Faro Scene and importing survey points

To chat about anything else.
Leandre Robitaille
I have made 90-100 posts
I have made 90-100 posts
Posts: 92
Joined: Sat Aug 03, 2019 1:53 am
Full Name: Leandre Robitaille
Company Details: Cima+
Company Position Title: Civil Technician
Country: Canada
Has thanked: 5 times
Been thanked: 13 times

Re: Faro Scene and importing survey points

Post by Leandre Robitaille »

This made me chuckle!

This is one common issue with scene registrations when using surveyed coordinates.

First thing, stretching the survey to the coordinates, aka controlling the error with a total station. To do this the ONLY WAY as of now is to do the whole survey with targets. This means spheres, chekerboards etc. For large projects your need a very elaborate target placement, leave targets behind and closing your loop to make sure it is balanced and hope the hell none of your hundreds of sphere placements moved. Sphere detection needs to be good as well, you can't have dirty or wet spheres and they cant be too far way or you wont have enough point density on those spheres. A good density would be in the 1000+ points on a sphere. I dont trust the chekerboard detection algorythms with the faro scanner, even at 7m away from the scans I have suspisions on the accuracy of a chekerboard target. For your survey I would use spheres for all your scan placement and chekerboards for your controlled points at the perimeter of the survey. You can still have cloud2cloud clusters done within your registration, the important point is to have those clusters linked with the others scans using targets. I also recommend leveling the faro scanner with a calibrated tribrach and not using the internal inclinometer. Now the hard part to understand and plan out is to have all your scans somewhat on the same level (Not split in 3 big clusters), they have to be in same cluster on the same level, all targets forced together between each scan, force the chekerboards from your scans to the appropriate .csv reference imported, so that when you do your target based, your control points spreads out the error between the scans.

It is not possible to stretch out the error with forced targets if doing a big cloud2cloud of all your survey (it would be amazing, but it is not possible as of yet, your survey will most likely do a U shape if doing a cloud2cloud, if it is large and you might get up to meters in error in Z value between the middle and the ends)

Now how the hell do you bring in those targets and not have them move?

Create a CSV file, has to be PXYZ and ; seperation

Drag and drop that csv file in your project, import on worskpace level. Now check the value for xyz of one of the coordinates, it didnt change yet amazing!

Have your cluster look something like this;

MASTER CLUSTER
-Scan manager
-scan1
-scan2
-scan3
-scan4
-scan5
-scan6
-scan7
-scan8
-maybe a random cluster of a few scans that you did a c2c
REFERENCE FOLDER

Have the reference not be in the master cluster, drag and dropping the reference in a cluster is where you might change your targets XYZ value since the cluster applies its transformation to that reference folder, have it be on the highest level, same as your master cluster.
You can use the force correspondance tool to force targets from one scan in the master cluster to a target in the Reference folder, it will work during your target based registration even if its outside your master cluster.

Once forced, redo the target based registration with inclinometer.

Look at the results, do a clipping box in correspondance view and wash your eyes with slightly badly registered scans cause maybe 1 sphere is badly detected but you cant figure out wich one it is because in scan manager they are all under 3mm tension, but your clipping box shows your allignement is not perfect, maybe the surveyor didnt measure that chekerboard accurately? what about those 7 stations that were done with that total station and how he spreads the errors between all the targets maybe thats where the error is from? Maybe the surveyed targets are fine and you should have spent 20min per scan and spend weeks on site to rescan all the targets with the rescan tool for spheres at 1/1 density 4x and create those awesome cluster folder per scans to get a higher precision on those spheres and spend hours trying to figure our what sphere corresponds with what cause you cant see them in the quick view they show up as individual scans and forcing the spheres with one another is extremelly tedious, but maybe then your target registration with controlled point would have been better and all you can do is wish there was a way to optimize a cloud2cloud registration with control points that this would make your life so much more simple.

Hope this helps

Rant over

-Leandre
fobos8
Forum Supporter
Forum Supporter
Posts: 145
Joined: Sat Jul 07, 2018 9:19 pm
Full Name: Andrew
Company Details: Medina Surveys
Company Position Title: Surveyor
Country: Uk
Linkedin Profile: No
Has thanked: 16 times
Been thanked: 12 times

Re: Faro Scene and importing survey points

Post by fobos8 »

Hi Leandre

Great rant and very informative. I had a chat with a Faro trainer today who said the same thing - that for this type of transfer to survey control to work you must have targets for registration and that c2c won't work.

Is it worth it doing it this way - seems like a hell of a lot more work?

Cheers, Andrew
Leandre Robitaille
I have made 90-100 posts
I have made 90-100 posts
Posts: 92
Joined: Sat Aug 03, 2019 1:53 am
Full Name: Leandre Robitaille
Company Details: Cima+
Company Position Title: Civil Technician
Country: Canada
Has thanked: 5 times
Been thanked: 13 times

Re: Faro Scene and importing survey points

Post by Leandre Robitaille »

Hi Andrew!

I wouldnt botter with it if your survey is small. I would only worry about it if you have a linear survey or if you cover a lot of ground.

There are other specific instances where you want to spread the error, there is a way to manage this with c2c, but sometimes using targets is the only solution. For exemple you could survey the whole interior of a building using c2c registration, but lets say you are in an industrial environnement and only a 10mm steel sheet is separating indoors from outdoors. When you are going to try and do a c2c for the scans outdoor they will try to fit the steel sheet as if it was the same wall and you might get errors from that. Hence placing spheres to survey the perimeter of the building is a good idea, then controlling the outdoors with targets in each doors captured by the interior survey. This is only an example, you have to addapt to every type of surveys.

Another alternative I heard from was Trimble Realwork, I think you can control the c2c registration using surveyed targets in that software.( someone once told me this but I never tested this).

Cloud2cloud registration does not take into account the inclinometer of the scan, it only averages the results at the end.

TLDR ;
small survey - c2c
Levels are important? - Can be c2c, but control with total station.
Big survey or linear : consider targets and controls.

Also note that the Faro scanners has its specific local coordinate related to the levels of its reference scan, if you would scan the whole planet your X and Y axis would be the same and Z values would not be from sea level. You can't use the earth angular compensation when controlling 3d scanning targets. I think it was something like 2cm error in Z over 2km but I might be wrong, if you were to compensate the targets with the earth's curve (I didnt personnaly use a total station for years, my collegue does the total station part for our 3Dsurveys). I would recommend doing freestations and do 3 points foward and back shot and not using the compensator of the total station when doing long distances survey control for a faro scanner. Once again, I think trimble doesnt have this issue.
RealVisuals
I have made 10-20 posts
I have made 10-20 posts
Posts: 16
Joined: Fri Jul 06, 2018 5:48 pm
Full Name: Vincent
Company Details: 3D scanning
Company Position Title: CEO
Country: Belgium
Linkedin Profile: Yes

Re: Faro Scene and importing survey points

Post by RealVisuals »

We have a similar project, where the staircases are the reference folders and we want to put the floorclusters in the higher level of the structure folder.
I did a test with C2C and it seems that the references remain in place. This project is too big and we want to split the data.
Normally we use it on the workspace level
You do not have the required permissions to view the files attached to this post.
Post Reply

Return to “General Chat”