Abstract
In this paper, we propose an approach that combines the unsupervised and supervised learning techniques for unconstrained handwritten numeral recognition. This approach uses the Kohonen self-organizing neural network for data classification in the first stage and the learning vector quantization (LVQ) model in the second stage to improve classification accuracy. The combined architecture performs better than the Kohonen self-organizing map alone. In the proposed approach, the collection of centroids at different phases of training plays a vital role in the performance of the recognition system. Four experiments have been conducted and experimental results show that the collection of centroids in the middle of the training gives high performance in terms of speed and accuracy. The systems developed also resolve the confusion between handwritten numerals.
Key words
Feature extraction self-organizing map learning vector quantiza-tion handwritten numeral recognition substitution error classificationPreview
Unable to display preview. Download preview PDF.
References
- Burr D J 1987 Experiment with a connectionist text reader.IEEE First Conf. on Neural Networks 4: 717–724Google Scholar
- Cai J, Liu Z-Q 1999 Integration of structural and statistical information for unconstrained handwritten numeral recognitionIEEE Trans. Pattern Anal. Mach. Intell. 21: 263–267CrossRefGoogle Scholar
- Cho S B 1997 Neural network classifiers for recognizing totally unconstrained handwritten numerals.IEEE Trans. Neural Networks 8: 43–53CrossRefGoogle Scholar
- David P S, Santhya P, Shyne T P 1991 Handwritten character recognition — An experience.Comput. Sci. Inf. 21:Google Scholar
- Elms A J 1996The representation and recognition of text using hidden Markov models. Ph D thesis, Dept. of Electronics and Electrical Engineering, University of Surrey, Guildford, Surrey, UKGoogle Scholar
- Favata J T, Srikantan G, Srihari S N 1994 Hand printed character/digit recognition using a multiple feature/resolution philosophy.Proc. Int. Workshop on Frontiers of Handwriting Recognition, 1994 (IWFHR’94), pp. 57–64Google Scholar
- Hu J, Brown M K, Turin W 1996 HMM based on-line handwriting recognition.IEEE Trans. Pattern Anal. Mach. Intell. 18: 1039–1045CrossRefGoogle Scholar
- Hush D R, Home B 1993 Progress in supervised neural networks — What’s new since Lippman.IEEE Signal Process. Mag.: 8–39Google Scholar
- Kohonen T 1990 The self-organizing map.Proc. IEEE 78: 111–123CrossRefGoogle Scholar
- Lee S W 1996 Off-line recognition of rotally unconstrained handwritten numerals using multi layer cluster neural network.IEEE Trans. Pattern Anal. Mach. Intell. 18: 648–652CrossRefGoogle Scholar
- Lee S W, Kim C H, Tange Y Y 1996 Multi resolution recognition of unconstrained handwritten numerals with wavelet transform and multi layer cluster neural network.Pattern Recogn. 29: 1953–1961CrossRefGoogle Scholar
- Oh I S, Suen C Y 1998 Distance feature for neural network-based recognition of handwritten characters.Int. J. Document Anal. Recogn. 1: 73–88CrossRefGoogle Scholar
- Nielson R H 1990Neuro-computing (Boston, MA: Addison-Wesley)Google Scholar
- Rocho J, Pavlidis T 1994 A shape analysis mode: With application to a character recognition system.IEEE Trans. Pattern Anal. Mach. Intell. 16: 393–404CrossRefGoogle Scholar
- Srikantan G, Lam S W, Srihari S N 1996 Gradient-based contour encoding for character recognition.Pattern Recogn. 29: 1147–1160CrossRefGoogle Scholar
- Strathy N W, Suen C Y 1995 A new system for reading handwritten ZIP codes.Proc. ICDAR, pp 74–77Google Scholar
- Tappert C C, Suen C Y, Wakahare T 1990 The state of the art in on-line handwriting recognition.IEEE Trans. Pattern Anal. Mach. Intell. 12: 787–803CrossRefGoogle Scholar
- Trier O D, Jain A K, Taxt T 1996 Feature extraction methods for character recognition — A survey.Pattern Recogn. 29: 641–662CrossRefGoogle Scholar