权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Developement of Supporting System for Knowledge Mining from Multimodal Image Database Using Image Processing Techniques.

利用图像处理技术开发多模态图像数据库知识挖掘支持系统。

基本信息

批准号：
10358006
负责人：
TORIWAKI Jun-ichiro
金额：
$ 13.89万
依托单位：
NAGOYA UNIVERSITY
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (A).
财政年份：
1998
资助国家：
日本
起止时间：
1998 至 2000
项目状态：
已结题

项目摘要

Various methods were developed for extracting knowledge useful for experts and aiding for creation of novel concepts and methodologies, using a set of new type of images. Also the possibility of "a tool of knowledge finding creative new knowledge using images as a cue" was examined and studied experimentally by implementing the proposed methods as a software system.(1)Visalization We developed a method to visualize the data distributing in the 4-dimensional space. A system for flexible manipulation of 2- and 3-dimensional images seamlessly was also developed.(2)Virtual space manipulation The systems as follows were developed and tested the performance experimentally.(a)The use of virtual scissors and manipulation of a virtual paper sheet in the virtual space.(b)Virtual sculpting.(c)Virtual woodcut printing including automated generation a woodblock and multi-colar printing.(3)Educational use of virtual space manipulation The above woodcut printing was provided to pupils of elementary and middle schools for free use. The virtual endoscope system for educational use was newly developed by modifying the system we developed for medical use. Both experiments clearly showed the effectieness of the use of virtual space manipulation as the aids for creative education.(4)Tool of knowledge discovery Knowledge discovery of the "algorithm type knowledge" was realized by integrating the image processing expert system IMPRESS we developed before and the interactive indication system of a 3D figure embedded in a 3D gray images. This system KDiSPS is considered as a supporting system for knowledge discovery by human and an automated knowledge discovery system by computer.

开发了各种方法，用于提取对专家有用的知识，并使用一组新型图像帮助创建新的概念和方法。此外，通过将所提出的方法作为一个软件系统来实施，对“一种利用图像作为线索发现创造性新知识的知识工具”的可能性进行了实验研究。（1）可视化我们提出了一种将数据在四维空间中的分布可视化的方法。还开发了一种用于无缝地灵活操纵2维和3维图像的系统。（2）虚拟空间操作开发了如下系统，并通过实验测试了其性能。（a）在虚拟空间中使用虚拟剪刀和操纵虚拟纸张。（B）虚拟雕刻。（c）虚拟木刻印刷，包括自动生成木刻和多色印刷。（3）虚拟空间操作的教育用途上述木刻印刷品提供给小学和中学的学生免费使用。虚拟内窥镜系统的教育用途是新开发的修改系统，我们开发的医疗用途。这两个实验都清楚地表明了使用虚拟空间操作作为创造性教育的辅助手段的有效性。（4）知识发现工具将我们开发的图像处理专家系统IMPRESS与嵌入在三维灰度图像中的三维图形交互显示系统相结合，实现了“算法类知识”的知识发现。该系统KDiSPS被认为是人工知识发现的支持系统和计算机自动知识发现系统。