传统分形图案的纹样在表现实际景物特征方面存在不足。针对这一问题,提出了一种基于纹理渲染的分形图案设计概念和技术,并改进了W e i的点匹配纹理合成算法,即首先用2维自相关函数先分析纹理的构成和分布,然后确定合成的L形邻域大小,最后将分形图案作为结构约束,并在HSL颜色空间上控制色彩渲染,实现了结构约束的纹理合成,并通过选择两个特殊的分形数集来进行实验,生成了既有样图纹理特征,又能控制色彩变化的分形图案。
This paper describes a new framework for reusing hand-drawn cartoon clips based on language understanding ap- proach. Our framework involves two stages: a preprocessing phase, in which a hand-drawn clip library with mixed architecture is constructed, and the on-line phase, in which the domain dependent language instructions parsing is carried out and clips in the clip library are matched by use of some matching values calculated from the information derived from instruction parsing. An im- portant feature of our approach is its ability to preserve the artistic quality of clips in the produced cartoon animations.
