A handwritten Bangla numeral recognition scheme based on expanded two-layer SOM

This paper proposes a system for handwritten Bangla numeral recognition based on expanded two-layer self-organising map (SOM), in which every map in the second layer expands from a corresponding neuron in the first layer map. It carries out multiple classifications in the second layer for each character sample. The exact classification of the character image is obtained by a fusion algorithm using confidence coefficients. The discriminability of SOM is improved by this structure. With the directional and density features as the input vector, the experiments on handwritten Bangla numeral samples have achieved satisfactory recognition performance.

[1]  Sargur N. Srihari,et al.  Class-wise multi-classifier combination based on Dempster-Shafer theory , 2002, 7th International Conference on Control, Automation, Robotics and Vision, 2002. ICARCV 2002..

[2]  Adam Krzyżak,et al.  Methods of combining multiple classifiers and their applications to handwriting recognition , 1992, IEEE Trans. Syst. Man Cybern..

[3]  Bidyut Baran Chaudhuri,et al.  Automatic Recognition of Unconstrained Off-Line Bangla Handwritten Numerals , 2000, ICMI.

[4]  Yue Lu,et al.  A high reliability classifier using decision trees and AdaBoost for recognizing handwritten Bangla numerals , 2007, 2007 International Conference on Wavelet Analysis and Pattern Recognition.

[5]  Adnan Amin,et al.  Off-line Arabic character recognition: the state of the art , 1998, Pattern Recognit..

[6]  Sung-Bae Cho,et al.  Neural-network classifiers for recognizing totally unconstrained handwritten numerals , 1997, IEEE Trans. Neural Networks.

[7]  Daming Shi,et al.  Handwritten Chinese character recognition using nonlinear active shape models and the Viterbi algorithm , 2002, Pattern Recognit. Lett..

[8]  Ponnuthurai N. Suganthan.  Hierarchical overlapped SOM's for pattern classification , 1999, IEEE Trans. Neural Networks.

[9]  Pengfei Shi,et al.  Handwritten Bangla numeral recognition system and its application to postal automation , 2007, Pattern Recognit..

[10]  Teuvo Kohonen,et al.  The self-organizing map , 1990 .

[11]  François Fouss,et al.  Yet Another Method for Combining Classifiers Outputs: A Maximum Entropy Approach , 2004, Multiple Classifier Systems.

[12]  Toru Wakahara,et al.  Handwritten Japanese character recognition using adaptive normalization by global affine transformation , 2001, Proceedings of Sixth International Conference on Document Analysis and Recognition.

[13]  Ching Y. Suen,et al.  Complementary algorithms for the recognition of totally unconstrained handwritten numerals , 1990, [1990] Proceedings. 10th International Conference on Pattern Recognition.

[14]  Bidyut Baran Chaudhuri,et al.  A Hybrid Scheme for Handprinted Numeral Recognition Based on a Self-Organizing Network and MLP Classifiers , 2002, Int. J. Pattern Recognit. Artif. Intell..

[15]  Ponnuthurai N. Suganthan,et al.  On the performance of the HONG network for pattern classification , 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium.

[16]  Sung-Bae Cho,et al.  Multiple recognizers system using two-stage combination , 1996, Proceedings of 13th International Conference on Pattern Recognition.