遥感大数据研究现状与发展趋势

发布时间：2018-07-21 19:44

【摘要】：目的遥感数据空间分辨率、时间分辨率、光谱分辨率以及辐射分辨率不断提高,数据类型也不断增加,从航天、航空、临近空间等遥感平台所获取的遥感数据量急剧增加,遥感数据已经具有明显的大数据特征。本文旨在从系统应用的角度分析遥感大数据处理中涉及的关键技术与问题,为相关研究人员提供有价值的参考。方法在参考大量文献的基础上,首先阐明遥感大数据的特点。其次,从GPU硬件加速、集群、网格、云计算、云格、复杂高性能计算等角度介绍了遥感大数据处理系统。再次,从分布式集群化存储技术等,分析了遥感大数据处理的关键技术。最后,从遥感大数据的多类不确定性、信息融合、机器学习、分析平台等出发,说明了目前研究存在的问题;从遥感大数据多类不确定性建模,面向遥感大数据的机器学习方法等角度说明了遥感大数据发展的趋势。结果本文详细梳理了遥感大数据的特点、典型的处理系统、核心技术,力图总结出在实际应用与学术研究中该领域需要解决的关键问题以及未来的发展趋势。结论大数据技术为遥感数据挖掘与知识获取带来了机遇与挑战,面向大数据的机器学习、数据统一分析框架、面向大数据的信息深度融合等问题的突破,将促进遥感知识挖掘的进一步发展。
[Abstract]:Objective the spatial resolution, temporal resolution, spectral resolution and radiative resolution of remote sensing data are increasing, and the data types are also increasing. The amount of remote sensing data obtained from remote sensing platforms such as spaceflight, aviation and adjacent space is increasing rapidly. Remote sensing data already have obvious big data characteristics. The purpose of this paper is to analyze the key technologies and problems involved in remote sensing big data processing from the point of view of system application, and to provide valuable reference for relevant researchers. Methods based on a large number of references, the characteristics of remote sensing big data were first expounded. Secondly, the remote sensing big data processing system is introduced from the aspects of GPU hardware acceleration, cluster, grid, cloud computing, cloud computing and complex high performance computing. Thirdly, the key technologies of remote sensing big data processing are analyzed from distributed cluster storage technology. Finally, from the view of multi-class uncertainty, information fusion, machine learning and analysis platform of remote sensing big data, the existing problems are explained. The development trend of remote sensing big data is explained from the point of view of machine learning method for remote sensing big data. Results in this paper, the characteristics, typical processing systems and core technologies of remote sensing big data are summarized in detail, and the key problems to be solved in the field of practical application and academic research and the development trend in the future are summarized. Conclusion big data technology has brought opportunities and challenges for remote sensing data mining and knowledge acquisition. It is a breakthrough in big data oriented machine learning, unified data analysis framework, and big data oriented information depth fusion. It will promote the further development of remote sensing knowledge mining.
【作者单位】：河南财经政法大学数学与信息科学学院;华中科技大学软件学院;武汉理工大学理学院统计系;武汉大学测绘遥感信息工程国家重点实验室;武汉科技大学城市学院实验实训中心;
【基金】：河南省自然科学基金项目(15A110011,14B110037) 中央高校基本科研业务费专项基金项目(2015IVA067)~~
【分类号】：TP79

【参考文献】