Unorganised point cloud dataset, as a transitional data model in several applications, usually contains a considerable amount of undesirable irregularities, such as strong variability of local point density, missing data, overlapping points and noise caused by scattering characteristics of the environment. For these reasons, further processing of such data, e.g. for construction of higher order geometric models of the topography or other sensed objects, may be quite problematic, especially in the field of object detection and three-dimensional surface reconstruction. This paper is focused on applying the proposed methods for reducing the mentioned irregularities from several datasets containing 3D point clouds acquired by LiDAR scanners and multibeam sonars. The good performance of the proposed methods has been shown along with illustration of the importance of the appropriate design of the point cloud data preprocessing step in the context of the final results of the 3D shape reconstruction procedure.
Authors
Additional information
- DOI
- Digital Object Identifier link open in new tab 10.1109/bgc.geomatics.2016.41
- Category
- Aktywność konferencyjna
- Type
- materiały konferencyjne indeksowane w Web of Science
- Language
- angielski
- Publication year
- 2016
