A Study on the environmental database and eits information retrieval system using meaning data of environment

在重视与遥感和环境信息的关系的同时，环境信息知识库的建设尤为重要。我们把环境信息数据的语义描述作为研究的主要对象，旨在建立使普通人能够访问环境数据库的环境数据库，这是我们的基础性研究。本研究以九州基多市的环境为例进行了具体的研究。《基塔九州市环境数据》报告每年正式发布一次。在本报告的基础上，制作了环境知识数据库系统的原型作为实验。具体研究内容如下：(1)陆地卫星影像数据分析(2)自然语言意义描述(3)融合交互式环境信息检索系统和虚拟现实(1)发展了多时相影像分类和大规模交互式影像分析。为了对遥感图像进行分析，提出了一种基于马尔可夫随机场模型的分割方法。对进入环境信息系统的自然语言的语义表达进行了研究。(3)构建了以SD-Form为语义描述语言的环境信息系统原型。

While it pays attention to the relations with remote sensing and the environmental information, the construction of the knowledge data-base which had environmental information is important. We put the principal object on describing environmental information data semantically, and it is the fundamental research that it aims at the building of "the environmental data base which makes access to the environmental database possible to the general person. The environment of Kita Kyushu City was featured in this research as a concrete instance. The report of "the environmental data of Kita Kyushu City" is published officially every year. The prototype of the environmental knowledge data base system was manufactured as an experiment based on this report. Concrete research items were as the following.(1) Analysis of landsat image data(2) Description of the meaning in natural language(3) Merging interactive environmental information retrieval system and virtual realityMulti-temporal image classification and large-scale interactive image analysis were developed about (1). In order to analyze remotely sensed image, the new segmentation method which employed MRF model was developed. The meaning expression to analyze the natural language entered into environmental information system was studied about (2). The prototype of the environmental information system which uses the SD-Form as meaning description language was constructed about (3).