Three-dimensional data quality assessment: Unmanned aerial vehicle photogrammetry and mobile laser scanner

With significantly rapid in acquiring data and sufficient data quality, unmanned aerial vehicle (UAV) photogrammetry and land based mobile laser scanner are actively implemented in three-dimensional (3D) data acquisition that involve with large area. Considering the final data provided by both measurement approaches are point clouds, accuracy assessment using several well-distributed control points are less significant. With intention to robustly evaluate the accuracies of both measurement approaches using point clouds, this study has established reference point clouds using terrestrial laser scanner (TLS) and tacheometry techniques. At the similar test site, fourteen (14) images were captured using UAV photogrammetry approach and georeferenced point clouds were directly acquired from MLS measurement. To produce quality point clouds from photogrammetry approach, six (6) ground control points (GCP) have been well-distributed at the test area to aid geometry correction in image processing phase. Obtained point clouds from both measurement approaches were deviated with the reference point clouds to determine values of mean deviations with the precisions. Based on law of propagation of variance (LOPOV) algorithm, final accuracy of the tested UAV photogrammetry and MLS were computed by propagating the accuracy of reference point clouds and yielded mean deviations of both approaches. Consider the theories and constraints for both approaches, it is found that the yielded accuracies are meet the measurement principles