Hi,
We have a garage (100m²) with cracks on the ground and a Faro s70 and I want to do cloud comparison to see the evolution of the cracks.
What settings do you recommend when scanning with Faro s70?
Thanks in advance for your help.
Cedrickb
What resolution to use when doing structural comparison
- Justin Richards
- V.I.P Member
- Posts: 159
- Joined: Wed Jan 27, 2021 9:03 pm
- 3
- Full Name: JUSTIN RICHARDS
- Company Details: Tribrach Solutions
- Company Position Title: Survey Tech
- Country: United States
- Linkedin Profile: No
- Has thanked: 38 times
- Been thanked: 51 times
- ProCro
- V.I.P Member
- Posts: 151
- Joined: Thu Jun 27, 2019 2:00 pm
- 4
- Full Name: Nino Skupnjak
- Company Details: SKIMI64
- Company Position Title: procurator
- Country: Croatia
- Linkedin Profile: Yes
- Location: Croatia
- Has thanked: 7 times
- Been thanked: 30 times
- Contact:
Re: What resolution to use when doing structural comparison
There is really small difference between 1/4, 2x (6.1mm/10m), and 1/5, 2x (7.7mm/10m). Is it worth extra time?
- Leandre Robitaille
- V.I.P Member
- Posts: 542
- Joined: Sat Aug 03, 2019 1:53 am
- 4
- Full Name: Leandre Robitaille
- Company Details: Cima+
- Company Position Title: Civil Technician - Surveyor
- Country: Canada
- Has thanked: 61 times
- Been thanked: 246 times
Re: What resolution to use when doing structural comparison
Yes. 1/4 is already at the minimum density for good sphere detection and propper accuracy.
1/5 is just a little short for sphere detection at proper range and should be used with knowing the limits you want to work with.
As for cracks evolution I dont think a 3d scanner is the tool you want to use. There are many alternatives with more sensitive accuracy you can use for crack monitoring
-
- V.I.P Member
- Posts: 544
- Joined: Fri Feb 24, 2017 10:47 am
- 7
- Full Name: Martin Graner
- Company Details: PointCab GmbH
- Company Position Title: Research and Development
- Country: Germany
- Linkedin Profile: No
- Has thanked: 160 times
- Been thanked: 175 times
- Contact:
Re: What resolution to use when doing structural comparison
Based on the sphere detection algorithm in our software I did some theoretical distances where spheres should automatically get extracted, using a 145 mm diameter sphere.Leandre Robitaille wrote: ↑Fri Jul 29, 2022 2:27 amYes. 1/4 is already at the minimum density for good sphere detection and propper accuracy.
1/5 is just a little short for sphere detection at proper range and should be used with knowing the limits you want to work with.
[...]
With 1/5 resolution they should get detected at 23 m, 1/4 should get detected at 29 m, guess the same numbers are roundabout valid for other software packages.
-
- V.I.P Member
- Posts: 401
- Joined: Mon Mar 23, 2009 4:04 pm
- 15
- Full Name: SAttaya
- Company Details: Sev1Tech
- Company Position Title: Sr Software Analyst-RemoteSensingEngr
- Country: USA
- Linkedin Profile: No
- Has thanked: 127 times
- Been thanked: 17 times
Re: What resolution to use when doing structural comparison
Martin,VXGrid wrote: ↑Fri Jul 29, 2022 11:24 amBased on the sphere detection algorithm in our software I did some theoretical distances where spheres should automatically get extracted, using a 145 mm diameter sphere.Leandre Robitaille wrote: ↑Fri Jul 29, 2022 2:27 amYes. 1/4 is already at the minimum density for good sphere detection and proper accuracy.
1/5 is just a little short for sphere detection at proper range and should be used with knowing the limits you want to work with.
[...]
With 1/5 resolution they should get detected at 23 m, 1/4 should get detected at 29 m, guess the same numbers are roundabout valid for other software packages.
I'm interested in how you came up with the 29m @ 1/4 resolution:
* what is it for 10m, 15m, 20m & 25m
Any chance you'd like to describe your sphere center point algorithm?
-
- V.I.P Member
- Posts: 544
- Joined: Fri Feb 24, 2017 10:47 am
- 7
- Full Name: Martin Graner
- Company Details: PointCab GmbH
- Company Position Title: Research and Development
- Country: Germany
- Linkedin Profile: No
- Has thanked: 160 times
- Been thanked: 175 times
- Contact:
Re: What resolution to use when doing structural comparison
So this is not totally sound:
The first assumption is how many points do we need to have on the sphere to automatically extract it. I used the value of around 50 points here, because then you can take a 7x7 pixel grid aka 49 points. More is of course better, less should work as well, the 50 works well, I have seen it doing great with a 5x5 patch as well.
Then we have the scanner resolution, 1/8 is around 5000 columns to 2125 rows (this may vary as well), taken from a FARO Focus S350.
There we have a point spacing of around 0.001256 m in 1 meter distance to the scanner (I didn't check if the point spacing varies in between rows and columns, I used the columns, because I thought rows might be denser, since this is the rotation direction of the laser, and columns is the direction of the complete device).
Now more simplifications:
Sphere diameter of 0.145 m is a circle area of 0.01651 m².
Pi * 0.145^2 / 4
Treating this area not as a circle, but as a square, we have a square side length of 0.1285 m
sqrt(circle area)
Now one side of the square needs to equal 7 points, or to put it this way: On the rectangle we need every 0.01835 m one single point
square side / 7
Lastly we need to consider the point spacing to get the distance
single point / point spacing in 1 meter distance
14,608 m
--------------------------------
Things which make this theoretical value worse:
- Considering point spacing and area size, a square and a circle won't "collect" the same amount of points
- Treating a sphere as a circle and discarding the depths
- Sampling issues
- The need to discard bound points due to ghost effects
As I have written, for automatic extraction, I mean if we do manual extraction a lot less points are feasable to get "a" sphere center. The more redundant points on the sphere, the better the sphere middle point and sigma estimation of course.