| With the significant changes in arable land area in recent years,online change detection of remote sensing images has become a key technology and research hotspot for identifying changes in arable land area.However,for the real-time dynamic update of cultivated area,coarse-grained online services can no longer meet the ever-changing user needs.With the gradual popularisation of microservices technology,some microservices platforms provide basic and common services for users to process online.But in the face of complex transformations in arable areas and endless algorithmic models,simple atomic services and coarse-grained services can no longer meet the ever-changing needs of users.Therefore,it is an important topic of current research to study how to divide coarse-grained services into microservices that meet the needs and to reasonably orchestrate microservices through specific combination mechanisms to solve complex application problems such as online arable land change detection.In this thesis,based on the research on online farmland change detection calculation,we start from the idea of microservice architecture,reasonably split the change detection services,encapsulate and extend the implementation with the OGC WPS standard,realise the graph planning service combination modelling,propose a data-driven optimal service chain query mechanism based on the study area data,and design and develop a prototype system for validation.The specific research work in this thesis is as follows:(1)Analyze the current research status of online calculation methods for farmland change monitoring and service combination under microservice architecture,analyze the difference between microservice architecture and traditional architecture,summarize the characteristics of microservice architecture,and propose key issues for online calculation of farmland change detection in combination with the characteristics of farmland change monitoring services and user requirements.The current microservice classification scheme is analysed,and a microservice classification index is proposed for the high data transmission characteristics of the farmland change monitoring service,and coarse-grained services are classified,compared with the traditional classification scheme for verification,and the classified microservices are encapsulated and extended with the OGC WPS standard.(2)On the basis of microservice partitioning for cultivation change detection,a combination of service modelling for intra-domain graph planning is designed,Qo S pre-estimation of microservices is implemented based on data from the study area,and a data-driven optimal service chain query mechanism is proposed.The Qo S values of microservices are used as weights and user requirements are encapsulated into JSON format for communication,and the optimal service chain to meet the requirements is constructed by combining the shortest path query idea under JSON data drive.(3)With the data-driven optimal service chain query mechanism as the theoretical foundation,and with the help of Spring Cloud framework,front-end visualization library and other related development technologies,the application analysis system for monitoring farmland changes was designed and developed,and integrated into the current project’s big data research and management platform for smart agriculture.Through testing and concrete experiments,the timeliness and applicability of the above methods were verified,and it was determined that the system could improve research efficiency and meet user needs.There are 33 figures,16 tables,and 87 references in this thesis. |
PDF Full Text Request