An User-feature Based Approach to Dynamically Building Personal Chinese Calligraphy Fonts
#This is a project originally created for a science project#
#Chinese Abstract# 摘要 電腦字體近來流行許多近似手寫字體的字形，像是少女體、硬筆書法體等等。這些字形都是由書法家一字字寫入電腦再編輯完成。這樣的製作方法不僅讓個人用戶無法製作自己的個人化字形，也讓專業的類手寫字體制作極為費時。結合型態學演算以及模式識別、中文字資料庫等等，我製作了利用使用者少數部件的重組，模擬該使用者書寫其他中文字的系統。此系統不僅可以用在製作個人化字形，亦可以用在文化資產的保存與研究，凡如佚漏古籍之校定及重製。 我的系統從使用者的部件擷取到字庫分析及重新組字，皆經過調校已符合中文字的特性，可利用三百個中文字部件組合出超過3500個常用中文字，涵蓋超過95%的中文字使用率。
#English Abstract# Abstract Though user-feature based font is easy to obtain in phonetic symbol languages like English, creating a similar font in Chinese is not an easy task. With over thirty thousand characters, user-feature based font is limited to only professional font designing companies with sky-high cost. My system is a whole new approach to creating personal font. While only requesting a little amount of user features, we can analyze and capture the information and components needed to simulate other characters not written by the user. My system not only provides a way for individual users to create personal font, it is also a tool to preserve Chinese calligraphy culture. When certain characters by a calligrapher are missing, my system provides the ability to reproduce these characters. After carefully manipulating algorithms and parameters, only 350+ characters are needed to produce more than 3500+ Chinese characters, covering more than 95% of everyday usage.