Evaluation of mobile mapping point clouds in the context of height difference estimation

Abstract In recent years, the usage of point cloud data for various mapping and other civil engineering tasks has become increasingly popular. The detailed acquisition of the environment forms a great advantage compared to point-wise methods using e.g. total station measurements. The major drawback is, that the uncertainty analysis of the measured points and accordingly the derived parameters is not straightforward. A variance propagation of the observations would not lead to plausible results, since the stochastic model is unknown in most of the cases. In this work, we present an empirical way to determine uncertainty information of the point cloud data captured by a mobile mapping system (MMS) related to height differences by using mainly the road surface, where the system drives on. Height differences between objects are often considered in the context of monitoring of land subsidence and engineering structures or mapping tasks. For the evaluation, height differences between points are analyzed, which differ in three major aspects from each other: the distances between the height observations, the environmental conditions, and the locations in the measurement volume of the system. Repeated measurements of the road surface and artificial targets are used to evaluate the precision of the height differences. Using reference values enables an analysis of the full uncertainty information. The results from two data sets show, that the environmental conditions severely influence the GNSS quality and consequently the precision of height differences decreases. Due to positive correlations between neighboring points, which are caused by the trajectory information, the height difference uncertainty increases concerning the traveled distance between the points. Because of remaining calibration errors, the location of the objects within the measurement volume of the profile laser scanner also influences the uncertainty of the height values and thereby also of height differences.