- Simplification Algorithm for Airborne Point Clouds
Simplification Algorithm for Airborne Point Clouds
Aerial and Terrestrial Laser Scanning
Full Paper Review
Matthew Ross Westaway, University of Cape Town, South Africa, firstname.lastname@example.org
George Sithole, University of Cape Town,South Africa,email@example.com
The processing of point cloud data is influenced by the size of the point cloud. Today the size of point clouds has increased due to improvements in laser scanning technology; therefore processing of point cloud data has become computationally exhaustive and storage demands have increased. Consequently, accurate point cloud simplification algorithms are being sought. In this paper, a point cloud simplification algorithm is proposed. In the presented algorithm, important points (feature points) are preserved and redundant points (non-feature points) are removed, whilst ensuring the point cloud satisfies a minimum density requirement globally.
The importance of a point can be evaluated by the relationship it has with its surrounding points. This relationship is quantified by using measures of information theory. These measures use the surface geometry about a point as a proxy for information. Five measures from past work are compared by quantitatively and visually measuring the accuracy of the simplified point cloud. The measure of information theory that results in the most accurate detection of feature points is found.