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

Hierarchical methods for universal image coding and intelligent database systems

通用图像编码和智能数据库系统的分层方法

基本信息

批准号：
05402042
负责人：
YASUDA Yasuhiko
金额：
$ 15.42万
依托单位：
Waseda University
依托单位国家：
日本
项目类别：
Grant-in-Aid for General Scientific Research (A)
财政年份：
1993
资助国家：
日本
起止时间：
1993 至 1995
项目状态：
已结题

来源：
https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-05402042/
关键词：
Universal image coding Hierarchical image coding Sub-band coding Distributed image database B-ISDN Data cache 3D image Shape from motion WWW MPEG2 シーンチェンジ重畳直交変換階層的画像構造フィルタバンク画像縮小方式

项目摘要

As represented by the Internet with WWW (World Wide Web), multimedia network systems or services are recently growing very rapidly and they are expected to become one of the most important infrastructures in the coming society of the 21st century. However, there are many problems to be solved to realize this expectation even from the viewpoint of system structure. In fact, as WWW systems have become into world wide practical use, the data traffic of the Internet has become heavily congested. B-ISDN is expected as a solution of the Internet traffic jam. This expectation may be doubtful, because it is anticipated that the number of information providers and data quantity which each of them generates increase as well as the number of information consumers according to the increase of network bit-rates. Therefore, some skillful techniques, which utilize properties of data, such as images and voices, are needed to reduce congestion problems of such networks as Internet.Computer vision techn … More iques address the reconstruction of the three dimentional (3D) world to make navigation and scene recognition possible. Users in 3D worlds, such as virtual reality systems, need 3D information for visual communication. Even in the Internet environment, new protocols such as VRML (Virtual Reality Modeling Language) have been proposed for 3D virtual reality communication on the World Wide Web system.In this project, we proposed a new structure of image database systems featuring storage style imitating the human memory mechanism, which reduces required database memory capacity as well as generated traffic and time needed for retrieval. Thus, the proposed scheme, which was named "Image Database System Featuring Graceful Oblivion, " would be able to contribute to construction of large scale multimedia systems by reducing the overall system cost including those of database and transmission.Furthermore, we proposed a method to reconstruct dynamic objects using a coaxial camera system. This method avoids the problems of pixel correspondence that plague stereo analysis because the optical axs of the cameras are precisely adjusted. We developed a coaxial camera system that consists of one moving and one static camera, with the optical axs of both having the same orientation and adjusted using an optical system with true and half mirrors. Less

以Internet和WWW （World Wide Web）为代表的多媒体网络系统或服务近年来发展迅速，有望成为21世纪即将到来的社会中最重要的基础设施之一。然而，即使从制度结构的角度来看，实现这一期望也有许多问题需要解决。事实上，随着万维网系统在世界范围内的实际应用，因特网的数据流量已经变得非常拥挤。B-ISDN有望成为解决网络拥堵的一种方法。这种预期可能值得怀疑，因为预计随着网络比特率的提高，信息提供者的数量和每个信息提供者产生的数据量都会增加，信息消费者的数量也会增加。因此，需要一些巧妙的技术，利用数据的属性，如图像和声音，来减少像互联网这样的网络的拥塞问题。计算机视觉技术…更多的技术解决了三维世界的重建，使导航和场景识别成为可能。用户在三维世界，如虚拟现实系统，需要三维信息进行视觉交流。即使在互联网环境下，为了在万维网系统上进行三维虚拟现实通信，也提出了新的协议，如VRML（虚拟现实建模语言）。在这个项目中，我们提出了一种新的图像数据库系统结构，其存储方式模仿人类的记忆机制，减少了所需的数据库内存容量，以及产生的流量和检索所需的时间。因此，该方案被命名为“优雅遗忘的图像数据库系统”，可以通过降低包括数据库和传输在内的整体系统成本，为大型多媒体系统的建设做出贡献。此外，我们还提出了一种利用同轴相机系统重建动态目标的方法。该方法避免了困扰立体分析的像素对应问题，因为相机的光学轴是精确调整的。我们开发了一种同轴相机系统，由一个移动相机和一个静态相机组成，两者的光学轴具有相同的方向，并使用带真镜和半镜的光学系统进行调整。少