Toward the development of a documentary standard for derived-point to derived-point distance performance evaluation of spherical coordinate 3D imaging systems

This paper describes ongoing research work at the National Institute of Standards and Technology (NIST) in support of the development of a documentary standard for derived-point to derived-point distance performance evaluation of spherical coordinate 3D imaging systems (laser scanners). The proposed standard is being developed by the subcommittee E57.02 on Test Methods within the ASTM E57 committee on 3D Imaging Systems. The overall approach of the testing strategy is to compare derived-point to derived-point distances obtained by an instrument under test against those obtained by a reference instrument (such as a laser tracker) through the measurement of reference artifacts. This paper presents a geometric error model and a sensitivity analysis based approach to identify sensitive test positions (i.e., where to place targets in space) for performance evaluation. We show that it is not only important for target spheres to allow repeatable determination of their centers, but it is also necessary they have good form. We highlight key issues pertaining to processing laser scanner data, such as segmenting the data, identification and removal of outliers, and choice of algorithms to determine sphere centers. An SMR walking method to establish reference distance and its uncertainty is discussed. All of these are key components of the proposed documentary standard.