测绘通报 ›› 2018, Vol. 0 ›› Issue (12): 46-51.doi: 10.13474/j.cnki.11-2246.2018.0382

Previous Articles     Next Articles

The 3D Roadway Alignment Measurement for High Definition Map

LUO Wenting, LI Lin   

  1. School of Transportation and Civil Engineering, Fujian Agriculture and Forestry University, Fuzhou 350000, China
  • Received:2018-05-19 Revised:2018-09-17 Online:2018-12-25 Published:2019-01-03

Abstract: 3D high definition (HD) map is a critical part of driverless systems and a basis of intelligent transportation. This study collects road alignment data by LiDAR system and INS, and provides a series of algorithms for roadway alignments extraction and calibration, which can eliminate the impacts of vehicle vibration and wandering. The lane markings are identified and positioned based on dilation and erosion operation and Canny edge detection. Based on the lateral position of lane marking the center line of the lane is obtained. The PT and PC are detected by linear fitting analysis, and curve radii is calculated using chord offset method. The cross slope is calibrated by transverse profile data, and the longitudinal grade is calibrated by bilateral filtering algorithm. This study provides a case study of 3D HD maps creation for Pushang grade-separated interchanges on Fuzhou. The validation test results show that the errors of PT detection, PC detection and curve radii measurement are 2.97%, 2.54%, and 1.16% respectively. The P-values for validation tests of cross slope and longitudinal grade measurement are 0.631 and 0.945 respectively based on Dunnett's test. It implies that the proposed methods have high accuracy on roadway alignment measurement. This study provides data basis for driverless system.

Key words: 3D high definition map, grade-separated interchanges, LiDAR system, inertial navigation system, driverless system

CLC Number: