Anyone have recommendations on scanning software? I'm looking at getting us into scanning and I'd like to hear what people are using and why; what you like and dislike.
I have only seen various versions of the Cyrax/Leica package called Hurricane or Typhoon or something like that, nothing in the past 5-8 years, but the "DECIMATE" command seemed to be a fantastic tool to make the data cloud actually manageable by a human. The raw clouds are way too dense to be useable, in my opinion.
Perhaps they have continued to improve that package ... as long as Americans are doing the programming and not the Swiss!
you can find lots of information here:
http://www.laserscanningforum.com/
Leica cyclone is commonly used from what I observed
Kubit did a real time webinar demonstration for us, with a point cloud that we supplied.
I don't know how they compare, but they work as an ACAD module.
"...The raw clouds are way too dense to be useable, in my opinion..."
The density of the point cloud is what makes it valuable. I have yet to see any manual surveys surpass a laser scanner in industrial or commercial as builts.
Random errors begin to drift away to nothing considering the number of points collected. Using a laser scanner will make you rethink what you REALLY know about 3d measurements.
> Random errors begin to drift away to nothing considering the number of points collected.
I don't think I'd characterize it that way. Blunders may disappear, but the instrument's angular and distance accuracy limits will continue to introduce error into the point cloud. Nearby points may be tightly-correlated, but that's more a matter of precision than accuracy.
> Kubit did a real time webinar demonstration for us, with a point cloud that we supplied.
>
> I don't know how they compare, but they work as an ACAD module.
You can't register a scan with Kubit.
Cyclone is the most used and probably the most expensive. It's what we have. We like it's functionality but our workflow requires us to further supplement the cloud by processing in a second package namely Amberg Tunnel Management
Jim
In concept Random error is closely related precision. When measuring, one of the ways to tighten your measurements, or, lessen the effect of random errors is to repeat your observations and average the outcome.
When Laser Scanning a surface for instance, instead of repeating the same angular and distance observation, the Laser Scanner repeats the measurement across the entire face of the surface. And, thus we have averaged the readings across the face of the surface.
Try looking at the 61 software venders attending SPAR
Many of the software companies have passed the Leica product Cyclone. Try FARO, LFM, Areva, All Point Systems, etc.
Leica Cyclone is the one that I've always used and the one that other companies I've worked for used as well.
We have used Faro Scene to process scans and export to different formats.
Autocad Recap can view and manipulate processed scans easily if you have Civil3d 2014 or 2015. Kubit does modeling from processed scans in quick fashion for cad work. The software is changing so quickly you really need to see which firm's latest and greatest version does what you need.
> The software is changing so quickly you really need to see which firm's latest and greatest version does what you need.
:good:
Once you've processed/registered the scan data in the software specific to your hardware; the question is less "what software do you like" and more "what software produces the best result for the deliverable/client you're aiming for".
Your software decisions are going to be driven by what scanning market segment you focus on.
That is true it does average out the surfaces. But the edge of objects, such as pipes, walls, equipment and such are still based on the timing and rotational accuracy of the scanning device. It is assumed, and many times rightfully so, that the scanner's points are all relative and accurate within a 1/4" throughout. That holds true up to a certain distance. For example, the scanner's point density is 1/4" at 30-40 feet typically, so it is hard to justify 1/4" accuracy past 30-40 feet and considering that the surfaces built are based on the points then point density rather than point 'precision' is more critical than our surveying notion of a points relative positional accuracy.