In order to solve the problem of delayed rescue due to untimely alarm,inaccurate alarm positioning,and biased alarm narrative,the paper designs a We Chat-based emergency alarm system for highway rescue.Based on the spatio-temporal properties of data points and considering the information value contained in the multimedia properties,a point aggregation scheme based on multimedia data features is proposed.Further,in order to solve the problems of large amount of alarm data,diverse data storage methods and unified data structure,middleware and JSON technology are introduced in data integration,and a JSON-based middleware model for data integration is designed.The main tasks of the paper are as follows.(1)An algorithm for point aggregation based on multimedia data features is proposed.Based on the comprehensive and accurate information contained in the multimedia data,to address the actual problems of repeated alarms and false alarms in the alarm process,considering the time of occurrence of data points and the state at the time of occurrence,design an aggregated calculation scheme.The algorithm integrates the geographic location information,multimedia features and time information of the data points to achieve the effective aggregation of alarm points,which in turn visualizes the alarm data and reduces the misjudgment and resource scheduling waste caused by the repeated display of alarm data by the police officers.(2)A JSON-based data integration middleware was designed.The JSON data exchange format is introduced in the data integration middleware for highway alarm data volume,data source heterogeneity and other problems;JSON-LD ontology description is used to eliminate the problem of data semantic heterogeneity;at the same time,the query rewrite algorithm Mini Con and select replicator are added to the middleware model to establish a virtual view for query requests with large commonalities,improving the operational efficiency of the system.(3)Design and implementation of We Chat-based highway emergency response system.The system is mainly designed as a one-button alarm module,multimedia alarm acquisition module,alarm details visualization module,event processing and other modules to complete the unified whole of the emergency rescue report and police.The front-end subsystem is developed using We Chat Web developer tools,and the back-end subsystem is implemented in Java language.In summary,the paper combines the idea of spontaneous geographic information with the data collected by drivers or pedestrians as a data source to construct a "driver/passengerhighway-traffic police" scheme.The system is built to provide more comprehensive information for aggregation computation through the design of point aggregation algorithm based on multimedia data features,which solves the problem of alarm display disorder at the receiving end;the JSON-based data integration middleware is applied to eliminate data semantic and syntactic heterogeneity and improve system query and operation efficiency.The design and development of the system realizes the efficient connection between the highway emergency and police officers,and provides better technical support for the command and dispatch platform. |