Abstract This study investigates the combination of different classifiers to improve Arabic handwritten word recognition. Features based on Discrete Cosine Transform (DCT) and Histogram of Oriented Gradients (HOG) are computed to represent the handwritten words. The dimensionality of the HOG features is reduced by applying Principal Component Analysis (PCA). Each set of features is separately fed to two different classifiers, Support Vector Machine (SVM) and Fuzzy K-Nearest Neighbor (FKNN) giving a total of four independent classifiers. A set of different fusion rules is applied to combine the output of the classifiers. The proposed scheme evaluated on the IFN/ENIT database of Arabic handwritten words reveal that combining the classifiers results in improved recognition rates which, in some cases, outperform the state-of-the-art recognition systems.

References

[1] Abdel-Azeem S. and Ahmed H., Effective Technique for the Recognition of Offline Arabic Handwritten Words using Hidden Markov Models, International Journal on Document Analysis and Recognition, vol. 16, no. 4, pp. 399- 412, 2013.

[2] Abdulla S., Al-Nassiri A., and Abdul-Salam R., Off-Line Arabic Handwritten Word Segmentation Using Rotational Invariant Segments Features, The international Arab Journal of Information Technology, vol. 5, no. 2, pp. 200-208, 2008.

[3] Alalshekmubarak A., Hussain A., and Wan Q., Neural Information Processing, Springer Link, 2012.

[4] Alginahi Y., A Survey on Arabic Character Segmentation, International Journal on Document Analysis and Recognition, vol. 16, no. 2, pp. 105-126, 2013.

[5] Al-Hajj R., Likforman-Sulem L., and Mokbel C., Combining Slanted-Frame Classifiers for Improved HMM-based Arabic Handwriting Recognition, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 31, no. 7, pp. 1165-1177, 2009.

[6] AlKhateeb J., Jiang J., Ren J., Khelifi F., and Ipson S., Multiclass. Classification of Unconstrained Handwritten Arabic Words Using Machine Learning Approaches, The Open Signal Processing Journal, vol. 2, pp. 21-28, 2009.

[7] Al-Rashaideh H., Preprocessing phase for Arabic Word Handwritten Recognition, Information Transmissions in Computer Networks, vol. 6, no. 1, pp. 11-19, 2006.

[8] Benouareth A., Ennaji A., and Sellami M., Semi- continuous HMMs with Explicit State Duration for Unconstrained Arabic Word Modeling and Recognition, Pattern Recognition Letters, vol. 29, no. 12, pp. 1742-1752, 2008.

[9] Chang C. and Lin C., LIBSVM: A Library for Support Vector Machines, ACM Transactions on Intelligent Systems and Technology, vol. 2, no. 3, pp. 1-39, 2011.

[10] Chen S. and Liu C., Computer Analysis of Images and Patterns, Springer Link, 2011.

[11] Chergui L., kef M., and Chikhi S., Combining Neural Networks for Arabic Handwritten Recognition, The international Arab Journal of Information Technology, vol. 9, no. 6, pp. 588- 595, 2012.

[12] Cheriet M., Lecture Notes in Computer Science, Springer Link, 2008.

[13] Cortes C. and Vapnik V., Support-vector network Machine Learning, Springer Link, 1995.

[14] Dalal N. and Triggs B., Histograms of Oriented Gradients for Human Detection, in Proceeding of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, California, 2005.

[15] El-Abed H. and M rgner V., Comparison of Different Preprocessing and Feature Extraction Methods for Offline Recognition of Hand- Written Arabic Words, in Proceeding of 9th International Conference on Document, Parana, 2007.

[16] El-Feghi L., Elmahjoub F., Alswady B., and Baiou A., Offline Handwritten Arabic Words Recognition using Zernike Moments and Hidden Markov Models, in Proceeding of International Conference on Computer Applications and Industrial Electronics, Kuala Lumpur, pp. 165 - 168, 2010.

[17] El-Melegy M. and Abdelbaset A., Global features for offline recognition of handwritten Arabic literal amounts, in proceeding of 5th International Conference on Information and Communications Technology, Cairo, pp. 125- 129 2007.

[18] Farah L., Farah N., Bedda M., Arabic word Recognition by fuzzy classifier, Journal of applied sciences, vol. 6, no. 3, pp. 647-650, 2006.

[19] Farou B., Hallaci S., and Seridi H., Syst me Neuro-Markovien Pour la Reconnaissance de L criture Manuscrite arabe Vocabulaire Limit , in Proceeding of 2nd Conf rence Internationale sur l'informatique et ses Applications, Saida, pp. 1-11, 2009.

