基于图的书法字笔画提取研究

Graph Based Stroke Extraction for Chinese Calligraphy

中国书法艺术源远流长,随着计算机技术的发展,利用计算机进行书法识别成为可能。为使书法能够在计算机中精确识别,书法笔画的正确分割与提取必不可少。采用基于点到边界方向距离(PBOD)的笔画分割算法,对得到的分割笔画结合原图像构建无向图,根据无向图的连接区域最终组合成单独的笔画。采用200幅随机书法字图像进行实验,结果表明笔画提取的正确率达到95.35%,证明基于图的书法字笔画提取算法准确率高。该算法不仅能正确地对笔画进行提取,而且很好地解决了复杂汉字存在多个笔画交叉区域、在笔画分割后混淆相似子笔画的问题,提高了笔画提取效率。

Chinese calligraphy has a long history, and the recognition of calligraphy in the computer has always been a subject of study. In order to make calligraphy recognized accurately in the computer, the correct segmentation and extraction of calligraphic strokes is essential. In this paper, a stroke segmentation algorithm based on point-to-boundary orientation distance (PBOD) is adopted. An undirected graph is constructed on the basis of the original image, and the connected area of the undirected graph is finally combined into a single stroke. We adopt 200 random Chinese calligraphy images to the experiments. The experimental results show that the accuracy of stroke extraction can reach 95.35%, which proves that the accuracy of the graph-based algorithm is higher. It is concluded that this algorithm can not only extract strokes correctly, but also solve the problem of confusing similar sub-strokes after stroke segmentation of complex Chinese characters with multiple stroke regions, and improve the efficiency of stroke extraction.