ASTM E2938-15 – Putting Laser Scanners to the Test
(Thanks to Dr. Kamel Saidi at the National Institute of Standards and Technology (www.nist.gov) for the following summary of the work that was done to develop a first-of-its-kind standard test for laser scanners.)
The ASTM E57.02 sub-committee on Test Methods for 3D Imaging Systems has recently published the first standard for time-of-flight, spherical coordinate laser scanners. These instruments measure a range and two angles (azimuth and elevation) for each point in a point cloud. The spherical coordinates are then often reported in Cartesian coordinates for most applications. This new standard is concerned with the range-measurement performance of laser scanners. The standard provides a way to determine the relative range error for a laser scanner in the medium-range (2 m to 150 m).
Because the origin of a laser scanner’s coordinate system is often a point inside the scanner and cannot be physically measured, the absolute range from a scanner to a scanned surface cannot be measured by any means other than with the scanner itself. So, for example, you cannot use a laser tracker to measure the reference (or “true”) distance between the scanner and a target. This reference distance is needed to determine the absolute range error of a laser scanner. Therefore, instead of an absolute range error, the new standard provides a way of evaluating the relative range error of a laser scanner.
Fig. 1 The standard being exercised at NRC-Canada using a rectangular planar target mounted on a rail
As defined in the standard, the relative range is the distance between two objects located along a straight line nominally passing through their centers and the instrument’s origin. In this standard, the relative range error is the difference in the distance between two planar targets measured with the instrument under test (IUT) and a reference distance between the same two targets. The reference distance is obtained using other instruments or methods that have a lower distance-measurement uncertainty than the laser scanner.
This standard makes it possible for a user to evaluate the relative range error of a laser scanner at any range between 2 m and 150 m; however, the standard recommends that the first target (closest to the scanner) be placed at a distance that is provided by the laser-scanner manufacturer. This distance will typically be the laser scanner’s “sweet spot” at which the ranging performance is optimal. The user is then free to choose the placement of the second target.
Placing the first target at the “sweet spot” ensures that the error from the range measurement of target 1 has a minimal contribution to the relative range error between target 1 and target 2. This, in turn, ensures that the relative range-measurement performance of the scanner between target 1 and target 2 approximates the absolute-range measurement performance from the laser scanner to target 2 (assuming that any bias – or constant offset – has already been eliminated from the system during calibration).
This standard also allows a user to choose any target material, provided the reflectance, flatness and mounting of the target are reported and that the required measurement uncertainty is met (the standard describes how an uncertainty budget can be calculated for a particular test setup). The option for the user to choose a target material addresses the need to accommodate the numerous types of materials properties that are found in real world applications.
Although the standard requires the evaluation of the relative range error for one distance, to more fully characterize an instrument’s performance the test can be repeated for multiple distances and target materials.
The sub-committee that developed this standard is made up of a balanced mix of laser-scanner manufacturers, users, consumers and others with a general interest in laser scanning. These volunteers[*] have been working on this standard for several years and we believe that the new standard is a first step to leveling the playing field for evaluating laser scanners on the market, for which no standards currently exist. Sub-committee E57.02 is also working on a new standard for evaluating the volumetric performance of medium-range, spherical coordinate laser scanners.
More information about ASTM E57 is available at http://www.astm.org/COMMITTEE/E57.htm. The full standard is available for purchase from ASTM (http://www.astm.org/) as ASTM E2938-15, Standard Test Method for Evaluating the Relative-Range Measurement Performance of 3D Imaging Systems in the Medium Range.
[*] Some of the past and present sub-committee members who contributed to the development of this standard include (in alphabetical order): Kevin Ackley, Alan Aindow, Paul Banks, John Battaglia, Angelo Beraldin, Robert Bridges, Geraldine S. Cheok, Luc Cournoyer, Jerry Dimsdale, Charles Fronczek, Tad Fry, Michael Garvey, Brent Gelhar, Janice Gerde, David E. Gilsinn, Thomas Greaves, Steven Hand, Tom Hedges, Darin Ingimarson, Geoff Jacobs, Mark Klusza, Ted Knaak, Alan M. Lytle, David McKinnon, Alan Metzel, Spike Milligan, David Ober, John Pamateer, Fred Persi, Michael Raphael, Kamel S. Saidi, Jonathan Saint Clair, Mitch Schefcik, William C. Stone, Thomas Valenti, John Villanueva, Gregory Walsh, and Christoph J. Witzgall.