Abstract: Traditional point cloud section extraction is carried out on the basis of 3D products. It is necessary to obtain ground points by filtering point cloud firstly and then sampling to obtain sections. The process suffers low work efficiency and unguaranteed quality of the section production, requiring a lot of manual refinement work. This paper proposes a new production process and method for the production demand of highway cross section thematic production. Firstly, a cross section is extracted based on digital surface model (DSM), and the point cloud set of the cross section is two-dimensionally projected to obtain a two-dimensional point array. Then, the point cloud array is filtered based on the skewness-balanced filtering algorithm, and finally a feature preserving cross section point cloud is extracted. Practice has proved that the new process reduces the amount of data processing and effectively improves the efficiency and precision of cross section production.

Key words: airborne LiDAR, cross section, point cloud filtering, skewness balance, point cloud thinning