Abstract
In this paper, an optical character recognition system for hand-written rotated digits in land registry maps is presented. It is based on a neural network and trained by a constructive learning rule, the Hyperbox Perception Cascade (HPC). The HPC classifier can design complex, possibly nonconvex, disjoint, and bounded decision regions and treat the rejection problems of outliers and unanticipated patterns, which would otherwise tend to be classified positively in an incorrect class. We use “shape features” and a novel approach to select the most promising features to attain a low generalization error. The numerous experiments show that a subset of 24 of the 46 features obtains a good classifier with a high rate of correct classification and a low rate of rejection.
Similar content being viewed by others
References
Anderson TW (1984) An introduction to multivariate statistical analysis. Wiley, New York
Bianchini P, Ferragina P, Granieri Notturno M, Tarricone L (1993) New techniques for speech understanding. IEEE Conference on Acoustics, Speech and Signal Processing, 2: 127–130
Boatto L, Consorti V, Del Buono M, Di Zenzo S et al. (1992) An interpretation system for land register maps. IEEE Computer, 25: 23–33
Brown RM, Fay TH, Walker CL (1988) Handprinted symbol recognition system. Patt Recogn, 21: 91–118
Burgess N, Granieri Notturno M, Patarnello S (1992) 3D Object classification: application of a constructive algorithm. Int J Neural Sys2: 275–282
Di Zenzo S, Del Buono M, Meucci M, Spirito A (1992) Optical recognition of hand-printed characters of any size position, and orientation. IBM J Res Devel 36: 487–500
Di Zenzo S, Burgess N, Ferragina P, Granieri Notturno M (1993) Recognition by constructive neural algorithms. Patt Recogn Lett 14: 997–1007
Fahlman SE, Lebiere C (1990) The cascade-correlation learning architecture. In: Kaufman, Morgan (eds) Advances in neural information processing system II
Frean M (1990) The upstart algorithm: A method for constructing and training feed-forward neural networks. Neural Computation 2: 198–209
Guyon I, Vapnick V, Boser B, Bottou L, Solla SA (1992) Structural risk minimization for character recognition. N.I.P.S. 91
Jain AK, Bhattacharjee S (1992) Text segmentation using gabor filters for automatic document processing. Machine Vision Appl 5: 169–184
Kohonen T, Chrisley TR, Barna R (1988) Statistical pattern recognition with neural networks: benchmarking studies. Neural networks: from models to applications
Le Cun Y, Boser B, Denker JS, Henderson D, Howard RE, Hubbard W, Jackel LD (1990) Back-propagation applied to hand-written zip-code recognition. Neural Comput 4
Lippman RP, Huang WY (1987) Comparison between neural nets and conventional classifiers. Proceedings of the 1st IEEE International Conference on Neural Networks
Minsky M, Papert S (1969) Perceptrons. MIT Press
Rumheiart DE, Hinton GE, Williams RJ (1986) Learning internal representation by error propagation. In: Rumheiart DE, McClelland J (eds) Parallel distributed processing, MIT Press
Sirat JA (1991) A fast neural algorithm for principal component analysis and singular value decomposition. Int J Neural Syst 2: 147–155
Smith RW (1987) Computer processing of line images: a survey. Patt Recogn 20: 7–15
Vapnick V (1992) Principles of risk minimization for learning theory. N.I.P.S. 91
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Calabró, C., Ferragina, P. & Granieri, M.N. Recognition of hand-written rotated digits by neural networks. Machine Vis. Apps. 8, 351–357 (1995). https://doi.org/10.1007/BF01211497
Issue Date:
DOI: https://doi.org/10.1007/BF01211497