基于轮廓拟合的新风格书法字合成

CALLIGRAPHIC CHARACTER SYNTHESIS BASED ON CONTOUR FITTING

限于当前有限的数字书法资源,书法数据库中有时不一定有用户所需风格的书法字图像。针对这个问题,提出了基于两个书法字轮廓合成一系列具有中间风格的新书法字的方法。首先提取同一字不同风格的两个样本字的轮廓,接着用贝塞尔曲线拟合轮廓,再把矢量化轮廓字并归一化以获取可比性;然后,通过形状匹配法找到两个书法字对应的轮廓点对,选取关键点作为控制点,合成具有中间风格的书法字矢量轮廓;最后,用填充算法填充轮廓得到一个新风格的书法字。

Limited to the current limited number of calligraphy resources, calligraphy database does not necessarily have the user needs the style of calligraphy images. To solve this problem, a new method of synthesizing a series of calligraphic characters with intermediate style based on two calligraphic outline is proposed. First, the contours of two sample words with different styles of the same word are extracted, then the contours are fitted by Bezier curve, then the vectorized contour words are normalized to obtain the comparability. Then, the contour points of the two calligraphic words are found by the shape matching method, and the key points are taken as the control points to synthesize the calligraphic vector outline with the intermediate style. Finally, fill the outline with the fill algorithm to get a new style of calligraphy.