Quality Evaluation of Spatial Point-Cloud Data Collected by Vehicle-Borne Laser Scanner

Abstract

Vehicle-borne laser scanning is used to collect urban street and building data for various modeling and mapping applications. The quality of laser point-cloud data is important for regenerating 3D urban street scene. Quality evaluation includes spatial structure analysis and positioning accuracy analysis. In spatial structure analysis, we evaluate the quality of point-cloud data high or low by analyzing the objectspsila spatial structure projection at each of the three coordinate axis. In positioning accuracy analysis, comparing sign pointspsila 3D coordinates collected by laser scanner with the precise coordinates collected by Total Station, we get the point-cloudspsila positioning accuracy. A new sign point was also designed to overcome the difficulty of extracting sign points coordinates from point cloud data. After the evaluation, we make a suggestion about how to improve the data quality.