[20] Iamsa S. and Horata P., Handwritten Character Recognition Using Histograms of Oriented Gradient Features in Deep Learning of Artificial Neural Network, in Proceeding of International Conference on IT Convergence and Security, Macao, pp. 1-5, 2013.

[21] Keller J., Gray M., and Givens J., A Fuzzy K Nearest Neighbor Algorithm, IEEE Transaction on Systems, Man, and Cybernetics, vol. 15, no. 4, pp. 580-585, 1985.

[22] Khalifa M. and BingRu Y., Advanced Research on Electronic Commerce Web Application, and Communication, Communications in Computer and Information Science , Springer Link, 2011.

[23] Kuncheva L., Combining Pattern Classifiers: Methods and Algorithms, Wiley-Interscience, 2004. 720 The International Arab Journal of Information Technology, Volume 14, No. 5, September 2017

[24] Lawgali A., Bouridane A., Angelova M., and Ghassemlooy Z., Handwritten Arabic Character Recognition: Which Feature Extraction Method?, International Journal of Advanced Science and Technology, vol. 34, pp. 1-8, 2011.

[25] Lorigo L. and Govindaraju V., offline Arabic Handwriting Recognition: a Survey, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 28, no. 5, pp.712-724, 2006.

[26] Ludwig O., Delgado D., Goncalves V., Nunes U., Trainable Classifier-Fusion Schemes: An Application to Pedestrian Detection, Proc. 12th International IEEE Conference on Intelligent Transportation Systems, St. Louis, vol. 1, pp. 432-437, 2009.

[27] Mirvaziri H., Javidi M., and Mansouri N., Visual Informatics: Bridging Research and Practice, Springer Link, 2009.

[28] Nemouchi S., Meslati L., and Farah N., Image and Signal Processing, Lecture Notes in Computer Science, Springer Link, 2012.

[29] Pechwitz M., Maddouri S., M rgner V., Ellouze N., and Amiri H., IFN/ENI-Database of Arabic Handwritten Words, in Proceeding of 7th Colloque International Francophone sur l Ecrit et le Document, Hammamet, pp. 129-136, 2002.

[30] Rao K. and Yip P., Cosine Transform: Algorithms, Advantages, Applications, Academic Press, 1990.

[31] Srivastava D. and Bhambhu L., Data Classification Using Support Vector Machine, Journal of Theoretical and Applied Information Technology, vol. 12, no. 1, pp. 1-7, 2010.

[32] Williams K., Suleman H., and Jorgina K., A Comparison of Machine Learning Techniques for Handwritten |Xam Word Recognition, in Proceeding of the South African Institute for Computer Scientists and Information Technologists Conference, East London, pp. 37- 46, 2013.

[33] Xu L., Krzyzak A., and Y-Suen C., Methods of Combining Multiple Classifiers and Their Applications to Handwriting Recognition, IEEE Transactions on Systems, Man, and Cybernetics, vol. 22, no. 3, pp. 418-435, 1992.

[34] Zhang T. and Suen C., A Fast Parallel Algorithm for Thinning Digital Patterns, Communication of the ACM, vol. 23, No. 3, pp. 236-239, 1984. Rachid Zaghdoudi received his MSc degree in Electronics from University of 08Mai 1945, Guelma, Algeria, in 2008. Currently, he is a researcher in the laboratory of Science and Information Technologies and Communication LabSTIC , University of 8 Mai 1945 of Guelma and member in GADM team at LabSTIC laboratory, Guelma, Algeria. His current research interests include Handwriting recognition, Classification, artificial Intelligence, pattern recognition, and document processing. Hamid Seridi received his Bachelor s degree with honours in 1981, from the University of Annaba, Algeria, and the Master s degree from the Polytechnic Institute of New-York, USA in 1984, both in Electrical Engineering. He received his PhD in Computer Science with distinction in 2001 from the University of Reims, France. He was Vice Dean of the Post-Graduation, Scientific Research and External Relations in the University of Guelma. Currently he is Professor and Director of Laboratory of Science and Information Technologies and Communication LabSTIC . He is also Chairman of the Scientific Council of the Faculty of Mathematics and Computing and Material Sciences. He is an expert member at the national committee for evaluation and accreditation national projects research. His research interests include approximate knowledge management, pattern recognition and artificial intelligence, data mining, video mining, machine learning and cryptography.