Updated: Sep 12, 2018
As with just about any technology that requires precision — from gas pumps to inkjet printers — you can’t count on LiDAR collection without proper sensor calibration. All sensors should be calibrated routinely, and calibration is recommended after every installation into a platform. How you calibrate depends on the system and the LiDAR provider, but the results are similar.
In airborne systems, there’s not a big difference between the calibration processes for fixed-wing and helicopter-mounted systems. Terrestrial LiDAR scanners, however, have self calibrations that take place before every collection.
In addition, all LiDAR systems (whether in the air or on the ground) are initially calibrated by the manufacturer in a lab and during field trials. Whenever a hardware component is repaired or replaced on a LiDAR sensor head, the system should be calibrated. Just as the brake shop will test drive your car after replacing the pads and turning the rotors, all LiDAR providers will do a calibration test flight or drive after doing any kind of repair to the LiDAR sensor.
It takes a good bit of flying to do a system calibration for an airborne system, typically at three different altitudes based on the LiDAR system configuration. The calibration flight will include a series of lines flown over an object, such as a building or a runway. The process usually involves flying one or more flight lines in opposing directions to determine differences in all the LiDAR calibration parameters. These differences or errors will be corrected by using the relationship between flight lines and ground survey information.
The things that get corrected the most are roll and pitch differences, though additional corrections will be made to heading, range correction, pitch slope error, and scan factor or torsion. If you want to understand scan factor or torsion, imagine if a LiDAR swath was profiled perpendicular to flight direction. The scan would appear to have what looks like a frown or a smile, when in reality the scan should follow the terrain. Pitch slope error is detected by seeing a building at nadir (or vertically below) in relation to the scanner in two opposing flight lines. The placement is exactly replicated with both flight lines but, at the edge of the swath, buildings or objects are offset in flight direction.
Range error occurs when the range or distance of the pulse isn’t correct from the sensor to the object it is hitting. This is often confused with an elevation correction of the sensor, which is a uniform difference of the absolution elevation of the survey of the ground and the reading of the LiDAR system of the ground.
Young, James. LiDAR for Dummies. Hoboken: Wiley Publishing, Inc., 2011.