SVM Classifier for Recognition of Handwritten Devanagari Numeral

This paper is published in proceedings of International Conference on Image Information Processing (ICIIP 2011), an IEEE conference

Title: SVM Classifier for Recognition of Handwritten Devanagari Numeral

SVM Classifier for Recognition of Handwritten Devnagari Numeral

Authors:
Mahesh Jangid, Renu Dhir, Rajneesh Rani, Kartar Singh

Abstract:
In this manuscript we recognize the handwritten Devnagari numerals. In our implementation we have used density and background directional distribution features for the zones, in which we divided the numeral samples already. We used the normalized images of samples of varying sizes of 32*32, 40*40 and 48*48. We divided these normalized images into 4*4 (16), 5*5 (25) and 6*6 (36) zones respectively to compute the features for each zone. In all the cases each zone is of size 8*8 pixels. Each zone contains 9 features consisting of one density feature and 8 backgrounds directional distribution features. The zonal density feature is computed by dividing the number of foreground pixels in each zone by total number of pixels in the zone i.e. 64. The other 8 features are based on directional distribution values of background in eight directions. These directional values are computed for each foreground pixel by summing up the value corresponding to neighbouring background pixels given in the specific mask for each direction. For each direction these directional distribution features are summed up for all pixels in each zone. Thus numbers of features finally used for recognition are 144, 225 and 324 for samples of respective sizes in increasing order. For classification purpose we have used SVM classifier with RBF kernel. Our dataset of handwritten Devnagari numerals used is provided by Indian Statistical Institute (ISI), Kolkata. Training data size is 18783 and testing data size is 3763, totally 22546. The optimum 5-fold cross validation accuracies of training data obtained for varying sizes of samples in increasing order are 98.76%, 98.91% and 98.94% respectively. By observing the cross validation results it is conclusive that at the cost of increasing the features size there is only minute increases in the performance. So we recommend 144 sized feature vector to recognize testing samples. The testing accuracy by using 144 features for 32*32 normalized samples observed is 98.51% which is prominent and cost-efficient.

Keywords:
Devanagari Numerals, Zoning density, Background directional distribution, SVM classifier

Citation:
Mahesh Jangid, Renu Dhir, Rajneesh Rani, Kartar Singh, “SVM Classifier for Recognition of Handwritten Devanagari Numeral”, Proceedings of International Conference on Image Information Processing (ICIIP 2011), 2011.

References:

1. U.Pal and B B Choudhuri, “Indian script character recognition: A survey” Pattern Recognition ,Vol 37,pp 1887-1899,2004.
2. R M K Sinha, “ A journey from Indian scripts processing to Indian language processing “, IEEE Ann. Hist. Computer, vol 31, no 1, pp 831, 2009.
3. I.K, Sethi and B. Chatterjee, “Machine Recognition of constrained Hand printed Devanagari”, Pattern Recognition, Vol. 9,pp. 69-75, 1977.
4. G S Lehal, Nivedan Bhatt, “A Recognition System for Devnagri and English Handwritten Numerals”, Proc. Of ICMI, 2000.
5. Reena Bajaj, Lipika Day, Santanu Chaudhari, “Devanagari Numeral Recognition by Combining Decision of Multiple Connectionist Classifiers”, Sadhana, Vol.27, Part-I, 59-72, 2002.
6. R.J.Ramteke, S.C.Mehrotra, “Recognition Handwritten Devanagari Numerals”, International journal of Computer processing of Oriental languages, 2008.
7. U. Bhattacharya, S. K. Parui, B. Shaw, K. Bhattacharya, “Neural Combination of ANN and HMM for Handwritten Devnagari Numeral Recognition”.
8. U.Pal, R.K.Roy and F. Kimura,”Indian muilti-script full pin-code string recognition for postal automation” in Proc. 10 th conf. Document Analysis Recognition, 2009, pp 456-460
9. U. Pal, T. Wakabayashi, N. Sharma and F. Kimura, “Handwritten Numeral Recognition of Six Popular Indian Scripts”, Proc. 9th ICDAR, Curitiba, Brazil, Vol.2 (2007), 749-753.
10. S. Knerr, L. Personnaz, and G. Dreyfus. Single-layer learning revisited: a stepwise procedure for building and training a neural network. In J. Fogelman, editor, Neu-rocomputing: Algorithms, Architectures and Applications. Springer-Verlag, 1990.
11. U. H.-G. Kressel. Pairwise classication and support vector machines. In B. Scholkopf, C. J. C. Burges, and A. J. Smola, editors, Advances in Kernel Methods { Support Vector Learning, pages 255{268, Cambridge, MA, 1998. MIT Press.
12. Chih-Chung Chang and Chih-Jen Lin, LIBSVM: a library for support vector machines, 2001. Software available at http://www.csie.ntu.edu.tw/~cjlin/libsvm
13. Mahesh Jangid, Kartar Singh, Renu Dhir, Rajneesh Rani, “Performance Comparison of Devanagari Handwritten Numerals Recognition”, Internation Journal of Computer Applications (IJCA), Vol. 22, No.1, May 2011.
14. Kartar Singh Siddharth, Mahesh Jangid, Renu Dhir, Rajneesh Rani, “Handwritten Gurmukhi Character Recognition Using Statistical and Background Directional Distribution Features”, International Journal of Computer Science and Engineering (IJCSE), Vol. 3, No. 6, June 2011.
15. Chih-Chung Chang and Chih-Jen Lin, LIBSVM: a library for support vector machines, 2001. Software available at http://www.csie.ntu.edu.tw/~cjlin/libsvm.
16. Chih-Wei Hsu, Chih-Chung Chang, and Chih-Jen Lin, “A Practical Guide to Support Vector Classification”, [Online]. Available: http://www.csie.ntu.edu.tw/~cjlin/papers/guide/guide.pdf
17. http://www.isical.ac.in/~ujjwal/download/database.html.
18. Gonzalez, Woods and Eddins,”a book on Digital Image Processing Using MATLAB”

