【摘要】 针对CJKOutline字体在存储量上存在的不足 ,本文提出一种结构式压缩算法。算法对CJK字体进行集合变换 ,得到笔划集合元素 ;并利用聚类算法得到模板笔划 ;对相似数据进行统一存储与调用。同时 ,本文还提出了一种基于笔划段的笔划抽取算法 ,从图论角度实现了集合变换。结果显示 ,算法取得了较好的效果 ,而且适用于多种字体更多还原

【Abstract】 Because the ability to display and output high quality characters for Chinese,Japanese,and Korean(CJK)languages is still limited now,we propose a structure based compression scheme.The approach uses "Stroke Extraction"to transform CJK outline fonts from Contour Set to Stroke Set.And the strokes are clustered to find appropriate templates.Then the similar strokes are saved as templates and rebuilding information.We also present a stroke extraction algorithm based on stroke segment topologically.The results reveal that the scheme achieves a high compression ratio without reducing the output quality of the fonts, and the stroke extraction algorithm is not only efficient but also propitious to multi fonts.更多还原