基于图像注意力标注技术的空间GIS数据可视化研究  

Research on visualization of spatial GIS data based on image attention annotation technology

作　　者：唐桂彬[1] TANG Guibin(Yangling Vocational&Technical College,Xianyang Shanxi,712100,China)

机构地区：[1]杨凌职业技术学院,陕西咸阳712100

出　　处：《自动化与仪器仪表》2021年第11期15-18,共4页Automation & Instrumentation

基　　金：杨凌职业技术学院2019年度研究基金项目(编号:GJ19025)。

摘　　要：空间GIS数据可视化应用作为地理信息系统应用领域中一直被研究的对象,当前空间GIS数据可视化方法在过去的使用过程中可以满足人们的需求,但随着空间数据种类不断增加,当前可视化方法由于数据分类性能不足,造成可视化精度较差的问题。因此,设计基于图像注意力标注技术的空间GIS数据可视化方法。使用神经形态视觉工具箱为空间数据预处理载体,构建高斯模型完成预处理过程。根据预处理结果构建三维符号库,并在其中增设二维符号。采用多样性学习算法确定三维数据的所属种类,根据种类差异完成标注过程。至此,基于图像注意力标注技术的空间GIS数据可视化方法设计完成。经实验结果证实,图像标注可视化方法性能与使用效果均得到了提升,可使用此方法完成空间数据处理。As the object of GIS application,spatial GIS data visualization can meet people's needs in the past.However,with the increasing of spatial data types,the current visualization method is due to the lack of data classification performance,which leads to the poor visualization accuracy.Therefore,the spatial GIS data visualization method based on image attention annotation technology is designed.The neural morphology visual toolbox is used as the carrier of spatial data preprocessing,and Gaussian model is constructed to complete the preprocessing process.According to the preprocessing results,the 3D symbol library is constructed and two-dimensional symbols are added.The diversity learning algorithm is used to determine the types of 3D data,and the annotation process is completed according to the difference of types.So far,the visualization method of spatial GIS data based on image attention annotation technology is designed.The experimental results show that the performance and effect of the visualization method are improved,and the spatial data processing can be completed by using this method.

关 键 词：地理信息系统 图像注意力标注技术 可视化 显著性检测 

分 类 号：TP39[自动化与计算机技术—计算机应用技术] P283[自动化与计算机技术—计算机科学与技术]