Research Of Writing Process Recurring Of Chinese Calligraphy And Its Application In Digital Library

Writing process recurring of Chinese Calligraphy is one of the most active researches in computer art application. It aims to recover stroke order and other features from static calligraphic character image and to present the recurring of writing process of calligraphic character. Recurring of Chinese calligraphic character has a wide spectrum of promising application in many areas such as virtual reality, character recognition and distance education. In recent years, research of Chinese calligraphy in the respect of computer art received increasing attention from both academic and industrial, many research about computer calligraphic and Chinese painting simulation made a rapid achievement. But the area of recurring of calligraphic writing process is still not widely focused on. One key subject of calligraphic image is the recovery of stroke order which is an important subject in this field, which not only involves special representation of Chinese character and but also involves the subject of pattern recognition.This thesis present a general framework of the writing process recurring of Chinese Calligraphy oriented to digital library. Under this framework we analyze the important modules including contour extraction, stroke order recovery and brush model, describe some novel algorithms, and draw some useful conclusions. Based on these researches the Chinese Calligraphy Writing Process Recurring system is developed and now it is put into use in the portal of CADAL. Our contributions are summarized as follows.A method of character structure analysis is proposed to analyze regional feature and orientation relation feature of calligraphic character. Based on these features the orientation relation function is used to extract structure information and we can get the writing order of each part of a character based on the normal writing rules of Chinese character.Chinese Calligraphic Model is is proposed. The model includes the features stroke's thickness extracted from contour and other features extracted from the skeleton of a callgraphic image.A new algorithm of stroke's thickness is proposed.We propose two methods to recover stroke order from Chinese calligraphic image. One is based on the Minimum Curvature, and the other is used for some special fonts of Chinese character such as running hand script, fluent script as these scripts have deformed and the number of stroke number reduced hard. For these reasons, a CCM (Chinese Calligraphic Model) is proposed and the Condensation Algorithm is used to extract order.We develop the 3D Chinese Calligraphic Handwriting Recur System which includes two modules: the Calligraphic Image Pre-processing Module and the Client Module. The Client Module is responsible for animating the writing process by communicating with the Calligraphic Data Engine to retrieve strokes order and stroke's thickness information from database. This module is an applet coded in Java with Java 3D API. Users can view the animation of calligraphic handwriting process by their browsers.