项目名称: 基于语义网络的街区场景相似性研究
项目编号: No.41501498
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 天文学、地球科学
项目作者: 孟妮娜
作者单位: 长安大学
项目金额: 20万元
中文摘要: 空间场景由具有一定空间结构特征的对象集构成,具有组合性、多样性、结构性的特点,将人类对空间场景相似性评估的定性认知过程转换为定量计算是GIS领域关于地图理解及空间认知的挑战性问题。本研究以地图空间中街区场景的相似性为研究对象,在分析视觉认知规律和街区场景空间上下文环境的基础上,提出以语义网络来表示街区场景相似性判断的知识,通过构建街区场景相似性语义网络以及在此基础上的知识识别与推理,建立适合人类视觉认知规律的街区场景相似性智能化计算模型,同时为地理信息科学领域里多个研究问题的解决提供方法支持,如空间数据库的相似性检索、地理数据的匹配、多源数据的集成、制图综合等。研究的主要内容包括:街区场景视觉认知模式研究及语义分析,街区空间场景上下文环境的研究,以街区场景相似性判断为目标的语义网络构建,以及以街区场景相似性判断为目标的语义网络识别方法研究。
中文关键词: 地图理解;视觉认知;空间场景;语义网络;相似性
英文摘要: Spatial scenes comprise objects arranged in a particular structure, and which have some characteristics, such as compositionality, diversity and configurability. The challenging problem of spatial cognition in GIS domain is to translate the cognitive process of a qualitative similarity assessment into the quantitative realm. With the street block scene in spatial database as the research object, this study construct semantic network of street block scene and study the method of knowledge identification and knowledge reasoning based on the semantic network in artificial intelligence to solve the calculating and matching similarity of spatial scene. This research aims to establish the model of processing intelligently street block scene similarity which is in conformity with the human cognition and to provide the solutions of research related to the comparison of spatial scene in geographic information science, such as integration and matching of volunteered geographic information, the spatial data updating, spatial data query and retrieval, etc. . The main research contents includes: visual cognitive pattern research of street block scene and the semantic analysis, spatial context study of street block scene, semantic network building and semantic network identification method research aiming at street block scene similarity judgment.
英文关键词: Map Comprehension;Visual Cognition;Spatial Scene;Sementic Network;Similarity