针对基于物理的水墨扩散模型计算复杂、效率低、难以实现图像的水墨画自动转换等问题,提出一种基于图像的水墨笔画合成方法.将水墨笔画分为基本笔画区、主扩散区和不规则边界区.首先从基本笔画区的边缘向四周等距离放大形成主扩散区,然后根据水墨的含水量和纸张特性合成不规则的扩散边界;为了表现笔画的粗糙效果,通过画家手绘的笔触样本统计颜色变化误差,将误差叠加到笔画上.最后,给出了该方法在水墨画渲染中的应用实例和应用方法.实验结果表明,该方法效率高、合成效果好,并具有良好的时间连续性,可应用于交互式绘画系统、图像的水墨画风格转换以及三维水墨画的渲染.
Physical-based models of ink diffusion are complex,inefficient and difficult to be applied to artistic image transformations.In the paper,an efficient image-based synthesis method for Chinese wash-ink strokes is proposed.A wash-ink stroke is divided into three regions:primary stroke region,diffused region,and irregular boundary region.The diffused region is generated by magnifying the outline of the primary stroke region outward with the same distance.Then the irregular stroke border is created based on the percentage of water of ink and paper properties.To simulate pigment granulation in strokes,we make the statistics in varieties of pixel luminance using hand-painted brush strokes and add the variety errors on the synthesized stroke images.Finally,some applications of using our method to render Chinese wash-ink paintings are introduced.The rendering results show that our method is efficient to synthesize realistic wash-ink brush strokes,and can be used in several fields,such as interactive painting systems,image transformations to wash-ink styles,and wash-ink rendering of three-dimensional scenes.