There are two different types of technology behind static terrestrial laser scanners. One technology is time of flight (TOF), scanners such as the Leica C10 or Riegl VZ-400 use this technique. The other is Phase-shift (sometimes referred to as phased based) – which Leica HDS7000 and the Faro Focus 3D use.
A simple way to find out if a scanner is phased based or TOF is by learning how quickly a particular scanner is able to scan. If the scanners scan rate is around 1 million points per second, then it is phased based, while typically TOF scanners will have a slower scan rate of between 50k and 150k per second*.
Most assume that the technical differences between TOF and Phase-shift are merely an issue of range and speed. And while that is partially true, there are important practical differences that should be understood before sending either out on a project.
TOF is terrible at close range
The accuracy of a TOF scanner at less than 6 feet is very poor. I learned this lesson the hard way. While I was still in school, we were tasked with scanning an interior engine room of the USS North Carolina (documented here). We had just received delivery of two Leica C10 TOF scanners, and were brimming with excitement to deploy them. We scanned for half a day with two scanners in the engine room, which (because of the narrowness of the walkways) required the scanner setups be located in close proximity to the features we wanted to capture. After registration we noticed that the data was almost unusable (because the point cloud had many centimeters of noise). Due to the data quality issues the mesh model we had planned to create in 3DReshaper was scrapped.
Phased based data can be weird
In my experience TOF point clouds are less noisy, and more consistent in the noise they produce. Phased based point clouds however, tend to be more noisy and also produce a range of spurious points that are not always predictable or consistent.
As an aside, when I speak about noise, my mind considers noise in two separate categories; measurement noise and spurious point noise. Measurement noise is caused by the inaccuracy of the scanner as a measurement device; when a Leica/Faro/Reigl rep scans a wall, fits a plane to the wall point cloud and determines that all points are within 1-3mm of the plane – that is measurement noise. Spurious point noise represents a (usually) much larger measurement error whereby the scanner receives a false laser return. Spurious points include things like edge effects and reflections.
At the scale of a building, measurement noise created by either phased based or TOF is more or less insignificant regarding the definition of building features. However, the difference in spurious point noise between phased based and TOF scanners can be problematic. Firstly, if you’re used to dealing with TOF data and are tasked with modeling or cleaning a phased-shift generated point cloud it may be difficult in some areas to differentiate between the real and false returns, so it is important that you have accurate and detailed field notes and pictures. Secondly, understanding the noise your scanner creates can alter the location of your setups in the field. For instance, a C10 can shoot through most glass windows and acquire accurate point cloud data – however, our HDS7000 usually cannot. Meaning, I need to add additional setups inside a room with windows when I use the HDS7000.
Phased based scanners dont have true dual axis compensators
A dual axis compensator is the device that accounts for scanner movement while the scanner is collecting measurements. TOF scanners have true dual axis compensators, like a total station – while phased based scanners use some form of less accurate tilt sensor. This is a controversial subject, and we can argue about the details; but compare the spec sheet of the C10 and the HDS7000 / Z&F 5010 and youll find a less accurate dual-axis compensator on the HDS7000. What does this mean? It means you should consider using external control on your phased based scanner projects.
TOF and Phased based scanners cant use all the same targets
We often use our C10 in conjunction with our HDS7000 on large building projects, where we will scan small repetitive inside rooms with the faster HDS7000, and use the C10 for longer range scans along the facade. When using both on the same project it is important to remember that the HDS7000 cannot pick-up Leica HDS targets. Both can use spheres and/or Black and white targets however.
* one caveat to this is the recent release of the Leica P20, which is a TOF scanner with a scan rate of 1 million points per second, albeit with a significant reduction in scanner range compared with other TOF scanners.
Thad Wester is the founder of Clarity Scanning. Look for him at Spar International 2013, where he will be giving a presentation on 3D printing from Point Clouds