摘要
乱针绣是一种极具艺术和文化价值的工艺产品,计算机辅助乱针绣制作对这一文化遗产的传承和发展具有重要意义.文中将计算机辅助乱针绣制作技术归结为绣线颜色子集选择和交叉针迹参数计算两大关键技术问题.首先,提出了基于遗传算法的绣线颜色子集优化选择方法,通过对标准交叉算子和变异算子的改进实现了绣机绣线颜色种类与绣品视觉效果间的有效折衷.其次,设计了计算机辅助乱针绣制作铺色和精绣两阶段针迹生成策略,并分别提出了一种基于图像亮度梯度的铺色交叉针迹参数算法及一种基于试探和混色原理的精绣交叉针迹参数计算方法,降低了机绣少量颜色而造成绣品与原图像间的色差.最后,给出了一个完整的计算机辅助乱针绣制作处理流程,并以实验验证了文中提出方法的有效性,为计算机辅助乱针绣制作技术的进一步研究奠定了基础.
The irregular needling embroidery is a kind of Chinese craftwork that has cultural value and artistic value; it is significant for its inheritance and evolution to aid artworks by computers. In this paper, the technology of computer-aided irregular broidering is resolved into two core techniques, the best color subset selection and the cross-stitches parameters computation. Firstly, a method of the best color subset selection from available colors is presented based on the genetic algorithm, where the crossover and mutation operators are improved to make a effective tradeoff between the little color types and the better visual effects. Then, a two phases strategy of cross-stitches generation, named color filling and fine sewing respectively, is designed. Two cross-stitches parameters computation algorithms are introduced, according to the gradient of the luminance image for color filling and based on trial-and-error and the spatial color mixing principle for fine sewing. This abates the chromatic difference between the original photo and embroidery caused by little color types of the broidering machine. Finally, a framework of computer-aided irregular broidering is described and the experimental results prove both effective and efficient of the proposed methods. The idea proposed in this paper draws especially a novel attention for computer-aided broidering.
出处
《计算机学报》
EI
CSCD
北大核心
2011年第3期526-538,共13页
Chinese Journal of Computers
基金
国家"八六三"高技术研究发展计划项目基金(2007AA01Z334)
国家自然科学基金(69903006
60373065
60721002)
教育部新世纪优秀人才资助计划(NCET-04-04605)
江苏省自然科学基金(BK2009230
BK2010375)
江苏省科技支撑(工业)计划(BE2010072)资助~~
关键词
计算机辅助刺绣
乱针绣
交叉针迹计算
遗传算法
computer-aided broidering
irregular needling embroidery
cross-stitches computation
genetic algorithm