Visual analysis of topological features for character recognition

Klekovska, Mimoza and Martinovska Bande, Cveta (2023) Visual analysis of topological features for character recognition. International Journal of Scientific and Research Publications, 13 (6). pp. 66-72. ISSN 2250-3153

Full text not available from this repository.

Abstract

The text presented hereinafter is describing the process of extraction of several characteristics required for recognition of Old Slavic Letters. The analysis is conducted over old Slavic scripts written with Cyrillic alphabet named “Constitutional Script”. It is a unique style of writing. The contour of the graphemes has been observed from several aspects: structural, graphic–aesthetical, geometrical and topological. As a result of such analysis, we have got semiotic style descriptions, harmonic proportions as well as characteristics of the letters that are used in the recognition process. The actual letter recognition is mainly founded on pixel recognition in particular segments like spots or blemishes, presence of vertical and horizontal lines as well as determination of compactness, complexity of the contour line and symmetry of the entire letter.

Item Type: Article
Subjects: Natural sciences > Computer and information sciences
Divisions: Faculty of Computer Science
Depositing User: Cveta Martinovska Bande
Date Deposited: 25 Dec 2023 10:29
Last Modified: 25 Dec 2023 10:29
URI: https://eprints.ugd.edu.mk/id/eprint/33112

Actions (login required)

View Item View Item