Research On OpenStreetMap Based Map Matching Algorithm And Its Implementation

Application-based mapping services are being developed continuously,such as peripheral search,navigation applications,etc.which are the specific performance of the map service.However,the map service can not be better developed for the map service being monopolized by a small number of large companies.Therefore,in the use of navigation,we often have to endure some troubles caused by mismatching.There are obvious differences in the process of Baidu and Gaode navigation,and there are some problems in the research of data source acquisition and research system,both for researchers and users,based on map matching service.In order to solve the above problems,this paper studied some basic map matching algorithms,and proposed a weight-based map matching algorithm which can achieve better matching results under the condition of poor GNSS signal quality.The traditional map matching algorithm is not comprehensive enough to consider the calculation factor in the matching,and the weight distribution is improper,which makes the results of the map matching one-sided.In addition to the proposed map matching algorithm,a map matching system was designed and implemented.The system uses some of the existing open source technology.It not only provides the road network data,but also a platform for map matching.So the optimization process of map matching algorithm is promoted.The specific work content is provided as follows:First,a weighted composite map matching algorithm was proposed.For the determination of the candidate range,in addition to the first grid partition,the radius of the circular confidence region of radius R is divided again.For the calculation of the weight of candidate road,in addition to the perpendicular distance and direction information,it also integrates the average distance parameter,segment belonging parameter and perpendicular parameter.Then for the possible mismatch which caused by the higher influence of belong parameters on the initial total weight,the algorithm was verified again to determine the optimal matching path.This algorithm not only inherits the efficiency of the original algorithm,but also improves the matching precision.Second,in view of the current lack of map matching system,this paper designs and implements a map matching system.Getting open source map data from OSM,and parsed it by OSMAPI.In order to reduce the amount of computation in the process of map matching,we divided the parsed road network data by geo-hash.After that we stored the road network information in the POSTGRESQL database,so that we can use it in the future.For the map display module,using the mapbox to render and the maperitive to slice,it shows the results of the map matching better.Third,analysis and summary experiment.According to the designed system,we implemented the map matching algorithm using Beijing floating car GPS data provided by the Data Hall,and the GPS trajectory data downloaded from Hong Kong University Science & Technology official website.The experimental result indicates that the algorithm can get better matching results under the condition of satisfying the real time,whether it is a high sampling frequency or low sampling frequency,and whether it is normal or crossing.It proves the algorithm can improve the map-matching accuracy effectively.