基于统计模型的书法笔画轮廓恢复*

获取书法汉字笔画的完整轮廓是进行计算机书法模拟和创作的重要前提,而笔画间的相互覆盖使得获取的笔画轮廓不完整。结合一个已有的统计模型,提出了一种新的笔画轮廓恢复方法。该方法通过建立轮廓模型、统计学习相似笔画来实现有缺失笔画轮廓的恢复。其主要过程包括笔画轮廓建模、建立并训练统计模型、运用统计模型恢复轮廓。实验表明该方法既有广泛的适应性又呈现良好的效果。

Stroke contour recovery based on statistical model for Chinese calligraphy

It is an important premise to extract the stroke contour from the Chinese calligraphy character for using computer to simulate the action of calligraphy learning and creating. However, the overlapping of strokes makes the extracted contour incomplete. This paper introduced a stroke contour recovery method based on an existing statistical model. The process included stroke contour modeling, building and training the statistical model, and using the statistical model to recover the missing part of the contour. The experiments show that this method not only has a wide adaptability but also presents good results.