An Advanced Scheme of Information Presentation Based on Integration and Conversion of Information and its Application to Recognition and Understand of Concepts

基于信息整合和转换的高级信息表示方案及其在概念识别和理解中的应用

• 批准号: 10558051
• 负责人: KITAHASHI Tadahiro
• 金额: $ 6.4万
• 依托单位: Osaka University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (B)
• 财政年份: 1998
• 资助国家: 日本
• 起止时间: 1998 至 1999
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-10558051/
• 关键词: media information processing; media conversion; media integration; geographical information processing system (GIS); document image processing; image coding; navigation system; continuous media processing; 機械部品交換; 3次元構造物内案内システム; メディア交換; 書体・字体・字

In this project, the concept of the media conversion that we have previously proposed is applied to several materials. The most meaningful results are as follows. One is the conversion of a route guidance in Japanese to a visual representation on a map. It is a rare case of successful example of the media conversion of linguistic media to image one that is the most difficult conversion among the various kinds of them. Besides, an interesting example of the image to symbolic media conversion is developed in terms of coding of academic document images into text files. The technique is expected to provide a significant information compression of images, but was not successfully attained because of the lack of the accuracy rate in recognition of characters and also in extracting paraphrase, and mathematical formulas' regions, tables and so on in documents. However, some significant results are obtained as byproducts of the development in terms of type-face identification and of distinguishing between European alphabets and Japanese ones, and they have attained satisfactory accuracy.A dialogue approach adopted in recognition of logical structure of advertising documents of products is also a unique and fruitful trial in this area.A couple of navigation systems in the field of mechanical parts exchange and personal guidance in buildings can be viewed as applications of another aspect of media processing, that is, the concept of media integration.

在这个项目中，我们以前提出的媒体转换的概念被应用到几种材料。最有意义的结果如下。一个是将日语的路线导航转换为地图上的视觉表示。语言媒体到图像媒体的转换是语言媒体到图像媒体转换中最难的一种，这是一个罕见的成功案例。此外，一个有趣的例子，图像到符号媒体转换的学术文档图像编码成文本文件。该技术有望为图像提供有效的信息压缩，但由于在文字识别、释义提取、数学公式区域、表格提取等方面的准确率不足而未能成功实现。然而，在字体识别和区分欧洲字母和日本字母方面，作为发展的副产品，获得了一些重要的结果，本文还介绍了一种基于对话的产品广告文本逻辑结构识别方法，并给出了在机械零件交换和个人信息交换领域中的导航系统建筑物中的引导可以被视为媒体处理的另一个方面，即媒体整合概念的应用。

项目成果

Fumihisa SHIBATA: "Concept Learning in Linguistic Navigation of Robots"Proc. of The 9th European-Japanese Conf. on Inf. Modelling and Knowledge Bases. 347-358 (1999)

Fumihisa Shibata：“机器人语言导航中的概念学习”Proc。

Fumihisa SHIBATA: "Mobile Robot Navigation Considering Ambiguities in Landmark Recognition"Journal of the Robotics Society of Japan. Vol.17, No.7. 993-1000 (1999)

Fumihisa Shibata：“考虑地标识别中的模糊性的移动机器人导航”日本机器人学会杂志。

柴田史久: "道標認識の曖昧性を考慮した移動ロボットのナビゲーション"日本ロボット学会誌. 17・7. 993-1000 (1999)

Fumihisa Shibata：“考虑路标识别模糊性的移动机器人的导航”日本机器人学会杂志 17・7（1999）。

Wei MING: "Font Identification by Adaptively Modeling of Characters"Proc. of Third IAPR Workshop on Document Analysis Systems DAS'98. 193-196 (1998)

明伟：“通过字符的自适应建模进行字体识别”Proc。

F. Shibata et al.: "Concept Learning in Linguistic Navigation of Robots"Proc. The 9th European-Japanese Conference on Information Modelling and Knowledge Bases. 347-358 (1999)

F. Shibata 等人：“机器人语言导航中的概念学习”Proc。