语义关联的地理视频数据自适应组织方法

项目名称： 语义关联的地理视频数据自适应组织方法

项目编号： No.41471332

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 天文学、地球科学

项目作者： 周艳

作者单位： 电子科技大学

项目金额： 90万元

中文摘要： 地理视频(GeoVideo)作为虚拟地理环境和智慧城市与城市安全十分重要的关键数据，是新一代GIS系统- - 视频GIS管理的主要空间数据类型。现有GIS系统中的地理视频主要作为多媒体属性数据进行分散独立存储管理和可视化应用，缺乏全局关联表达能力，难以支持地理视频数据的深度挖掘和关联分析，急需地理视频数据组织的新方法。本项目系统研究地理视频数据高效组织的关键问题，主要内容包括：地理视频语义关联模型、语义约束的地理视频数据自适应聚合方法、地理视频多层次语义关联索引等，并以地理视频数据公共安全应用案例进行实验验证。本研究将宏观的地理语义与微观的视频内容语义相结合，实现语义关联的地理视频数据自适应组织，为提高地理视频数据时空关联分析与计算能力、支持多维关联约束的智能高效检索提供有效途径，丰富和发展视频GIS基础理论方法。

中文关键词： 地理视频；自适应组织；语义关联；时空索引；视频GIS

英文摘要： As the key data of virtual geographic environment, smart city and urban safety, GeoVideo is a very important spatial data type managed by the new generation of GIS - VideoGIS. In existing GIS systems, GeoVideo is mainly stored as a kind of multimedia attribute data, and is mostly used for visualization applications of multimedia attributes retrieve. As lack of a spatiotemporal associated Geovideo organization, it is difficult to support data mining and data association analysis based on GeoVideo data in existing GIS. therefore a new GeoVideo data organization method is urgently needed. The project studies key issues on efficient organization of GeoVideo data, the main research contents include GeoVideo semantic association model, semantic constrained adaptive aggregation method of GeoVideo and a multi-level semantic association spatial index on GeoVideo. A public safety application case is given as experimental verification based on GeoVideo data associated analysis. This research focuses on semantic associated GeoVideo data adaptive organization by combining macro-semantics of geographic space and micro-semantics of video content, which provides an efficient way to improve the capabilities of Geovideo data spatiotemporal associated analysis and calculation. The results of research could support multi-dimensional associations constrained GeoVideo data intelligent retrieve, enrich and develop the basic theory and approach on VideoGIS.

英文关键词： Geovideo;self-adaptive organization;semantics association;spatiotemporal index;VideoGIS