We have a traverse survey methodology for terrestrial laser scanning, using total station, level and scanner observations, processed with least squares, for registration alignment, it provides very good results, especially in open areas where there are insufficient features for cloud based registration.
The scanner is a Topcon GLS-2000, capable of measuring prisms up to 200m, accuracy is what you'd expect from a 3 second total station. The scanner has a 1 second dual axis level compensator.
The scanner has a resection workflow, which allows you to move around freely, while targeting prisms mounted over control points, you can perform least squares adjustments on the resection, the vertical accuracy of the scanner compared a level isn't as good as we need.
The level gives us sub mm elevation accuracy.
With our current workflow, there's quite a bit of time required to setup tripods, put the scanner back in its case, move to the next location, leap frog the prisms, and measure side target control points. We measure instrument height in each position and come back and measure all the scan locations with the level. In each position we perform a low resolution full dome scan, then we take window scans of things we need to capture at higher resolution.
If we're pressed for time, we can use only the scanner and level, without the total station.
It takes us about an hour for each scan position.
Having used scanners with VIS and IMU's that assist in tracking the scanner location, these scanners can move through an environment much faster, with their light weight carbon fibre tripods, it only takes a few minutes for each position, however these scanners require sufficient features for registration and registration can take a long time if things go wrong.
I've been thinking about how I can speed up our scanning workflow.
The scanner itself only requires 1 minute and 45 seconds to perform a 12mm@10m full dome scan in greyscale. I've purchased a Ricoh Theta X, which I can mount on top of the scanner, it takes 360 degree 60MP HDR images in 10 seconds. Topcon makes a tubular compass, which I can use to point the scanner's azimuth to magnetic north, which will help me orientate the photo's later.
I've made contact with one of the carbon fibre tripod manufacturers, the tripod I've chosen is a hunting tripod, with a shoulder support carry pad, that makes it easy to move the tripod between scan positions, while keeping the scanner vertical. This is the tribrach support, underneath is a 25mm diameter Topcon type prism adapter.
This allows us to mount a prism underneath the scanner, so we can use the total station to track it's position. We could use a 360 prism, however a nodal prism is more accurate and it's easy enough to point it roughly at the total station, this is Sokkia's AP12 60mm prism:
However we still want to level all our scan positions, the prism blocks off the scanners laser plummet, we could use an offset prism to give the plummet a clear path to ground, but the instrument height on the scanner is only accurate to 3mm, which isn't good enough, so we normally use a Leica Height hook, but we don't want to remove the scanner from the tribrach, and we don't want to take angle instrument height measurements, the scanner has a trunnion axis mark, but it's not really good enough either as the tape has to bend around the scanner to reach it.
So I designed something similar to a height hook which attaches to the AP12 prism, so we don't have to remove it.
Continued...
