Vehicle-mounted LiDAR Road Detection Method

The vehicle-mounted LiDAR system can measure large-scene 3D spatial data quickly,conveniently and with high accuracy,and has unique advantages in acquiring streetscape and road 3D data,which can collect road 3D data efficiently without affecting transportation.Nowadays,the transportation industry is developing rapidly,and roads are spread all over the corner,so the timely detection of roads is especially important.In this paper,based on the team’s self-developed vehicle-mounted LiDAR measurement system,we propose to use the vehicle-mounted LiDAR system for road surface inspection.In order to ensure the accuracy of road surface data,we need to check and calibrate the system,improve the accuracy of the vehicle-mounted LiDAR point cloud and measure the road surface point cloud data afterwards,based on the vehicle-mounted point cloud data,we can analyze the road surface flatness,identify cracks,identify ruts and analyze the depth of the road surface anti-skid performance structure.The main research work of this paper is as follows:1.Research on pavement inspection methods and analysis of vehicle-mounted LiDAR point cloud data can provide the possibility of testing certain indicators of pavement,establishing the calculation of pavement smoothness,the identification of pavement cracks and ruts using vehicle-mounted point clouds,and the calculation of pavement skid resistance construction depth.2.From the hardware integration principle and data decoding principle of vehicle-mounted LiDAR measurement system,as the trajectory points play a controlling role on the vehicle-mounted point cloud,a method of optimizing the trajectory is proposed,mainly for the vehicle-mounted GNSS/INS in the area of bad GNSS signal or when the signal is out of lock,two positioning and fixing methods of GNSS/INS and INS/odometer are proposed,and the microcontroller is used to develop a positioning mode that can be switched independently for the presence or absence of GNSS signal,and a complete frame of data is obtained experimentally for the combined GNSS/INS navigation method.3.From the principle of in-vehicle point cloud data solving,the check calibration model based on the special target ball for LiDAR and IMU placement 6 parameters is proposed.According to the model principle,it is known that the check calibration is not only for LiDAR and IMU poses but also for optimizing other multi-source errors,and the necessity of check calibration is verified.4.The vehicle-mounted LiDAR system after calibration measures the pavement data more reliably,and proposes a construction depth calculation method by relying on a section of the road for vehicle-mounted experimental point cloud data for pavement leveling detection,crack identification,rut identification and analysis.Through the above research content,this paper mainly combines the vehicle-mounted LiDAR system with road surface detection,and provides the feasibility of applying the vehicle-mounted LiDAR system to road surface detection.Using our team’s self-developed vehicle-mounted LiDAR measurement system to analyze its principle in depth,from hardware integration to data processing,from trajectory optimization to point cloud optimization,the possibility of vehicle-mounted LiDAR system accuracy improvement is proposed from multiple angles to provide a more accurate data source for road inspection,and finally the feasibility of vehicle-mounted LiDAR system for pavement leveling detection,the effect of pavement cracks in rutting identification and The method of structure depth calculation.