILLAN COMPANY NEW YORK BOSTON CHICAGO DALLAS ATLANTA SAN FRANCISCO MACMILLAN & CO., LIMITED LONDON BOMBAY CALCUTTA MELBOURNE THE MACMILLAN CO. OF CANADA, LTD. TORONTO A HISTORY OF ELEMENTARY MATHEMATICS WITH HINTS ON METHODS OF TEACHING BY FLORIAN CAJORI, PH.D. PROFESSOR OF MATHEMATICS IN COLORADO COLLEGE REVISED AND ENLARGED EDITION THE MACMILLAN COMPANY LONDON : MACMILLAN & CO., LTD. 1917 All rights reserved Engineering Library COPYRIGHT, 1896 AND 1917, BY THE MACMILLAN COMPANY. Set up and electrotyped September, 1896. Reprinted August, 1897; March, 1905; October, 1907; August, 1910; February, 1914. Revised and enlarged edition, February, 1917. o ^ PREFACE TO THE FIRST EDITION "THE education of the child must accord both in mode and arrangement with the education of mankind as consid- ered in other the of historically ; or, words, genesis knowledge in the individual must follow the same course as the genesis of knowledge in the race. To M. Comte we believe society owes the enunciation of this doctrine a doctrine which we may accept without committing ourselves to his theory of 1 the genesis of knowledge, either in its causes or its order." If this principle, held also by Pestalozzi and Froebel, be correct, then it would seem as if the knowledge of the history of a science must be an effectual aid in teaching that science. Be this doctrine true or false, certainly the experience of many instructors establishes the importance 2 of mathematical history in teaching. With the hope of being of some assistance to my fellow-teachers, I have pre- pared this book and have interlined my narrative with occasional remarks and suggestions on methods of teaching.
    [Show full text]
  • The Project Gutenberg Ebook #31061: a History of Mathematics
    The Project Gutenberg EBook of A History of Mathematics, by Florian Cajori This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at www.gutenberg.org Title: A History of Mathematics Author: Florian Cajori Release Date: January 24, 2010 [EBook #31061] Language: English Character set encoding: ISO-8859-1 *** START OF THIS PROJECT GUTENBERG EBOOK A HISTORY OF MATHEMATICS *** Produced by Andrew D. Hwang, Peter Vachuska, Carl Hudkins and the Online Distributed Proofreading Team at http://www.pgdp.net transcriber's note Figures may have been moved with respect to the surrounding text. Minor typographical corrections and presentational changes have been made without comment. This PDF file is formatted for screen viewing, but may be easily formatted for printing. Please consult the preamble of the LATEX source file for instructions. A HISTORY OF MATHEMATICS A HISTORY OF MATHEMATICS BY FLORIAN CAJORI, Ph.D. Formerly Professor of Applied Mathematics in the Tulane University of Louisiana; now Professor of Physics in Colorado College \I am sure that no subject loses more than mathematics by any attempt to dissociate it from its history."|J. W. L. Glaisher New York THE MACMILLAN COMPANY LONDON: MACMILLAN & CO., Ltd. 1909 All rights reserved Copyright, 1893, By MACMILLAN AND CO. Set up and electrotyped January, 1894. Reprinted March, 1895; October, 1897; November, 1901; January, 1906; July, 1909. Norwood Pre&: J. S. Cushing & Co.|Berwick & Smith.
    [Show full text]
  • Support Vector Machine for Handwritten Devanagari Numeral Recognition Shailedra Kumar Shrivastava Sanjay S
    International Journal of Computer Applications (0975 – 8887) Volume 7– No.11, October 2010 Support Vector Machine for Handwritten Devanagari Numeral Recognition Shailedra Kumar Shrivastava Sanjay S. Gharde HOD, Information Technology Department Research Scholar, Information Technology Department Samrat Ashok Technological Institute Samrat Ashok Technological Institute Vidisha, (M. P.) INDIA Vidisha, (M. P.) INDIA ABSTRACT characters by applying different techniques, but very less work Support Vector Machines (SVM) is used for classification in has been performed on Handwritten Devanagari numerals. So, pattern recognition widely. This paper applies this technique for this research work has been conducted on Handwritten recognizing handwritten numerals of Devanagari Script. Since Devanagari numeral. benchmark database does not exist globally, this system is constructed database by implementing Automated Numeral Recognition of Handwritten Devanagari Numerals or Extraction and Segmentation Program (ANESP). Preprocessing is Characters [1, 2] is a complicated task due to the unconstrained manifested in the same program which reduces most of the shape variations, different writing style and different kinds of efforts. 2000 samples are collected from 20 different people noise. Also, handwriting depends much on the writer and because having variation in writing style. Moment Invariant and Affine we do not always write the same digit in exactly the same way, Moment Invariant techniques are used as feature extractor. These building a general recognition system that would recognize any techniques extract 18 features from each image which is used in digit with good reliability in every application is not possible. In Support Vector Machine for recognition purpose. Binary recent time, specialists have made use of different techniques classification techniques of Support Vector Machine is such as Modified Discrimination Function (MQDF) [10], implemented and linear kernel function is used in SVM.