Access the paper:

On IEEE Xplore
Hosted on this blog

Handwritten Gurmukhi Character Recognition Using Zoning Density and Background Directional Distribution Features

This is my first research paper published in May-June issue of IJCSIT (International Journal of Computer Science and Information Technologies). IJCSIT Vol 2 Issue 3

Handwritten Gurumukhi Character Recognition Using Zoning Density and Background Directional Distribution Features

Abstract:
This paper presents an approach to recognize isolated handwritten Gurumukhi characters. Extracted features used are 16 zonal densities and the 8 background directional distribution features for each of 16 zones. All sample images of Gurumukhi characters used are normalized to 32*32 pixel sizes. Zoning density is computed by dividing number of foreground pixels in each zone by number of total pixels in the zone. Background directional distribution values are calculated for each foreground pixel by directional distribution of its neighbouring background pixels. The value for each directional distribution is computed by summing up the values specified in a mask for corresponding neighbour background pixels. A specific mask is used in particular direction. These feature values for each direction are summed up for all pixels in each zone. Thus adding 16 zoning density features and 128 background directional distribution features (8 features in each of 16 zones) we used feature vector containing 144 features in our experiment. SVM classifier with RBF kernel is used for classification. We have taken 200 samples of each of 35 basic Gurumukhi characters in our experiment, which are collected from 20 different writers each contributing to write 10 samples of each of 35 characters. Thus we have used total 7000 character samples. By training the classifier with whole dataset we obtained 95.04% 5-fold cross validation accuracy. By using dataset taken by 16 writers (size 5600) to train the system and testing the 1400 samples by another 4 unknown writers to classifier, we obtained 89.14% accuracy. By splitting the handwritten samples of each writer to train and test the system in the ratio of 4:1, we received 99.93% test accuracy for known writers. Thus we got 94.53% average test accuracy for known and unknown writers and 95.04% 5-fold cross validation accuracy.

Authors:
Kartar Singh Siddharth, Renu Dhir, Rajneesh Rani

Keywords:
Isolated Handwritten Gurumukhi Character Recognition, Zoning Density, Background Directional Distribution, SVM Classifier.

Citation:
Kartar Singh Siddharth, Renu Dhir, Rajneesh Rani, “Handwritten Gurumukhi Charater Recognition Using Zoning Density and Background Directional Features”, (IJCSIT) International Journal of Computer Science and Information Technologies, Vol. 2, Issue 3, pp. 1036-1041, May-June 2011.

Full access of the paper:
IJCSIT
Scribd
Google Docs
on this blog