Using simulated Terrestrial Laser Scanning to analyse errors in high-resolution scan data of irregular surfaces

Terrestrial Laser Scanning (TLS) is increasingly being used to collect mm-resolution surface data from a broad range of environments. When scanning complex surfaces, interactions between the surface topography, laser footprint and scanner precision can introduce errors into the point cloud. Quantification of these errors is, however, limited by the availability of independent measurement techniques. This research presents simulated TLS as a new approach to error quantification. Two sets of experiments are presented. The first set demonstrates that simulated TLS is able to reproduce real TLS data from a plane and a pebble. The second set uses simulated TLS to assess a methodology developed for the collection and processing of field TLS data. Simulated TLS data is collected from surfaces up to similar to 1 m(2) created from regular arrays of uniform spheres (sphere diameters of 10 to 100 mm) and irregular arrays of mixed spheres (median sphere diameters of 16 to 94 mm). These data were analysed to (i) assess the effectiveness of the processing methodology at removing erroneous points; (ii) quantify the magnitude of errors in a digital surface model (DSM) interpolated from the processed point cloud; and (iii) investigate the extent to which the interpolated DSMs retained the geometric properties of the original surfaces. The processing methodology was found to be effective, especially on data from coarser surfaces, with the retained points typically having an inter-quartile range (IQR) of point errors of similar to 2 mm. DSM errors varied as a function of sphere size and packing, with DSM errors having an IQR of similar to 2 mm for the regular surfaces and similar to 4 mm for the irregular surfaces. Finally, whilst in the finer surfaces point and DSM errors were a substantial proportion of the sphere diameters, geometrical analysis indicated that the DSMs still reproduced properties of the original surface such as semivariance and some percentiles of the surface elevation distribution. (C) 2010 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). Published by Elsevier B.V. All rights reserved.