FARO S150 Leveling

Discuss all FARO related issues here.
tbwester
V.I.P Member
V.I.P Member
Posts: 878
Joined: Fri Apr 20, 2012 1:49 pm
Full Name: Thad Wester
Company Details: WeWork
Company Position Title: Head of Reality Capture
Country: USA
Linkedin Profile: Yes
Location: SC/NYC

Re: FARO S150 Leveling

Post by tbwester » Fri Jan 11, 2019 8:59 pm

Slewis33 wrote:
Fri Jan 11, 2019 7:37 pm
Does Cyclone tilt e57s? I think we found the culprit.
It tilts any file from a faro even if you run through scene first.

Slewis33
I have made 40-50 posts
I have made 40-50 posts
Posts: 47
Joined: Thu Jun 06, 2013 2:50 pm
Full Name: Shaun Lewis
Company Details: Clark Construciton
Company Position Title: Senior Manager - Reality Capture
Country: USA
Linkedin Profile: Yes

Re: FARO S150 Leveling

Post by Slewis33 » Fri Jan 11, 2019 9:21 pm

10-4. We will get creative and spend a little more time to level it out. ;)

steveholly
I have made <0 posts
I have made <0 posts
Posts: 4
Joined: Mon Nov 30, 2015 2:14 pm
Full Name: stephen holly
Company Details: Pure Real Estate
Company Position Title: Architectural Technician
Country: England

Re: FARO S150 Leveling

Post by steveholly » Sun Jan 13, 2019 1:18 pm

Slewis33 wrote:
Fri Jan 11, 2019 9:21 pm
10-4. We will get creative and spend a little more time to level it out. ;)
@Slewis33

We have the same issues with our Faro Unit. We did an external project of 200 scans over a distance of maybe 300m x 300m. When we checked this against our GPS and control points, there were some errors of 200mm! looks great in the x and y but the z-axis is tilted.

We have tried every registration software, Recap, Scene and Cyclone, every file type, E.57, RCP, PTX, PTG and the problem has not been solved. Please let me know if you have any success. First thing this week we will book in our scanner for a service to rule this out as the source of the issue.

KR
Stephen Holly
www.modellingarchitecture.co.uk

User avatar
JohnBunnFARO
I have made 20-30 posts
I have made 20-30 posts
Posts: 27
Joined: Thu Aug 11, 2016 6:18 pm
Full Name: John Bunn
Company Details: FARO
Company Position Title: Account Manager
Country: USA
Linkedin Profile: Yes

Re: FARO S150 Leveling

Post by JohnBunnFARO » Thu Jan 31, 2019 3:10 pm

steveholly wrote:
Sun Jan 13, 2019 1:18 pm
Slewis33 wrote:
Fri Jan 11, 2019 9:21 pm
10-4. We will get creative and spend a little more time to level it out. ;)
@Slewis33

We have the same issues with our Faro Unit. We did an external project of 200 scans over a distance of maybe 300m x 300m. When we checked this against our GPS and control points, there were some errors of 200mm! looks great in the x and y but the z-axis is tilted.

We have tried every registration software, Recap, Scene and Cyclone, every file type, E.57, RCP, PTX, PTG and the problem has not been solved. Please let me know if you have any success. First thing this week we will book in our scanner for a service to rule this out as the source of the issue.

KR
Stephen Holly
www.modellingarchitecture.co.uk
Hi Stephen,

This should definitely not be the case! Please make sure the inclinometer is turned on during processing/registration, and also a CSV file with the known control points would also automatically level this out.

Please feel free to reach out for assistance.
Half Nerd, Half Amazing

SAttaya
V.I.P Member
V.I.P Member
Posts: 310
Joined: Mon Mar 23, 2009 4:04 pm
Full Name: Steve Attaya
Company Details: Geocent
Company Position Title: Sr. Software Analyst [RemoteSensingEngr]
Country: USA
Linkedin Profile: Yes
Location: New Orleans, LA

Re: FARO S150 Leveling

Post by SAttaya » Wed Feb 06, 2019 8:58 pm

I don't know if this is a problem with the Sx50 series,
but in the 120 series, if you stopped a scan with the button
then the inclinometer would be "frozen".

So whenever we stopped the 120 (2012 calibrated every year) scanner, we knew we'd have to
power it down and back up again. Needless to say, we almost always finished the scan and
pitched it.

Guess I should try it with the S150. :lol:

PS: mostly, we just would check that it was still working before starting the scan.
(Manage>Sensors>inclinometer and manually tilt it out of vertical)
This also would catch the error where someone checked the level and as the last
thing would mean to push the GoBack button, but instead pushed the TurnInclinometerOff
button.

mikajahkola
I have made 10-20 posts
I have made 10-20 posts
Posts: 15
Joined: Wed Jul 12, 2017 9:27 pm
Full Name: Mika Jahkola
Company Details: Myymore Oy
Company Position Title: Managing Director
Country: Finland
Linkedin Profile: Yes

Re: FARO S150 Leveling

Post by mikajahkola » Fri Feb 08, 2019 9:55 pm

steveholly wrote:
Sun Jan 13, 2019 1:18 pm

@Slewis33

We have the same issues with our Faro Unit. We did an external project of 200 scans over a distance of maybe 300m x 300m. When we checked this against our GPS and control points, there were some errors of 200mm! looks great in the x and y but the z-axis is tilted.

We have tried every registration software, Recap, Scene and Cyclone, every file type, E.57, RCP, PTX, PTG and the problem has not been solved. Please let me know if you have any success. First thing this week we will book in our scanner for a service to rule this out as the source of the issue.

KR
Stephen Holly
www.modellingarchitecture.co.uk
Do I understand right that you registered the project cloud-to-cloud without targets, and then tried to match the cloud to control points but got large mismatches in z-direction?

I have seen this phenomenon many times. Cloud-to-cloud registration tends to drift so that scans get more and more tilted as you go farther from the starting point. I have tested Recap, Scene, and Cyclone 360 Register, and all show the same effect. Which is worse, you get different results depending on your starting (reference) position.
I have also tested Trimble Realworks. It gave much better results in this scenario, minimal drift in z direction.

There seems to be one fundamental difference in how the products refine cloud-to-cloud registration.
Realworks levels each FLS scan on import based on the inclinometer value, and keeps the scan upright through the entire registration. The other products (R,S,C) let the scans rotate three-dimensionally when optimizing registration. Only one scan (reference) is level, all other scans are floating freely and just matched to each other. I have found no way around it. Recap and Cyclone 360 Register don't give the user any choice at all, registration will always be refined and the scans will float freely. Scene gives the user more options but there is only one way Scene will optimize the registration, namely by ignoring the inclinometer and rotating scans three-dimensionally.

Registration refinement using all 6 degrees of freedom may give the best fit locally, but it can lead to big global errors towards the edges of the registered cloud. You can see this if you look at the Scene registration report. It shows the inclinometer mismatch values per scan, which can get way higher than the specified inclinometer accuracy 0.019 degrees. A significant inclinometer mismatch indicates that the registration is obviously wrong. Even though Faro's inclinometer is not super accurate, it would anyway keep the error within certain bounds, if it would be used in registration.
The Realworks approach works better but there you have to accept that neighbouring scans are not perfectly matched. In indoor scans with 10 m typical visibility you probably don't see any difference but outdoors with longer ranges small alignment errors start to appear.

I think the optimum solution would be that registration would be refined using 3-dimensional rotation but inclinometer mismatch would be used as one error term in the process, preferably so that the user could set the relative weight of the inclinometer.

Post Reply

Return to “FARO”