In the present project it is investigated how UAV data can be used for making local updates to the Danish Elevation Model (DK-DEM). It was found that a change detection and categorisation of the UAV pointcloud is essential parts of an algorithm for updating the DK-DEM. It was experienced that the UAV data have several quality issues and for this reason a quality assessment of the data was required. It was chosen to prioritise the development of an algorithm for quality assessment and change detection. The quality assessment removes outliers in the UAV data by performing a robust adjustment of a plane on the UAV data and the data quality is evaluated by statistical measures from the adjustment. The change detection evaluates elevation differences by a standard deviation of the difference. The standard deviation of the difference is based on the quality of both the DK-DEM and the UAV data.
The quality assessment successfully removed outliers in the UAV data and it was seen that the dataset in general had a higher precision after it was performed. The change detection was able to detect changes of terrain and from inspection of the change detection it was seen that the algorithm achieve a satisfying result. As a categorisation have not been performed several issues regarding non-terrain objects is present. For this reason the developed algorithms is only seen as part of a final algorithm.