Trajectory Visual Analysis System Based On OD Data

Visual analysis is widely used in the information mining and analysis of all kinds of data.With the development of smart cities and the advent of the era of information explosion,a large amount of trajectory data is generated in many industries every day.All sorts of useful information closely related to people are hidden in these data.Therefore,it is necessary to mine and analyze the unique information rules in the data by means of vivid and intuitive visualization.In order to deal with the analysis of various moving trajectory data,experts have researched a large number of processing methods about trajectory data and developed the corresponding analysis system software about trajectory visualization.However,with the gradual expansion of data volume,the direct drawing of trajectory data after clustering,binding and other processing will also cause serious visual confusion due to crossover and overlap.Which makes it very difficult for researchers to analyze and mine the hidden information in the data.In order to reduce visual confusion and clearly analyze human behavior in the data,it is very necessary to research and design a visual analysis method for OD trajectory data.To solve the above problems,this thesis makes the following work based on the existing methods:1.This thesis proposed an analysis method of trajectory visualization based on OD data.On the basis of the existing trajectory processing method,this method combines the algorithm of edge binding.All process is as follows: Compare the original OD data with the word,and obtain the corresponding word vector by using model training,and make low-dimensional projection of a higher-dimensional vector by t-SNE,and apply the edge binding algorithm to the data which is optimized after blue noise sampling algorithm.Which is to improve the OD trajectory of a large number of overlapping visual confusion,and analyze visually human flow pattern.2.Based on the method in this thesis,a prototype system of trajectory data visualization based on Echarts and Leaflet is designed and implemented.The system is mainly divided into two modules: trajectory visualization and information visualization.In the two modules,a number of visual charts are designed for different information,and the interactive linkage function between visual charts is provided.The method proposed in this thesis and the visualization system designed are used to reduce the visual confusion in the rendering of trajectory data,and to analyze the rules of human flow.The experimental results and case studies show that the method proposed in this thesis has advantages in some aspects.