Skip to main content
notice

Doctoral Seminar Jehan Janbi

November 27, 2014
|


Speaker: Jehan Janbi
Supervisor: Dr. C. Y. Suen
Supervisory Committee: Drs. N. Dumont, P. Grogono, T. Kasvand
Title: Arabic Fonts’ Similarity Analysis Based on Visual Characteristics
Date: Thursday, November 27, 2014
Time: 10:15
Place: EV 3.309

ABSTRACT

In desktop publishing design, a font is an essential component of each design. Font file contains several information that facilitates font manipulating such as font name, designer or date of issue. For design choices, the visual properties of a font are more meaningful than file properties and information. Therefore, in Latin fonts, to provide an option to deal with font based on its visual properties, PANOSE number is added to its information. It composed of 10 digits where each digit represents a font’s visual characteristics. To extend this system to cover Arabic fonts, similarity analysis for Naskh style fonts was conducted. The analysis has been conducted within two levels on 112 fonts, using 16 glyphs that contain all strokes that used to compose all Arabic characters. The first analysis was over font’s level to cluster them into groups in such a way that fonts in each group are similar to each other. While the second analysis has been done on a glyph level to determine which glyphs have more information about font design. The similarity matrixes for both analyses had been calculated by using font matching tools that support Unicode fonts to include Arabic characters




Back to top

© Concordia University