Website Content by Liying Huang 2013

Copyright © 2013 UNSW. Designed by LIH

Text Box:

BORESIGHT AND SCANNER CALIBRATION OF A MOBILE TERRESTRIAL LIDAR SYSTEM

SCANNER CALIBRATION

We all know the significance of the equipment calibration and check if the accuracy of the equipment is within the tolerance regularly. For the scanner in MLS, which uses line mode measurement, it is different. This kind of test is done in a strict factory test by manufacture. There is no common method to test the performance of the scanner besides independent measurement by using other techniques. The users need to send back the scanner every year for service. It not only takes time to do so, but also may take a risk. How about like the graph shown, after comparing with the road profile from a 3D terrestrial scan at the control, we find this difference and it is already close to the deadline? In the Queensland Project early of this year, two independent scan with different scanner, Riegl and Z&F with similar specification, are completed for the same road section. There are 25mm over 20m difference. Since Z&F agree with the road profile, an inference is Riegl is out of calibration. If it is right, two months will be taken to send back the scanner to Austria for service. Therefore, it is necessary to construct a local test to determine the performance of the scanner prior to the project or after the project and find the corrections to scan data if the scanner is out of calibration.

Let us know our target scanner Riegl VQ250 firstly. As you can see, itís a typical 2D-line scanner to rotate around its z axis, the scanner head rotates at 100 Hz and measures 300k / sec. The output of the scanner is a X, Y and Z coordinate relative to the centre of rotator head spin axis as well as intensity.

The most direct error source is the direct range errors. It can be a function of many variables such as beam divergence along distance, intensity, angle of incidence, refraction through the atmosphere. For instance, the beam spot size become larger and larger from close to far; the measurement return may not reflect the same point as in the close. If the scanner is in good working order, the z values for the measurements should be zero since it is assumed that the output from the scanner has been subject to the factory calibration corrections regards to small rotation errors in the spin axis. In other words, a 360 measurement should form a plane which is perpendicular to z axis. However, if the mirror is not setting properly, as the graph shown, cone shape is supposed to be constructed.