EECS 349, Machine Learning

  Insturctor: Prof. Doug Downey
By Majed Valad Beigi
Northwestern University
home Introduction Design Back Propagation NN LAMSTAR NN SVM Results Conclusion

Offline HandWritten Character Recognition

You can download the full project report and source code here:
FullReport_Source Code


An off-line handwritten character recognition using Back Propagation Neural Network, LAMSTAR Neural Network and Support Vector Machine has been described in this report.

A summary of this project is as follow:

1- Scanning the paper page with the handwritten characters on it

2- Extracting sub-images of individual characters form the scanned image using the image processing toolbox of the MATLAB

3- Resizing the sub-images either to a 50 pixel *70 pixel image or a 90 pixel*120 pixel image (Since the cropped sub-images of characters from the last step can have different sizes, they have to be resized to a same standard size to be given as the input to the classifiers)

4- Creating a 50*70 or a 90*120 matrix of Boolean values from each sub-image by assigning ‘0’s to white pixels and ‘1’ to black pixels.

5- Resizing the original large matrix to a smaller matrix (by running a 10*10/15*15 window on the 50*70/90*120 original matrix and finding the average of the values in the window)

6- Feeding the 5*7/6*8 matrix in the form of a 1*35/1*48 matrix to the input of the BP NN or SVM. For the LAMSTAR NN, I have only considered the 6*8 input matrix.

 7- Finally, we can get the results of the classification


As the results from the previous section suggests, LAMSTAR NN is the most efficient and the fastest classifier for solving this problem compared to the two other techniques that have been examined in this project. Moreover, for the BP NN and SVM, as I mentioned earlier I have considered two different sizes for the input matrix and I got higher accuracy for the input that had higher resolution (6*8 input matrix).


[1] E. Tautu and F. Leon, “Optical Character Recognition System Using Support Vector Machines,” pp. 1-13, 2012.

[2] Designing an Intelligent System for Optical Handwritten Character Recognition using ANN

[3] R. Plamondon and S. N. Srihari, “On-line and off- line handwritten character recognition: A comprehensive survey,”IEEE. Transactions on Pattern Analysis and Machine Intelligence, vol. 22, no. 1, pp. 63-84, 2000.

[4] N. Arica and F. Yarman-Vural, “An Overview of Character Recognition Focused on Off-line Handwriting”, IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews, 2001, 31(2), pp. 216 - 233.

[5] U. Bhattacharya, and B. B. Chaudhuri, “Handwritten numeral databases of Indian scripts and multistage recognition of mixed numerals,” IEEE Transaction on Pattern analysis and machine intelligence, vol.31, No.3, pp.444-457, 2009.

[6] U. Pal, T. Wakabayashi and F. Kimura, “Handwritten numeral recognition of six popular scripts,” Ninth International conference on Document Analysis and Recognition ICDAR 07, Vol.2, pp.749-753, 2007.

[7] M. Simner, W. Hulstijn, and P. Girouard (Eds.), “Forensic, developmental and neuropsychological aspects of handwriting,” Special issue of of the Journal of Forensic Document Examination, 1994.

[8] R. Plamondon, “Pattern recognition,” Special Issue on Automatic Signature Verification, vol. 8, no. 3, June 1994. IACSIT International Journal of Engineering and Technology, Vol. 5, No. 2, April 2013

[9] G. P. Van Galen and P. Morasso, “Neuromotor control in handwriting and drawing," Acta Psychologica,, vol. 100, no. 1-2, p. 236, 1998.

[10] R.G. Casey and E.Lecolinet, “A Survey of Methods and Strategies in Character Segmentation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 18, No.7, July 1996, pp. 690-706.

[11] C. L. Liu, H. Fujisawa, “Classification and Learning for Character Recognition: Comparison of Methods and Remaining Problems”, Int. Workshop on Neural Networks and Learning in Document Analysis and Recognition, Seoul, 2005.

[12] F. Bortolozzi, A. S. Brito, Luiz S. Oliveira and M. Morita, “Recent Advances in Handwritten Recognition”, Document Analysis, Umapada Pal, Swapan K. Parui, Bidyut B. Chaudhuri, pp 1-30.

[13] J.Pradeep, E.Srinivasan, S.Himavathi, “Diagonal Based Feature Extraction For Handwritten Character Recognition System Using Neural Network”,Electronics Computer Technology (ICECT),2011, Volume-4, 2011, pp. 364-368.

[14] C.J.C. Burges, “tutorial on support vector machines for pattern recognition. Data Mining and Knowledge Discovery”, 2, pp. 121–167,1998.