Abstract
ABSTRACTPoint clouds are a very detailed and accurate vector data model of 3D geographic information. In contrast to other data models, no standard has been defined for visualization and management of point clouds based on levels of detail (LOD). This paper proposes the application of the concept of LODs to point clouds and defines the LOD0 for point cloud classification (the lowest possible level of detail) as urban and non-urban. A methodology based on the use of machine learning techniques is developed to perform LOD0 classification to airborne LiDAR data. Point clouds acquired with airborne laser scanner (ALS) are structured in grid maps and geometric features related with Z distribution and roughness are extracted from each cell. Six machine learning classifiers have been trained with datasets including urban (cities) and non-urban samples (farmlands and forests). The influence of grid size, point density, number of features and classifier type are analysed in detail. The classifiers have been tested in three case studies. The best results correspond to a grid size of 100 m and the use of 12 geometric features. The accuracy is around 90% in all tests and Cohen’s Kappa index reaches 81% in the best of cases.
Highlights
Point clouds focus great research in recent years
In order to be able to structure the data from a lower levels of detail (LOD) to a higher one, this paper presents a LOD0 classification methodology for point clouds on a large-scale based on machine learning techniques
Each sample is represented by the 12 features explained in Section 3.2: five variables of the histogram related with the height and seven values of roughness
Summary
Point clouds focus great research in recent years. The technology responsible for their acquisition has become cheaper and more companies offer these services. Point clouds can be considered as vector data models of 3D points, they are not usually integrated into Geographic Information System (GIS) due to their large size and non-structured nature. Aerial Laser Scanner (ALS) has less density than the previous ones, but allows to acquire large areas of land quickly and from top view perspective (Yan, Shaker, & El-Ashmawy, 2015). Together, all these technologies can provide the user a complete point cloud of a scene with different densities
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
