Title :
Feature extraction circuit for optical character recognition
Author :
Caviglia, D.D. ; Valle, M. ; Rossi, Anna ; Bo, G. ; Pedrazzi, P.
Author_Institution :
Dept. of Biophys. and Electr. Eng., Genova Univ.
fDate :
5/12/1994 12:00:00 AM
Abstract :
A feature set for the encoding of optical images of handwritten characters is presented. It allows good recognition rates when coupled with a neural network classifier and can be easily implemented in analogue CMOS VLSI technology. A circuit for the extraction of such features is described
Keywords :
CMOS integrated circuits; VLSI; analogue processing circuits; backpropagation; feature extraction; image coding; image processing equipment; image recognition; neural chips; optical character recognition; OCR; analogue CMOS VLSI technology; encoding; feature extraction circuit; handwritten characters; neural network classifier; optical character recognition; optical images; recognition rates;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19940526
