Parallelizing OCR

Author

Montesinos, Eduardo ; Kienhofer, J.

Author_Institution

Inf. for Eng. & Sci., Karlsruhe Univ., Germany

fYear

1991

fDate

27-30 Mar 1991

Firstpage

46

Lastpage

52

Abstract

The authors describe how to parallelize optical character recognition (OCR) in EFECT, a parallel dataflow-oriented graphical programming language. OCR-systems are judged on two criteria: speed and reliability. To obtain both, a simple, reliable, and fast feature extraction method has been chosen to be executed in a parallel environment. As the EFECT environment has not been completed yet, the algorithms have been implemented and tested on a personal computer, ignoring the speed up when using a parallel architecture. The main features described are: feature extraction, storing knowledge in a neural network, and implementing a parallel learning and recognition algorithm in EFECT. Isolating words and symbols from the raw data is only roughly described. Some test runs and their results are presented and discussed. The raw data were produced without a scanner using a pixel-oriented painting program. The character isolation was done interactively

Keywords

learning systems; neural nets; optical character recognition; parallel programming; EFECT; character isolation; feature extraction; neural network; parallel dataflow-oriented graphical programming language; parallel learning; parallelizing OCR; pixel-oriented painting program; reliability; speed; Character generation; Data mining; Distribution functions; Feature extraction; Informatics; Microcomputers; Neural networks; Optical character recognition software; Testing; Text recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Computers and Communications, 1991. Conference Proceedings., Tenth Annual International Phoenix Conference on

Conference_Location

Scottsdale, AZ

Print_ISBN

0-8186-2133-8

Type

conf

DOI

10.1109/PCCC.1991.113790

Filename

113790