This thesis evaluates the accuracy of contour lines generated from UAV data using Agisoft Metashape Professional, ArcGIS Pro and QGIS, compared to contour lines from the Swedish National Land Survey (Lantmäteriet). Ground control points were measured with both 3- and 10- second intervals, where the 3 second measurements proved sufficiently accurate (1,7 centimetres standard uncertainty), while the 10 second measurements were unreliable, likely due to strong wind conditions. Therefore, the measurements were processed at 3 second intervals to create terrain models.

Contour lines were generated and analysed visually and statistically using ArcGIS Pro and Excel. The results indicate that UAV-based contour lines, with appropriate classification parameters, can achieve high accuracy, especially when processed in ArcGIS Pro and QGIS. ArcGIS provided a good balance between accuracy and processing efficiency, while QGIS offered a free alternative with comparable results. However, there was a methodological flaw in the Metashape Professional Metashape Professional classifier that affected the results of the terrain models in some areas. Metashape Professional also lacked tools for contour generalization, which negatively affected accuracy.

Although UAV-based contour lines had higher resolution that expected. With proper processing and classification, UAV data may be optimal for extracting contour lines which is suggested for further research.