SKELETIZATION AND APPROXIMATION TECHNIQUES APPLIED TO IMAGES OF HANDWRITTEN SYMBOLS BASED ON THE USE OF A MULTILEVEL THRESHOLD CELLULAR-AUTOMATON

Authors
ODINETS DN TATUR MM
Citation
Dn. Odinets et Mm. Tatur, SKELETIZATION AND APPROXIMATION TECHNIQUES APPLIED TO IMAGES OF HANDWRITTEN SYMBOLS BASED ON THE USE OF A MULTILEVEL THRESHOLD CELLULAR-AUTOMATON, Automatic control and computer sciences, 32(3), 1998, pp. 25-30
Citations number
11
Categorie Soggetti
Computer Science Artificial Intelligence","Robotics & Automatic Control","Computer Science Artificial Intelligence","Robotics & Automatic Control
Journal title
Automatic control and computer sciences → ACNP
ISSN journal
01464116
Volume
32
Issue
3
Year of publication
1998
Pages
25 - 30
Database
ISI
SICI code
0146-4116(1998)32:3<25:SAATAT>2.0.ZU;2-N
Abstract
A structural-algorithm model of a cellular automaton for preliminary p rocessing of images is proposed. The process of approximating handwrit ten alphanumeric symbols is carried out oil the basis of this model. T he images which are processed may be of different dimensions and conta in lines having different slopes and thicknesses. The approximation is performed by means of segments of straight lines with discrete slope of 45 degrees. The method is based on multilevel threshold selection o f individual sectors of the images and classical cellular logic.