RANGE PRECISION AND ACCURACY OF TERRESTRIAL LASER SCANNERS

Accuracy and precision of terrestrial laser scanners (TLS) is widely unknown at long ranges, with manufacturers providing a set of values for one distance, generally 100 meters, without information on how the trend varies with range. In order to allow researchers to better anticipate, quantify, and report these errors, a series of tests was conducted as a part of the NSF-funded Geo-Launchpad Internship Program to determine a scanner’s ability to detect a 1 cm motion of a highly reflective flat target at distances ranging from 100 to 900 meters.

The standard deviation of precision error and average accuracy error in the direction of the range were consistently low (~1 mm), as they are usually dependent only on environmental conditions that affect the speed of laser travel. Precision and accuracy error for target movement perpendicular to the range direction generally increased by a factor of 8 at a range of 900m. Larger error was expected, as the footprint of the laser beam grows with range, and errors in mirror and scanner head alignment are magnified with distance. Notably, the scanner picked up a damping oscillating pattern in the target’s vertical position over a set of 55 tests in which the target was stationary in respect to the vertical direction; this may be due to the instrument’s internal tilt sensor calibration improving as successive scans were performed. Environmental factors, such as the presence of dust particles in the air, appear to have a large impact on scanner precision and accuracy, especially in directions orthogonal to the range.

Additional testing will be carried out to determine whether the error trend is linear or exponential, controlling for environmental factors and using surveying equipment such as a Total Station to provide precise coordinates of the target. Further testing with a larger pool of instruments will also be performed to further understand anomalies detected in the vertical axis.