| Flood monitoring is a complex,multi factor,great harm disaster,to achieve comprehensive,thorough,and timely monitoring is a serious challenge.As a complex observation system,the data access protocols of remote sensing data system are different and the metadata encoding rules of remote sensing data system are different,such as NASA ECHO,NOAA CLASS,USGS LANDSAT,and INPE CBERS2.These data system are isolated,divided,and collaborate with each other.In addition,the data sources of remote sensing data system are different.However,to meet the requirement of data access of heterogeneous remote sensing data system and metdata encoding,there are lots of problems in remote sensing data system,such as the following questions:1)the data access protocols are too heterogeneous to form a unified data access way;2)the metadata encoding rules are different and there is no unified metadata encoding method.In the daily monitoring,the difficulty of remote sensing data resource acquisition caused by heterogeneous data system lead to the difficulty of management caused by the encoding of metadata.According to the key technology of massive remote sensing images online mapping,the purpose of this paper is to establish a meta-model for unified access and encoding of multi-source heterogeneous remote sensing data,to achieve data access and metadata encoding of heterogeneous remote sensing data system.To propose the remote sensing image slicing storage method based on Cloud Computing and desing remote sensing metadata storage model in Hadoop Distributed File System to achieve massive remote sensing images slicing storage;to propose the remote sensing images parallel mapping method based on Cloud Computing,design the parrel mapping algorithm based on map/reduce mechanism,and design the online mapping method based on Web Processing Service to accomplish the remote sensing images efficient mapping.The research contents contains four parts:1)study the multi-source and heterogeneous Remote Sensing data system access protocol and metadata encoding rules to access protocol and enhanced enhanced earth observation metadata profile of Observations&Mearurements model;2)study the Remote Sensing images sling storage based on Cloud Computing;3)study the Remote Sensing images parallel mapping based on Cloud Computing;4)desing and realize the massive remote sensing image online mapping efficient sharing service prototype system and implement the comprehensive application demonstration of Hubei Lake Basin flood disaster monitoring.(1)Unified data access protocol and metadata encoding model of multi-source and heterogeneous remote sensing data sytemFirstly,towards the data access requirement of flood monitoring,the study proposed the classification system of remote sensing data system access protocol and metadata encoding,established the multi-source and the unified data access and encoding framework of heterogeneous remote sensing data system,and defined five basic metadata components and seven tuple metadata description structure,which lay the foundation for unified access and efficient management of multi-source heterogeneous remote sensing data system.Lastly,taking Wuhan Waterlogging Monitoring as the scene,the study conducted the various types' remote sensing data access of multi-source and heterogeneous remote sensing data system and access performance testing and comparison.(2)Remote sensing slicing storage method based on cloud computingThe study analyzed the storage requirement of remote sensing images to summarize the consideration of remote sensing images storage,design the map/reduce sling storage algorithm of remote sensing images,and design the remote sensing images and metadata storage tables in Hadoop Distributed File System.Lastly,taking Wuhan Waterlogging Monitoring as the scene,the study conducted the remote sensing images slicing storage and storage performance and comparison.(3)Remote sensing images parallel mapping method based on cloud computingThe study analyzed the remote sensing mapping requirement of parallel mapping,design the remote sensing images parallel mapping algorithm based on map/reduce mechanism.In addition,the study analysed the web service interface of Web Processing Service and proposed the Web Processing Serivce extension method for remote sensing image parallel mapping.Lastly,taking Wuhan Waterlogging Monitoring as the scene,the study conducted the massive remote sensing images parallel mapping and mapping performance and comparison.(4)Massive remote sensing image online mapping and efficient sharing service prototype system and applicationThe study designed and realized the massive remote sensing image online mapping sharing service prototype system.Taking various kinds of remote sensing data resources in vegetation monitoring of flood disaster affected areas in Hubei as an example,the remote sensing data resources of multi-source heterogeneous remote sensing data system are efficiently accessed.Towards the massive remote sensing image storage requirements,the study conducted the massive remote sensing data storage of remote sensing data resources in vegetation monitoring of flood disaster.Towards the monitoring requirement of vegetation monitoring of flood disaster in Hubei Province,the study realized the massive remote sensing images fast mapping of flood affected vegetation.So the effectiveness and feasibility of the remote sensing image based on cloud computing in this study are verified by the application demonstration in Hubei province.The innovation of this paper is mainly reflected in three aspects:1)a meta-model for unified access and encoding of multi-source heterogeneous remote sensing data system is constructed to realize the unified access of heterogeneous remote sensing data reousources;2)proposed a massive remote sensing slicing storage method based on cloud computing to promote the storage efficiency of massive remote sensing images;3)proposed a massive remote sensing images parallel mapping method based on cloud computing to promote the massive remote sensing images mapping efficiency.In conclusion,in terms of the problems of massive remote sensing images,the designed online mapping method has strong technical,practical and promotional value,relating to the remote sensing image data center,cloud computing,geographic information sharing services in the field of the latest theories and methods. |
PDF Full Text Request