Dynamic Visualization Of Trajectory Data Under Spatial Temporal Scale Interactively Controlling

The rapid development of remote sensing technology,LBS(Location Based Services)technology and the widespread popularity of handheld mobile devices have resulted in a huge amount of spatiotemporal trajectory data.Trajectory data contains rich spatial and temporal information.The effective analysis of trajectory data can mine the trip rules of urban population,analyze the supply and demand relationship of transportation infrastructure,predict traffic congestion and special events,which provide decision support for urban traffic management departments in traffic dispatching and road construction layout.However,trajectory data usually has the characteristics of spatiotemporal sequence,differ-frequency sampling and poor data quality.How to organize these trajectory data effectively and display it reasonably is one of the difficulties in transportation research field.Among the many spatiotemporal analysis methods of trajectory data,visualization has gradually become an indispensable tool for the study of spatiotemporal trajectory data due to its intuitive,concise,objective,real-time,interactive and controllable advantages.The dynamic expression of geographical phenomena and spatiotemporal data has always been a hot issue in the field of GIS(Global Positioning System).The development of computer technology and the change of data acquisition methods have made the expression of map and map symbols change from static form to dynamic form,which has become an inevitable trend.The rise of concepts such as “digital city”,“smart city” and “intelligent transportation” puts forward higher requirements for the form of data representation.In order to meet the needs of government's quick decision-making and planning management,the visualization results need not only to express urban structure and geospatial information clearly and intuitively,but also to be intuitively perceived and interactive controlled by users.Data visualization is changing from twodimensional to three-dimensional.It is developing towards improving people's visual perception and realizing the interaction between users and information data.This study present a Web-based three-dimensional dynamic visualization method for spatial-temporal trajectory data taking the GPS(Global Positioning System)trajectory data of Shanghai taxis as an example.This method constructs a threedimensional visualization framework integrated by Mapbox GL and Three.js,and uses particle system to simulate the movement state of GPS trajectory data in large-scale urban three-dimensional scene.Through the dual control of spatial scale and temporal scale,the visual representations of different details at different scales can be realized.The complex taxi GPS trajectory data is converted into visualization results that are concise,intuitive,dynamic,interactive and consistent with users' visual cognitive habits in urban three-dimensional scene.The whole thesis contains the following aspects:(1)The Hidden Markov(HMM)algorithm is used to road network matching for trajectory data,and the algorithm is improved.In addition to finding the projection points of the original GPS points on the best matched road segments,nodes of the road segments through which the trajectory passes are inserted into the matching results to avoid the situation that the particle move directly to the next trajectory point without crossing the junction.(2)A 3d visualization framework integrating Mapbox GL and Three.js is built.By calling Three.js in Mapbox GL's custom layer through space coordinate transformation,a flexible and extensible 3d visualization framework is constructed by comprehensive utilizing the 3d geographic scene provided by Mapbox GL and the powerful 3d drawing function of Three.js.(3)A taxis simulated mobility model based on particle system is constructed.In the dynamic visualization process of GPS trajectory data,each taxi entity is abstracted into one particle.The processes of starting,driving and ending a taxi corresponds to the three stages of "generation","activity" and "extinction" of particles respectively.(4)The interactive operations of dynamic maps is designed.Contrasting with the concept of spatial scale,the concept of temporal scale in sequential dynamic map is proposed.According to the correlation between temporal scale and spatial scale in real world and cartographic representation,the interactive operations of spatial temporal dual-scale control and spatiotemporal multi-scale expression are designed.The experimental results show that the proposed method can intuitively reflect the temporal and spatial evolution of taxi trajectories,and has good interactivity and dynamic expression effects.It can be widely applied to real-time monitoring and historical analysis of various trajectory data through network platform.This paper provides the techniques and methods involved in the process from the original taxi GPS trajectory data to the dynamic and interactive 3d visualization results.The research in this paper provides a reference for the research of spatiotemporal trajectory data visualization.