Artifact for Testing Terrestrial Laser Scanners for Errors Arising from Target Optical Properties

Abstract

Terrestrial laser scanners (TLSs) are a type of coordinate metrology instrument which collects the three-dimensional coordinates of a scene by measuring the range that a laser beam travels and its orientation (azimuthal and elevation angles). TLSs are subject to a variety of error sources, including those which arise in the ranging, measurement of angles, and interaction of the laser beam with the scanned surface. Documentary standards exist which prescribe tests to expose some of those error sources, including ASTM E2938-15, which addresses errors arising from the ranging unit of a TLS, and ASTM E3125-17, which addresses those from any inherent optical misalignments (leading primarily to angular errors) inside a TLS. While ASTM E2938-15 does allow one to scan different materials, it only captures the impact of the materials’ optical properties in aggregate, along with all other sources’ contribution to ranging errors.

We propose an artifact which interrogates the effects of different materials’ optical scattering behavior on the performance of a TLS. The artifact consists of a planar, media-blasted aluminum plate onto which several, planar materials under test are affixed. The artifact is meant to be modular, so the materials under test can be swapped out easily. Further, the plate holds three to four media-blasted steel spheres from which a unique coordinate system may be established. The artifact is first calibrated on a coordinate measurement machine (CMM). The CMM records a three-dimensional point cloud with sufficient points to fit the radius and center point of each sphere and a plane to the faces of each material and the background plate. The TLS also collects a point cloud of data representing all surfaces on the artifact. From both sets of data, the distance from the approximate center of the surface of each sample to the plane of the background material can be computed and compared. Additionally, the root-mean-square error of each plane fitted to the TLS data is computed. The RMS error and the error in the offset distance can be used as a measure of the ranging uncertainty for that material.

We report on initial testing performed on a grayscale reflectance tile set made from sintered polytetrafluoroethylene. The measured distance from the front of each sample to the background plane shows a strong dependence on the target’s level of reflectance. These preliminary results show that the artifact we propose can be used effectively to examine some errors in TLS measurements. Future measurements must be taken to further examine the impact of range, ranging technology, angle-of-incidence, and additional materials.