Research And Application Of Passenger Hot Spot Visualization Based On Taxi GPS Data

With the increasing of the augment of city residents income and awareness of time management,more and more people choose taxi as its means of transportation,Therefore,the taxi market has been greatly developed.Usually taxis are equipped with GPS terminals,those terminal will periodically send some real-time status information to the taxi dispatching center,such as Latitude、longitude and Driving direction and load information and so on.After a while,Taxi companies collect and store a vast amount of taxi track data.These days,more and more people are taking part in the data industry,which lead to that it become one of the research hotspots of big data of urban traffic Previous taxi driver hot spot visualization method have the ability to quickly process taxi GPS data,but the visualizations are not ideal and the exact location of some key hot spots is not accurate enough,which will lead to the information provided to the user is inaccurate.At the same time,the design of visual interface needs to be improved,and other information hidden by taxi GPS data can not be directly reflected.Based on the above questions.The main contents of this paper are as follows:Firstly,data preprocessing and map matching: first processing the collected GPS data and OSM map data.Deletes nonconforming data and fixes lost data.Adjust the load point of the taxi read by using geometric map matching.Thus,the collected original GPS carrier hotspots are corrected.Complete the preparation of the data.Secondly,improving the DBSCAN algorithm : According to the passenger point in the road distribution.this paper Select DBSCAN algorithm based on density clustering.At the same time,Considering the road traffic restrictions such as green belts and traffic lights.This paper design An improved DBSCAN algorithm based on road restriction that The algorithm proposed in this paper introduces the travel distance of the taxi to the DBSCANalgorithm.soreplaceThe DBSCAN algorithm used to calculate the similarity measure of Euclidean distance 。Thirdly,visual function realization and result analysis.The purpose of this part is to meet the specific requirements of the visualization of this article,and to convert the large number of multi-dimensional taxi GPS passenger hotspot data into the shapes,symbols and colors that the observer can perceive.It provides a reliable basis for taxi company planning and residents’ travel.