Some buildings I've measured were just room sizes and wall thicknesses.
Disto is brilliant for those.
Measure from both ends, sides as a check.
I also stress where total station or similar isn't used there's no guarantee of verticality or squareness.
But, like most jobs, imperative on these are photos of doorways and where wall meets floor /ceiling, window openings.
Some of those ornate old chrome (or whatever they are) windows require careful measurements.
Skyline facades are interesting too, need to reproduce similar.
I don't have reflector less and did some intersections from 3 setups which proved themselves.
Expect the scanners would come into their own then.
I'd never make one of those pay for me.
Despite the quote, we got the job. I should have gone higher!
We are considering hiring a Geoslam ZEB1 to try.
http://geoslam.com/hardware-products/zeb1/
Has anyone here used a ZEB1 or heard anything about them?
I've rented a zeb1 once and found it to be usable for initial measuring but no detailing. I didn't have time to test it out and had to do my survey at once so i think a little expermenting would make your data better. I processed data with pointcab software.
You should rent a revo instead, this gives much better data with less noise. For high detail measuring you should use a traditional scanner.
Revo is a lot faster beacause of no setups.
If you're interested i can give you a downloadlink for some sample data.
beuckie, post: 379521, member: 2245 wrote: I've rented a zeb1 once and found it to be usable for initial measuring but no detailing. I didn't have time to test it out and had to do my survey at once so i think a little expermenting would make your data better. I processed data with pointcab software.
You should rent a revo instead, this gives much better data with less noise. For high detail measuring you should use a traditional scanner.
Revo is a lot faster beacause of no setups.If you're interested i can give you a downloadlink for some sample data.
Hi I am also looking into getting my hands into scanning. May have a job similar to his but on a smaller scale. Would like to find out the ease of use of scan data.
Registration of scan data is all quite straight forward, but my concern is from the point cloud to CAD drawings, how easy and straight forward is it?
My client still want standard CAD drawings and not point cloud models.
If you could share your sample data that would be good.
Cheers
Seb, post: 379487, member: 7509 wrote: Despite the quote, we got the job. I should have gone higher!
We are considering hiring a Geoslam ZEB1 to try.
http://geoslam.com/hardware-products/zeb1/
Has anyone here used a ZEB1 or heard anything about them?
Gavin Schrock has used one, and was quite impressed. Someone else indicated that they had closure problems, simply map outward and make your way back to your starting area for overlap. Loops and overlaps will help with LIDAR mapping.
Check Gavin out for further details on the ZEB1.
Leica makes a Pegasus on a backpack as well. Two LIDAR heads, five cameras, Novatel IMU, and multi band GNSS (I know it will not help inside, but it is there). My expectation is that this instrument can be processed in relative coordinates, so it would still be very powerful.
Seb, post: 379487, member: 7509 wrote: Despite the quote, we got the job. I should have gone higher!
We are considering hiring a Geoslam ZEB1 to try.
http://geoslam.com/hardware-products/zeb1/
Has anyone here used a ZEB1 or heard anything about them?
I have seen results etc; posted on linkedin. Looked impressive (from a distance). I suggest you contact the company and find out what the state of the art is. I imagine the technology is evolving rapidly.
sireath, post: 379522, member: 9370 wrote: Hi I am also looking into getting my hands into scanning. May have a job similar to his but on a smaller scale. Would like to find out the ease of use of scan data.
Registration of scan data is all quite straight forward, but my concern is from the point cloud to CAD drawings, how easy and straight forward is it?
My client still want standard CAD drawings and not point cloud models.
If you could share your sample data that would be good.
Cheers
Scanning data from a Faro is imported in or Pointcab or Autocad through Pointsense add-ons.
The scanning data from a zeb1 is so dense with noise that in autocad it results in a dark cloud. These data were processed in Pointcab (German software).
The scanning took approx. 15 min.
beuckie, post: 379527, member: 2245 wrote: Scanning data from a Faro is imported in or Pointcab or Autocad through Pointsense add-ons.
The scanning data from a zeb1 is so dense with noise that in autocad it results in a dark cloud. These data were processed in Pointcab (German software).
The scanning took approx. 15 min.
This is the result in Autocad (very simple layout plan)
beuckie, post: 379529, member: 2245 wrote: This is the result in Autocad (very simple layout plan)
What kind of accuracy did you get using this scanner?
1 to 1,5 cm
Thanks all.
The company here in Australia does not have a Revo for hire or we would have got that one.
Part of their hire fee is processing a certain amount of data, and I fully intend to walk though the job making spot checks with a disto before providing the finished product to the client.
We are also going to try it out on a clay stock pile job at a brick factory that we do every six months. New and exciting gear is fun. I'm hoping it all works out as this has the potential to be very useful.
beuckie, post: 379529, member: 2245 wrote: This is the result in Autocad (very simple layout plan)
This is interesting. For the floor layout did you manually draw out the lines or is there a software to automatically do it for you?