    [Show full text]
  • Preliminary Proposal to Encode Devanagari Letter Numerals L2/17
    Preliminary proposal to encode Devanagari letter numerals Srinidhi A and Sridatta A Tumakuru, India [email protected], [email protected] January 17, 2017 1 Introduction This is preliminary proposal to encode letter numerals found in Devanagari script. The objective is to bring them to the attention of the Unicode Technical Committee and to seek advice regarding the encoding of the numerals. 2 Background The modern Devanagari has a set of digits with the use of zero. In addition to digits, a distinct numerical notation system is attested early and medieval manuscripts. These numerals are referred as ‘letter numerals’ in the scholarly literature to distinguish from regular digits. The letter numerals were largely used for marking page numbers in manuscripts. These numbers are derived from Brahmi numbers (U+11052…U+11065). These Devanagari numerals are paleographically analogous to letter numerals found in closely related North Indic scripts like Bhaikṣukī, Tocharian, Khotanese, Ranjana, Newa and Sharada. Often the numbers were used alongside with digits. Letter numerals are written in the margin to the left, whereas digits are written in the margin to the right. North Indian letters numerals are distinct from alphasyllabic numeration used in of other parts of India such as Aksharapalli and Katapayadi systems. In these notations existing letters are reused, whereas the North Indian numerals are evolved from Brahmi numbers. The term ‘letter-numerals’ here does not symbolize that these numerals are evolved or conceptually derived from letters. These numerals were used predominantly in states of Gujarat and Rajasthan of Western India. These numerals occur in both ‘Jainanagari’ and regular styles of Devanagari.
    [Show full text]
  • Neural Combination of ANN and HMM for Handwritten Devanagari Numeral Recognition U
    Neural Combination of ANN and HMM for Handwritten Devanagari Numeral Recognition U. Bhattacharya, S.K. Parui, B. Shaw, K. Bhattacharya To cite this version: U. Bhattacharya, S.K. Parui, B. Shaw, K. Bhattacharya. Neural Combination of ANN and HMM for Handwritten Devanagari Numeral Recognition. Tenth International Workshop on Frontiers in Handwriting Recognition, Université de Rennes 1, Oct 2006, La Baule (France). inria-00104481 HAL Id: inria-00104481 https://hal.inria.fr/inria-00104481 Submitted on 6 Oct 2006 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Neural Combination of ANN and HMM for Handwritten Devanagari Numeral Recognition U. Bhattacharya S. K. Parui B. Shaw K. Bhattacharya CVPR Unit CVPR Unit CVPR Unit B-Tech Final Year Indian Statistical Indian Statistical Indian Statistical Dept. of CS&E Institute, Kolkata Institute, Kolkata Institute, Kolkata University of Calcutta [email protected] [email protected] mrbikashshaw kuntal.bhattacharya @yahoo.co.in @gmail.com Abstract recognition of hand printed Devanagari characters is found in [3]. For recognition of handwritten Devanagari In this article, a two-stage classification system for numerals, Ramakrishnan et al. [4] used independent recognition of handwritten Devanagari numerals is component analysis technique for feature extraction presented.
