A Hybrid Approach to Rendering Handwritten Characters

With the growing popularity of pen-based computers comes the need to display clear handwritten characters at small sizes on low-resolution displays. This paper describes a method for automatically constructing hinted TrueType fonts from on-line handwriting data. Hints add extra information to glyph outlines in the form of imperative constraints overriding side effects of the rasterization process. We use an aggressive matching strategy to find correspondences between an input glyph and a previously-hinted template, considering both global and local features to allow hinting even when they differ in shape and topology. Recognizing that stroke width statistics are among features that characterize a person’s handwriting, we recalculate global values in the control value table (CVT) before transfer to preserve the characteristics of the original handwriting.