Utvärdering av handhållen 3D-skanner DotProducts DPI-8

Sammanfattning: Within the measurement technology area, an easy to use 3D-scanner to create three-dimensional point cloud has entered the market. Compared to traditional laser scanning, data collection is done quickly and at low cost. This kind of equipment is carried with one hand, and registered objects are presented in the display of the equipment. The scanning is performed by the operator while walking around in the area, which makes this technique more cost effective than traditional laser scanning. The instrument is revolutionizing as far as data acquisition goes, but only a few previous studies about 3D-scanners that use real time feedback of uncertainties have been made. Therefore, it is important to evaluate the equipment and document achieved uncertainties to determine in which applications it is useful. The purpose of this study was to evaluate the achieved uncertainties of the DotProducts DPI-8 handscanner, compared to the Leica Nova MS50 by scanning different materials indoors, as well as elongated objects outdoors. Part of the study was performed by scanning a room, containing different types of materials, which were compared. Another part of the study investigated whether the scanner is capable of scanning a railing, which was performed outdoors to determine if the scanner is appropriate to scan long objects outdoors. In addition, different spheres were scanned to evaluate if they are adequate to use in the registering of pointclouds. The results showed that all compared surfaces are within the manufacturers given specification, which is 40 mm uncertainty at 3 m scanning distance. Surfaces with greater deviation were the ones not scanned perpendicular to the scanned object. Scanning outdoors is concluded not to be suitable using the DPI‑8, as the sunlight causes the instrument to lose tracking every now and then. Considering the spheres, it was determined that the 40 mm-spheres were not suitable to use for registration, since the centers of the spheres differed 10–40 mm between the center points. Furthermore, a six-inch sphere was scanned, but it could not be recognized by the DPI-8. In conclusion, the hand scanner is a worthy complement to a terrestrial laser scanner indoors where the sightline is partly obscured. However, the scanning is required to be orthogonal to the surface, and the instruments need common flat targets to get a good result.