PhD Defense: An Improved Representation for Stroke-Based Typefaces and a Method for their Creation

Abstract

Because a Chinese, Japanese, or Korean typeface can consist of more than 20,000 glyphs, traditional scalable outline-based typefaces require approximately 10-14 megabytes (MB) of memory. This requirement is particularly problematic in mobile devices (e.g., cell phones and PDAs) and embedded systems (e.g., car navigation systems) where memory is at a premium. Existing commercial solutions (e.g., by Bitstream and Monotype Imaging) represent glyphs using simplified uniform width strokes. However, these light-weight stroke- based fonts lack the detail, expressiveness, and variety needed for optimal legibility and true cultural acceptance.

We present an improved representation for stroke-based fonts, entitled Stylized Stroke Fonts (SSFs). SSFs overcome the shortcomings of existing scalable font representations by providing the detail and expressiveness of outline typefaces with a memory footprint comparable to current uniform width stroke typefaces. This reduced memory footprint facilitates the use of highly stylized typefaces on embedded devices and when transferring applets containing fonts over the internet. Furthermore, because creating a single CJK typeface from scratch requires multiple man-years of labor, we further present a novel method for automatically converting conventional outline typefaces to SSFs so that a collection of SSFs can be generated to match the wide variety offered by outline typefaces.