The Research And Implementation Of Data Visualization System Rased On Map

With the rapid development of Internet of Things and the continuous popularization of mobile devices and sensor devices,there are more and more means of obtaining geospatial data,and we are now in the age of geograpgic big data.Data visualization can display data graphically and help users analyze data in an intuitive and effective way.Currently,although many scholars have stuied the research direction of geospatial data visualization,there are still many problems.Especially when dealing with dense data visualization issues,the existing platforms that support the visualization of geospatial data still adopt coarse-grained,low-accuracy and time-complexity methods,which cause large errors for subsequent analysis of users.In addition,each visualization platform is not perfect for storing and handling multi-source,massive and heterogeneous geospatial data and does not support real-time data visualization.These problems above need to be solved seriously,thus,this paper studies how to deal with the problem of dense data visualization and builds a geospatial data visualization system that supports multi-source,massive and heterogeneous data sources and supports real-time data visualization.The main contents of this paper are as follows:Firstly,based on the deep understanding of the characteristics of electronic map service model and geospatial point data,this paper designs and implements a grid-based and K-means algorithm point clustering algorithm to improve the accuracy of point clustering aiming at the problem of dense point data visualization.Secondly,based on the machine learning clustering algorithm,combining DBSCAN and K-means two clustering algorithms,a two-stage edge-bound algorithm is implemented.The experimental results show that the proposed algorithm can be used in the case of no significant increase in time complexity under the edge than the commonly used binding algorithm with higher accuracy.Third,this paper designs and implements a common system solution that supports multi-source heterogeneous data sources,data real-time visualization and friendly interaction.The system provides a one-stop-shop for visualization options including data source uploading,map creation,visualization of attribute configurations,visualization results and other functional modules.The system supports the point clustering algorithm and edge binding algorithm implemented in this paper.The system combines the specific application scenarios with the related technical analysis results to realize the flexible and rapid visualization application,which can efficiently and accurately display the hidden information and rules of geospatial data and improve the data accessibility and understandability.