Quite interesting to note the amount of noise generated.
I hate to be the downer, but I feel that I must warn those not yet involved with scanning. It is an amazing tool amd can collect more data in a minute than a gun in a lifetime. Example : We have tested a Viametris (french company) who has a cart based MLS and trigger based, backpack 360 hand scanner. This data was compared to the existing terrestrial scanning done with a P40. The backpack scanner was so-so. The cart was better but need 2x the control of the P40. Then data was better but heights were useless as IMU only records horizontal data. So P40 was best. Problem is using that data
AutoCAD has feature extraction but it is not great for outlines like interal wall surveys. There are other plugins that do well. UnDET and pointssense do a good job.
My point being that scanning is awesome but QC is critical, remember revit hates large coord values, scan data is many times big so get a proper pc(think dell precision). A pointcloyd is useless if you dont know how to get data out and dont think that you will be effective if you haven't tested it and sorted out the kinks before you do a project.
It is a different game ti traditional survey
Enjoy and go big!
Sent from my D6503 using Tapatalk
Seb, post: 379619, member: 7509 wrote: Thanks all.
We are also going to try it out on a clay stock pile job at a brick factory that we do every six months. New and exciting gear is fun. I'm hoping it all works out as this has the potential to be very useful.
I'll be interested to see how that works out. I've seen some attempts at scanning heaps, but they all miss the point that you need to be above the stockpile to pick up any hollows in the top. Not so much of a problem with a nicely graded stockpile, but a disaster on a heap when some material has just been tipped and is full of depressions, or indeed on a nice graded heap with an edge bund where the scanning has been done from ground level.
For any form of scanning you still need to walk nearly all the surface, although it would save going into deep holes.
You might find the REVO has difficulty in resolving the surfaces due to lack of definitive points, possibly depends on having enough striations on the surface to aid matters.
We have flown stockpiles using an SUA and the results are very good, but obviously it is viewing a larger area from above and pattern matching then isn't a problem, plus we have plenty of control points. Cost wise it isn't really effective unless the heaps are very large (over 0.25M tonnes) or very difficult to walk.
If you have a pointcloud of the interior of all the rooms and the client wants a 2D floor plan then the way it is done in Cyclone is you do a horizontal clipping plane at some convenient height above the floor and below the ceiling. This way you only see a slice through the building (say 1' thick...the user sets this). Then you use the 2D drawing tools such as 2D polyline which draws on the active reference plane (which I use to set the clipping plane). The polylines can be moved to zero y later. Set the view to oblique (instead of perspective) and you can see the walls easily).
The other thing clipping planes are handy for is cleaning to ground. You set the clipping plane vertical, then at some convenient width (on large projects I have used as much as 20') you can turn the thing up on end and clean the weeds down to the ground. If I have the width set at 20' then I set the raise/lower distance a little bit less, say 19.5' then bump the clipping plane when each cross section is down. We have also forgone the cleaning and just used virtual surveyor to put points in along the cross section. With a narrow cross section is easy to see the bottom of the point cloud.
Topoing on the point cloud by adding vertexes then exporting those to an ascii file makes a better looking surface in AutoCAD and much smaller in size then the massive things you get out of Cyclone even if they are decimated down.
If anybody wants some sample data, send me a mail to info (at) lsbbvba.be and i'll send a downloadlink.
This is data from the first usage off the Zeb1 and not representative for data from a regular user. the data is as received from geoslam (laz format).
Saturday morning was spent walking around with the hired ZEB1.
These images are made up of four separate scans that we have pieced together roughly.
Checks with the total station are still to be done but it all looks promising.
Pretty good for a mornings work I think.
The clay stockpile job is on Wednesday.
Looks good (from a distance at least!). Amazing how much data you can collect in such a short space of time.
I would be particularly interested to know what software you are using and how it handles
a) registering and adjusting the scans to fit co-ordinated control points
b) extracting line work from the cross-sections to DWG
and how long such a process would take.
I imagine you will be doing some spot checks on the accuracy of co-ordinates. Would be interested to see what differences you get.
Seb, post: 380682, member: 7509 wrote: Saturday morning was spent walking around with the hired ZEB1.
These images are made up of four separate scans that we have pieced together roughly.
Checks with the total station are still to be done but it all looks promising.
Pretty good for a mornings work I think.
The clay stockpile job is on Wednesday.
That looks really good. I have got to try that scanner out. Would be interested to know what software you use for processing and the accuracy achieved.
I don't suppose you would be willing to share the scan or part of it?
Totalsurv, post: 380695, member: 8202 wrote: That looks really good. I have got to try that scanner out. Would be interested to know what software you use for processing and the accuracy achieved.
I don't suppose you would be willing to share the scan or part of it?
I think the software supplied for use with this scanner is GeoSlam Desktop.
This registers the scans using the internal IMU (inertial measurement unit) logs.
I don't know how much further processing can be done without transferring the point cloud to another package though.
