Research On Urban Road Extraction Based On Airborne LiDAR Point Cloud Data

As an infrastructure in the city,road plays an important role in transportation,emergency response and urban construction.With the rapid development of cities and the construction of “smart cities”,data acquisition and real-time updates of road networks have become particularly important.The airborne LiDAR system is characterized by its efficient,accurate and time-efficient 3D information acquisition capability and rich echo intensity information,making this technology a hot research direction for rapid urban road extraction.For the extraction of urban roads from airborne LiDAR point cloud data,domestic and foreign scholars have proposed various algorithms,but most of them have their own defects in the extraction of road point clouds.For example,the integrity of road point cloud data extraction is not high,and it is impossible to effectively eliminate parking lot,open space and other areas that are connected to the road and have very similar characteristics,and road center line extraction fracture phenomenon.This article focuses on how to effectively eliminate parking lot,open space and other areas to carry out the following research work:Researched achievements of domestic and foreign scholars in the extraction of LiDAR point cloud roads;outlines the composition and working principle of the airborne LiDAR system;the characteristics of the airborne LiDAR point cloud and the processing flow are analyzed,which lays the foundation for the subsequent extraction of road point cloud data.Several kinds of filtering algorithms that are currently popular are studied.Finally,a progressive triangular network encryption algorithm with good performance in urban areas is used to separate ground points and non-ground points,and the principle of the algorithm is introduced in detail.The intensity attributes of airborne LiDAR point cloud data are studied,several typical medium reflectivities are listed,and the intensity values of road sample data are used to statistically calculate the intensity thresholds of road area point cloud data.Thus,the original road point cloud data is extracted.It is difficult to eliminate areas that are connected to roads and have extremely similar characteristics such as parking lots and open squares.It is proposed to first use the distance segmentation method to eliminate discrete point cloud data scattered on both sides of the road,then point cloud segmentation method based on RANSAC algorithm is used to extract road point cloud data,and this method can effectively eliminate parking lots,open squares and other areas.In order to facilitate the use of extracted road point cloud data in the future,this paper uses the mathematical morphology and Hough transform detection method to extract road center point cloud data.At last,this paper takes the airborne LiDAR point cloud data of a city area in the Netherlands as the original data,realizes the above method in MATLAB 2014 and Microsoft Visual Studio 2010 environment.The accuracy,completeness,and overall quality were used to evaluate and analyze the accuracy of the experimental results.The result proves the effectiveness of the algorithm.