The DSFPN, a new neural network for optical character recognition

Author

Morns, Ian Phillip ; Dlay, Satnam S.

Author_Institution

Dept. of Electr. & Electron. Eng., Newcastle upon Tyne Univ., UK

Volume

10

Issue

6

fYear

1999

fDate

11/1/1999 12:00:00 AM

Firstpage

1465

Lastpage

1473

Abstract

A new type of neural network for recognition tasks is presented. The network, called the dynamic supervised forward-propagation network (DSFPN), is based on the forward only version of the counterpropagation network (CPN). The DSFPN, trains using a supervised algorithm and can grow dynamically during training, allowing subclasses in the training data to be learnt in an unsupervised manner. It is shown to train in times comparable to the CPN while giving better classification accuracies than the popular backpropagation network. Both Fourier descriptors and wavelet descriptors are used for image preprocessing and the wavelets are proven to give a far better performance

Keywords

Fourier transforms; learning (artificial intelligence); optical character recognition; wavelet transforms; DSFPN; Fourier descriptors; classification accuracies; counterpropagation network; dynamic supervised forward-propagation network; image preprocessing; recognition tasks; supervised algorithm; wavelet descriptors; Backpropagation algorithms; Character recognition; Neck; Neural networks; Optical character recognition software; Optical computing; Optical fiber networks; Pattern recognition; Testing; Training data;

fLanguage

English

Journal_Title

Neural Networks, IEEE Transactions on

Publisher

ieee

ISSN

1045-9227

Type

jour

DOI

10.1109/72.809091

Filename

809091