マルチモーダル仮想空間における形状と画像の創成および知的符号化

作为在人脑中实现形象想象的工具，有“雕刻”、“泥塑”和“口头指令”。本研究主要研究了三种图形生成方法：[1]虚拟雕刻几何建模、[2]粘土几何建模、[3]语言指令几何建模，旨在构建智能化的图形图像生成与合成的人机界面。此外，作为一种基于虚拟雕刻方法的图像合成技术，[4]虚拟木刻印刷及其应用（特别是浮世绘的智能编码）也进行了讨论。它是一种通过使用虚拟雕刻系统创建的虚拟木版、虚拟纸张、虚拟墨水和虚拟裸仁的打印过程来进行图像合成的方法。虽然我们已经研究了这个基本策略，但我们通过基于物理的渲染方法研究了更多的细节，我们特别关注浮世绘的智能编码，浮世绘是日本传统的木刻印刷方法。此外，虚拟铜版印刷方法介绍了使用基于物理的方法。

As tools to realize an imagination on a figure at the brain of a human, there have been "Sculpting", "Clay Modeling", and "Verbal Instructions". In this research we studied following three methods of figure creation : [1]geometrical modeling by virtual sculpting, [2]geometrical modeling by clay, [3]geometrical modeling by verbal instructions, to aim at construction of intelligent man-machine interfaces to create and synthesis of figures and images. In addition, as a technique of image synthesis based on the virtual sculpting method, [4]virtual woodcut printing and its application (especially intelligent coding of Ukiyo-e), is discussed. It is an image synthesis method by printing procedure using virtual woodblocks created by the virtual sculpting system, virtual paper, virtual ink, and virtual baren. Although this basic strategy had already been studied by us, we studied more details by a physics-based rendering approach, and we especially focused on intelligent coding of Ukiyo-e, a traditional woodblock printing method in Japan. In addition, virtual copperplate printing method is introduced by using a physics-based approach.