Research On Key Technologies And Applications Of Map Topology Extraction Based On GPS Data

How to obtain electronic maps at a low cost and with high efficiency and accuracy is a hotspot and difficulty in the current research field of electronic map production.Generally,there are two main methods for making electronic maps:(1)Digitizing paper maps;(2)Making images of remote sensing into maps by human-computer interaction.Both of these methods require a lot of manpower and financial resources.Therefore,the production of automated map is an important direction to reduce labor costs.Based on the research and analysis of GPS trajectories data,this paper summarizes and improves the existing GPS trajectories analysis methods,and designs and implements a map information extraction algorithm suitable for studying data sets.The algorithm mainly includes the application of stagnation point processing,intersection mining,intersection rule mining and road topology mining.The main contents of the research work done in this paper are as follows:(1)Data analysis and pre-processing.The characteristics and existing problems of GPS trajectories are analyzed,and data preprocessing is performed on the original GPS trajectories data.In this paper,pre-processing of transformation,discontinuity data pre-processing,stagnation point data pre-processing,missing data and abnormal data preprocessing are performed;and GPS trajectories data are rasterized to improve the efficiency of the algorithm.(2)Design of key technology and verification of road topology mining.This paper first completes the design of intersection mining algorithm based on k-order angle.Aiming at the problem of continuous fitting for spatio-temporal data,the QSF fitting algorithm is proposed to improve the efficiency of classical fitting algorithm.Second,on the basis of the algorithm,which is about intersection mining,the intersection rule extraction algorithm based on hierarchical clustering and the road topology mining algorithm using the k-order angle are proposed.Last,the above algorithms are applied to the actual GPS trajectories data.The effectiveness of the algorithms in this paper is shown by quantitative or qualitative analysis of the experimental results.(3)Application of the algorithm and system implementation.A map topology visualization system is designed and preliminarily implemented according to the proposed algorithm,which provides a reference for the design of a more efficient and convenient system of map visualization.It also demonstrates the effectiveness of the algorithms in this paper.The algorithm proposed in this paper has certain enlightening significance for the realization of the extraction of automated road network,and provides some design ideas for practical applications,and has a certain value in theory development and engineering application.