Real properties in Dalarna often consist of small lots of forest or agriculture. These lots are often elongated and consequently impractical, therefore the Swedish mapping, cadastral and land registration authority performs comprising land consolidations. These are to swap land between the owners of the properties, in order to form appropriate lots, for purpose of, inter alia, promoting investment and development opportunities in the region. In order to establish a map, a decision basis, for the valuation of the lot, the claimed proprietary right has to be surveyed. The work is performed with NRTK (network-RTK) or total station, and the maximum allowed planar deviation is 0.50 m (base level requirement). The purpose of this study was to examine whether there is a possibility of using Unmanned Aircraft Systems (UAS) when claimed property boundaries are going to be surveyed. UAS is a system consisting of an unmanned aerial vehicle, a digital camera, a Global Positioning System (GPS) receiver, an Inertial Navigation System (INS) and a control station.
In a delimitated area of about 24 ha in Norra Åbyggeby, north of Gävle, black and white and white markers, with the size of 0.40x0.40 m, were positioned in the terrain and then surveyed as ground control points or boundary points with network RTK. Mean planimetric coordinates have been calculated for the control points, which have been used as reference coordinates. These have been compared with the coordinates measured in an orthophotomosaic produced from flights with UAS at an altitude of 100 and 180 m. To get a further comparison, the coordinates were determined by a block adjustment in the software PhotoScan from Agisoft. The object identification of the markers, placed in different environments, and from different altitudes has been studied.
The digitalization in the orthophotomosaic resulted in a Root Mean Square error (RMS-value) of 0.083 m at an altitude of 100 m and a value of 0.049 m at an altitude of 180 m. Corresponding RMS values were 0.071 m at an altitude of 100 m and 0.077 m at an altitude of 180 m when computed in PhotoScan. F-test has been calculated using the four RMS values, the result of the F-test showed that coordinates obtained in an orthophotomosaic, in ArcMap, are equivalent to coordinates obtained by block adjustment, in PhotoScan. The F-test also showed that the coordinates are equivalent from altitudes 100 and 180 m by block adjustment, but they are not equivalent when they are obtained in an orthophotomosaic from altitudes 100 and 180 m. If we disregard the systematic error at three of the points (orthophotomosaic 100 m) the F-test did not show any statistically significant difference between the two altitudes.
All deviations were below the base level requirement. The largest planar deviation was 0.181 m at an altitude of 100 m and 0.083 m at an altitude of 180 m. Corresponding values for PhotoScan were 0.155 m and 0.148 m. How dense the forest was where the marker was placed and the impact of the sun, in terms of shadows and brightness, have affected the composition of the mosaic, and consequently the deviations. UAS can be used for surveying of claimed property boundaries, but there is no guarantee that all signalized boundary points can be surveyed directly in the orthophotomosaic. One recommendation is to use a less accurate method for the measurement of the ground control points (needed for the georeferencing of the point cloud/orthophotomosaic) than the method used in this thesis. The higher altitude is preferable because the time requirement of the flight will be shorter, and a smaller number of aerial photos need to be processed. Choose the method that the user is used to, manual digitizing in an orthophotomosaic or automatic calculation in a block adjustment.