结合空间位置和语义匹配的多源房屋灾害数据融合

Multi-source building disaster information fusion based on spatial and semantic matching

自然灾害综合风险普查工作中,房屋灾害信息数据来源众多,为调查底库快速建库带来很大挑战。本文针对多源房屋数据难以快速提取融合的问题,基于地理空间位置的信息匹配和基于地理空间语义的信息匹配等技术,通过空间数据提取、房屋建筑增补绘、属性信息匹配、属性信息选择,实现了房屋建筑多源信息快速自动融合,为自然灾害普查工作提供了可靠翔实的房屋建筑数据底板。实践证明,该技术成果数据准确可靠,可有效提升自然灾害风险普查工作效率,大量降低人工成本。

In the national survey on natural disaster risks,there are many sources of house disaster information data,which brings great challenges to the rapid construction of the investigation base database.In order to extract and fuse multi-source house data quickly,this paper realizes the rapid and automatic fusion and reliability evaluation of multi-source house information through the operations of basic data preprocessing,spatial data extraction and automatic fusion of attribute information by using technologies such as information matching based on geospatial location and information matching based on geospatial semantics.On this basis,the natural disaster census database building system is developed to provide reliable and detailed housing construction data for natural disaster census.Practice had proved that the platform results were accurate and reliable,which could effectively improve the efficiency of natural disaster risk survey and greatly reduce labor costs.