Street View Image Compact Structure Storage Retrieval Method Based On Road Coding

With the process of data informatizationand digitzation having accelerated,street view data emerges,as the important data resource.Street view data has many characteristics,such as scalability and the character of data mining,and it can provide the images of the fa?ade of the city’s streets.Additionally,street view data can excavate natural and humane society information.Street view data is usually stored in a folder pattern based on dates,which has many deficiencies.And traditional spatial index methods still has some problems,such as low index efficiency and low adaptability.For the above problems,this paper,on the basis of road coding,puts forward acompact structure storage model and an improvedmethod of Road Coding Index Structure-TREE(RCIS-TREE).In this paper,the method of storage data model with compact structure based on road coding is described in detail.Street view data is stored in five layer structures: road,road segment,camera number,window level,image data,and index filesare createdaccording to data structure,which forms a compact structure with data files.Moreover,in the light of storage model with road coding and R-TREE index,the paper buildsRCIS-TREE spatial index method,including specific road coding approach,the establishment of Field of View(FOV)and the structure of Minimum Bounding Titled Rectangle(MBTR),and the implement of the index method.Finally,ComparingRCIS-TREE spatial index method and R-TREE spatial index method,the results showed that the response time of R-TREE spatial index method increases with the change of data volume,which presents “S” type.Although,the response time of RCIS-TREE spatial index methodgoes up and down,it is within the range of forty-five milliseconds.Besides,the response time of RCIS-TREE spatial index method is very short and becomes horizontal gradually,but it does not increase with the rise of data volume.