Spatio-Temporal Pattern Mining And Visualization Of Movement Trajectory Data

To improve the spatiotemporal analysis and modeling techniques of the movement data,this research focused on the hardware construction,data acquisition and analysis,data mining and data visualization of spatial and temporal data to discuss methods and techniques for the collection,analysis,mining,and visualization of movement data.First,it introduced some basic techniques including time geography and visualization,because the paper will carry out quantitative and qualitative analysis of movement data based on the concept of time geography.This study completed the design and implementation of a space-time movement data analysis system.The space-time movement data analysis system consists of hardware and software: data acquisition system and software system.Our first study is about collecting data about moving objects,so the first part will focused on the acquisition of movement data sets and the construction of hardware devices.The second study developed a comprehensive spatio-temporal data analysis system software including four modules: a trajectory data preprocessing module,a database operation and management module,a space analysis module,and a spatiotemporal data visualization module.There are many problems with the raw data collected from the real world,and data analysis shows that its quality has not reached an acceptable level,which requires the addition of tasks to improve and simplify the collected data.In addition,the software system integrates the quantitative and qualitative representation of moving objects and visualizes the extracted information based on the concept of temporal geography.In particular,it uses space-time trajectories,spatio-temporal density,and volume rendering techniques to capture and analyze the spatio-temporal patterns and characteristics of moving objects.In order to evaluate the capabilities of the developed framework and system,an real movement data set collected by a GPS device was analyzed.That is,using the vehicle movement data at an intersection in the mining area as an example to explore the temporal and spatial patterns and characteristics of moving objects.This study aims to contribute to the study of spatio-temporal data collection,pattern exploration,and visualization techniques for movement data and is divided into five aspects.1)Develop a comprehensive spatiotemporal data analysis system to analyze and illustrate the spatiotemporal patterns and processes of moving objects.2)Design and implement data acquisition terminal and successfully carry on the open-pit mine transport vehicle and successfully collect data on moving vehicles.3)A qualitative and quantitative description of moving objects is proposed,as well as techniques and methods for the exploration of movement patterns.4)Develop new visualization methods for trajectory data and integrate time geography frameworks to visually explore spatiotemporal patterns and trends.5)Test the real movement data set of the moving object under the time geography framework based on the proposed framework and techniques.Specifically,this study proposes a new analysis framework integrating time geography,trajectory data mining and 3D volume visualization,as well as the tools that use this framework to study the theoretical and practical datasets of movement objects.The results show that the framework and tools have a good ability to identify and visualize a variety of movement activities in space and time.