Augmented multi-layer perceptron for rotation- and scale-invariant hand-written numeral recognition

Author

Kageyu, Satoshi ; Ohnishi, Noboru ; Sugie, Noboru

Author_Institution

Dept. of Electr. Eng., Nagoya Univ., Japan

fYear

1991

fDate

18-21 Nov 1991

Firstpage

54

Abstract

An OCR system that can recognize hand-written numerals regardless of changes in rotation and scale is proposed. The system consists of two phases. In the first phase, a binary input image is transformed with complex-log mapping followed by the Fourier transform into a rotation- and scale-invariant image. Then the transformed image is fed into a multi-layer neural network, the weights of which are modified by the error-backpropagation algorithm to absorb slight shape distortions. The system was implemented and tested using hand-written numerals. High recognition rates of 90 to 95% were obtained. A method for improving performance is also suggested

Keywords

Fourier transforms; neural nets; optical character recognition; Fourier transform; OCR system; augmented multi-layer perceptron; binary input image; error-backpropagation algorithm; multi-layer neural network; rotation invariant recognition; scale-invariant hand-written numeral recognition; slight shape distortions; Character recognition; Feature extraction; Fourier transforms; Multi-layer neural network; Multilayer perceptrons; Optical character recognition software; Pattern recognition; Pixel; Robot control; Shape;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1991. 1991 IEEE International Joint Conference on

Print_ISBN

0-7803-0227-3

Type

conf

DOI

10.1109/IJCNN.1991.170381

Filename

170381