    [Show full text]
  • Offline Handwritten Character Recognition Using Neural Networks
    International Journal of Computer Sciences and Engineering Open Access Research Paper Vol.-7, Issue-5, May 2019 E-ISSN: 2347-2693 Offline Handwritten Character Recognition using Neural Networks Hemant Yadav1*, Sapna Jain2 1 Department of Computer Science, MRK Institute of Engineering and Technology, Saharanwas, Rewari, India 2 Department of Computer Science, MRK Institute of Engineering and Technology, Saharanwas, Rewari, India *Corresponding Author: [email protected], Tel.: +91-98182-63577 DOI: https://doi.org/10.26438/ijcse/v7i5.838845 | Available online at: www.ijcseonline.org Accepted: 10/May/2019, Published: 31/May/2019 Abstract— Handwritten character recognition is currently a under research field. A lot research is getting done in this field in which the point of interest is to receive as higher accuracy as possible in a distorted writing. That is as we know the way of writing of different person is different, so to recognize every writing with a greater accuracy is the point of concern. In this paper we proposed a method for different languages handwritten character recognition. The main focus is to train the model with pre-set data and then using that trained model to test the handwritten character passed to it. In our proposed method we used MATLAB to design our code, in this the model can be trained on runtime also. Keywords— Handwritten Character, Character Recognition, Feature Extraction, Neural Networks, Image Recognition, Offline Character Recognition. I. INTRODUCTION Indian scripts' character recognition, most of the published works deal with pre-printed papers, but there are only rare Handwriting is an age-old and elegant way of research papers representing handwritten character communication and is a distinctive characteristic of an recognition.
    [Show full text]
  • Radial Basis Function for Handwritten Devanagari Numeral Recognition
    (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 2, No. 5, 2011 Radial Basis Function For Handwritten Devanagari Numeral Recognition Prerna Singh Nidhi Tyagi M.Tech Student Assistant Professor Shobhit University, ISCEIT. Shobhit University, Meerut, India Meerut, India Abstract—The task of recognizing handwritten numerals, using a different languages like English [2] ,Chinese[3] , Arabian classifier, has great importance. This paper applies the technique ,Persian , Bengali , Guajarati .Different classifiers have been of Radial Basis Function for handwritten numeral recognition of used for handwritten digit recognition using neural network. Devanagari Script. Lot of work has been done on Devanagari Recently, signifi- cant contribution towards the improvement of numeral recognition using different techniques for increasing the recognition rates have been made by means of different accuracy of recognition. Since the database is not globally combination strategies [4] and by the use of support vector created, firstly we created the database by implementing pre- machines [5] . processing on the set of training data. Then by the use of Principal Component Analysis we have extracted the features of Machine learning is a artificial system[6] capable of each image, some researchers have also used density feature autonomous acquisition & integration of knowle- dge.This extraction. Since different people have different writing style, so capacity to learn from experience, analy- tical observation other here we are trying to form a system where recognition of numeral means, results in system that can improve its own speed or becomes easy. Then at the hidden layer centers are determined performance and allows computer to “learn“. It can be and the weights between the hidden layer and the output layer of categorized into supervised learning, unsupervised learning, each neuron are determined to calculate the output, where output semi-supervised learning and reinforcement learning [7].
    [Show full text]
  • Comparison of Feature Extraction Techniques for Gujarati Isolated Numerals Dr
    International Journal of Modern Trends in Engineering and Research www.ijmter.com e-ISSN No.:2349-9745, Date: 28-30 April, 2016 Comparison of Feature Extraction Techniques for Gujarati Isolated Numerals Dr. Mamta Baheti Department of Computer Science, Hislop College, Nagpur [email protected] Abstract— This paper is for recognition of isolated Gujarati numerals. As per the reported work, Gujarati is widely spoken across world along with Indian states. In the proposed study, we have used noisy handwritten numerals in procedure of training and testing. After applying pre-processing to the scanned image, they are then treated with the proposed algorithm. In our proposed algorithm we have used affine invariant moments and invariant moments as feature extraction technique and Gaussian distribution function as classifier. We produced pleasing results for some numerals. We compared the results for both the extraction techniques and found that for our database Affine invariant moments gave better results but they could be improved by giving better quality images for training and testing. Keywords- Affine Invariant Moments; Invariant Moments; Gaussian Distribution Function I. INTRODUCTION In the present scenario, every day task is being computerized. Each user wants the computer to have user-friendly interactions. In recent years a good amount of reported work has been contributed for recognition of handwritten characters. Variation in writing styles and recognition accuracy has a milestone to achieve. It has been noted that there are approximately 22 regional official languages are used in India [1-4]. Every regional official language has the wide variety as well some peculiar similarities. Out of these, one language is Gujarati which is spoken in Gujarat and some parts of India and also in South Africa.
    [Show full text]