If you have any problems related to the accessibility of any content (or if you want to request that a specific publication be accessible), please contact us at email@example.com.
Methodologies and Applications of Data Coordinate Conversion for Roadside LiDAR
AltmetricsView Usage Statistics
Light Detection and Ranging (LiDAR) is becoming more popular in applications of the transportation field, including traffic data collection, autonomous vehicles, and connected vehicles. Compared with traditional methods, LiDAR can provide high-resolution-micro-traffic data (HRMTD) for all road users without being affected by the light condition. Unlike the macro data collected by traditional sensors containing traffic flow rates, average speeds, and occupancy information, the HRMTD can provide higher accuracy and more detailed multimodal all-traffic trajectories data. But there are still some limitations when using it. The first one is that the raw data is in LiDAR’s coordinate system, which greatly affects the visibility of the data. Secondly, the detection range limits its further development. Although LiDAR can detect data within 200 m from itself, the effective detection range is 50~ 60 m. What’s more, the occlusion issue occurred from time to time.To overcome these limitations, data mapping and integration methods are needed. This research proposed the data integration and mapping method for roadside LiDAR sensors. There is a total of six main steps in this method: reference points collection, reference points matching, transformation matrix calculation, time synchronization, data integration, and data mapping. The raw LiDAR data is in the Cartesian coordinate system. In this coordinate system, the position of each LiDAR point is represented by (x,y,z). To map these points on the GIS-based software based on the WGS 1984 coordinate system, the coordinate system of the LiDAR data should be transformed. After converting the LiDAR data into Geographic Coordinate Systems, the ICP method is applied to integrate the data collected by multiple LiDAR sensors. Compared with the original LiDAR data, the longitude, latitude, and elevation information are added to the processed dataset. The new dataset can be used as the input for the HRMTD processing procedures for roadside LiDAR. Other than benefiting the autonomous vehicle(AV) system and connected vehicle(CV) system, the HRMTD can also serve other transportation applications. This research provides an application using the HRMTD obtained from roadside LiDAR data to extract lane and crosswalk-based multimodal traffic volumes. This method has three main steps: start and endpoint selection, detection zone selection, and threshold learning. The second step is the primary step of the method, which can be divided into four sub-steps: location searching, data comparison, size searching, and best zone selection. A whole day of data collected in the real world is used to verify the method and compared with the manually counted traffic volume, and the result shows that the accuracy of this traffic volume extraction method reaches 95% or higher. This research will significantly change how traffic agencies assessing road network performance and add great traffic values to the existing probe-vehicle data and crowd-resourced data.