Automatic cross section extraction and cross slope measurement for curved ramps using light detection and ranging point clouds

Road cross section and slope are essential for ensuring both safe road operations and effective drainage. This paper presents a method that integrates the point clouds processing and geometric information computation derived from LiDAR point cloud data. The contributions are as follows: (1) a road marking detection method is developed to extract pavement points and remove the noisy points; (2) the centroid of road markings is extracted and defined as the road axis, determined by both road markings and boundaries utilizing the index-based classification technique; (3) knot and region of interest within the cross-sectional profile are generated to extract cross section points. A Principal component analysis algorithm is used to measure the cross slope. Based on the 13 locations of two sites within the interchange ramp, differences between the proposed method and field surveys range from 0.02 % to 0.46 %, indicating the accuracy of the proposed method for cross slope measurement.