Backgrounds This work emphasizes the current research status of the urban Digital Twins to establish an intelligent spatiotemporal framework.A Geospatial Artificial Intelligent(GeoAI)system is developed based on the G...Backgrounds This work emphasizes the current research status of the urban Digital Twins to establish an intelligent spatiotemporal framework.A Geospatial Artificial Intelligent(GeoAI)system is developed based on the Geographic Information System and Artificial Intelligence.It integrates multi-video technology and Virtual City in urban Digital Twins.Methods Besides,an improved small object detection model is proposed:YOLOv5-Pyramid,and Siamese network video tracking models,namely MPSiam and FSSiamese,are established.Finally,an experimental platform is built to verify the georeferencing correction scheme of video images.Result The MultiplyAccumulate value of MPSiam is 0.5B,and that of ResNet50-Siam is 4.5B.Besides,the model is compressed by 4.8times.The inference speed has increased by 3.3 times,reaching 83 Frames Per Second.3%of the Average Expectation Overlap is lost.Therefore,the urban Digital Twins-oriented GeoAI framework established here has excellent performance for video georeferencing and target detection problems.展开更多
This paper proposes a novel data indexing scheme,the distributed access pattern R-tree(DAPR-tree),for spatial data retrieval in a distributed computing environment.As compared to traditional distributed indexing schem...This paper proposes a novel data indexing scheme,the distributed access pattern R-tree(DAPR-tree),for spatial data retrieval in a distributed computing environment.As compared to traditional distributed indexing schemes,the DAPR-tree introduces the data access patterns during the indexing utilization stage so that a more balanced indexing structure can be provided for spatial applications(e.g.Digital Earth data warehouse).In this new indexing scheme,(a)an indexing penalty matrix is proposed by considering the balance of data number,topology and access load between different indexing nodes;(b)an‘access possibility’element is integrated to a classic‘Master-Client’structure for a distributed indexing environment;and(c)indexing algorithm for the DAPR-tree is provided for index implementations.By using a duplication of official GEOSS Clearinghouse system as a case study,the DAPR-tree was evaluated in a number of scenarios.The results show that our indexing schemes generally outperform(around 9%)traditional distributed indices with the utilization of data access patterns.Finally,we discuss the applicability of the DARP-tree and document DARP-tree shortcomings to encourage researchers pursuing related topics in Big Data indexing for Digital Earth and other geospatial initiatives.展开更多
Remote sensing satellites are playing very important roles in diverse earth observation fields.However,long revisit period,high cost and dense cloud cover have been the main limitations of satellite remote sensing for...Remote sensing satellites are playing very important roles in diverse earth observation fields.However,long revisit period,high cost and dense cloud cover have been the main limitations of satellite remote sensing for a long time.This paper introduces the novel volunteered passenger aircraft remote sensing(VPARS)concept,which can partly overcome these problems.By obtaining aerial imaging data from passengers using a portable smartphone on a passenger aircraft,it has various advantages including low cost,high revisit,dense coverage,and partial anti-cloud,which can well complement conventional remote sensing data.This paper examines the concept of VPARS and give general data processing framework of VPARS.Several cases were given to validate this processing approach.Two preliminary applications on land cover classification and economic activity monitoring validate the applicability of the VPARS data.Furthermore,we examine the issues about data maintenance,potential applications,limitations and challenges.We conclude the VPARS can benefit both scientific and industrial communities who rely on remote sensing data.展开更多
基金Supported by Key R&D Program of the Ministry of Science and Technology (2019YFC0810704)Key R&D Program of Guangdong Province (2019B111102002)Shenzhen Science and Technology Program (KCXFZ202002011007040)。
文摘Backgrounds This work emphasizes the current research status of the urban Digital Twins to establish an intelligent spatiotemporal framework.A Geospatial Artificial Intelligent(GeoAI)system is developed based on the Geographic Information System and Artificial Intelligence.It integrates multi-video technology and Virtual City in urban Digital Twins.Methods Besides,an improved small object detection model is proposed:YOLOv5-Pyramid,and Siamese network video tracking models,namely MPSiam and FSSiamese,are established.Finally,an experimental platform is built to verify the georeferencing correction scheme of video images.Result The MultiplyAccumulate value of MPSiam is 0.5B,and that of ResNet50-Siam is 4.5B.Besides,the model is compressed by 4.8times.The inference speed has increased by 3.3 times,reaching 83 Frames Per Second.3%of the Average Expectation Overlap is lost.Therefore,the urban Digital Twins-oriented GeoAI framework established here has excellent performance for video georeferencing and target detection problems.
基金funded by the National Key R&D Program of China[grant number 2018YFB2100704]Science,Technology and Innovation Commission of Shenzhen Municipality[grant numbers JCYJ20170412142239369,JCYJ20170818101704025]the National Natural Science Foundation of China[grant numbers 41701444,71961137003,41971341].
文摘This paper proposes a novel data indexing scheme,the distributed access pattern R-tree(DAPR-tree),for spatial data retrieval in a distributed computing environment.As compared to traditional distributed indexing schemes,the DAPR-tree introduces the data access patterns during the indexing utilization stage so that a more balanced indexing structure can be provided for spatial applications(e.g.Digital Earth data warehouse).In this new indexing scheme,(a)an indexing penalty matrix is proposed by considering the balance of data number,topology and access load between different indexing nodes;(b)an‘access possibility’element is integrated to a classic‘Master-Client’structure for a distributed indexing environment;and(c)indexing algorithm for the DAPR-tree is provided for index implementations.By using a duplication of official GEOSS Clearinghouse system as a case study,the DAPR-tree was evaluated in a number of scenarios.The results show that our indexing schemes generally outperform(around 9%)traditional distributed indices with the utilization of data access patterns.Finally,we discuss the applicability of the DARP-tree and document DARP-tree shortcomings to encourage researchers pursuing related topics in Big Data indexing for Digital Earth and other geospatial initiatives.
基金supported by National Natural Science Foundation of China(41974006)Shenzhen Scientific Research and Development Funding Program(KQJSCX20180328093453763,JCYJ20180305125101282,JCYJ20170412142239369)+1 种基金Open Fund of Key Laboratory of Urban Land Resources Monitoring and Simulation(KF-2018-03-004)Department of Education of Guangdong Province(2018KTSCX196).
文摘Remote sensing satellites are playing very important roles in diverse earth observation fields.However,long revisit period,high cost and dense cloud cover have been the main limitations of satellite remote sensing for a long time.This paper introduces the novel volunteered passenger aircraft remote sensing(VPARS)concept,which can partly overcome these problems.By obtaining aerial imaging data from passengers using a portable smartphone on a passenger aircraft,it has various advantages including low cost,high revisit,dense coverage,and partial anti-cloud,which can well complement conventional remote sensing data.This paper examines the concept of VPARS and give general data processing framework of VPARS.Several cases were given to validate this processing approach.Two preliminary applications on land cover classification and economic activity monitoring validate the applicability of the VPARS data.Furthermore,we examine the issues about data maintenance,potential applications,limitations and challenges.We conclude the VPARS can benefit both scientific and industrial communities who rely on remote sensing data.