Outdoor scan for ortho pictures

Discuss FARO hardware here.
Post Reply
myvisuals
V.I.P Member
V.I.P Member
Posts: 122
Joined: Tue Oct 07, 2014 12:05 pm
9
Full Name: Vincent Impens
Company Details: RealVisuals
Company Position Title: Managing Partner
Country: Belgium
Linkedin Profile: Yes

Outdoor scan for ortho pictures

Post by myvisuals »

Hello,

I have to make a quote for an outdoor scan of this building. Friday i will visit the building.
Very nice with a lot of detail... But i was wondering about te setting to use.
I was thinking of setting up te scanner on 180° and using 1/4 with 3X quality.
Normally i scan always targetless, works very wel.
But with this setting i think the ortho pictures will not have a good quality...

What do you guys think?
Google street location
https://www.google.be/maps/place/China+ ... 57!6m1!1e1
You do not have the required permissions to view the files attached to this post.
User avatar
David Zip
Forum Supporter
Forum Supporter
Posts: 117
Joined: Sat Jul 09, 2011 6:51 am
12
Full Name: David Zip
Company Details: 3D MobileScan
Company Position Title: Owner
Country: Canada
Linkedin Profile: Yes
Location: Edmonton, Alberta Canada
Contact:

Re: Outdoor scan for ortho pictures

Post by David Zip »

That setting has worked well for me. To deal with the protruding features in the building, extra scan positions (such as elevated positions) will be needed to avoid noticeable occlusions - this will give more point density as well.
The ortho feature will utilize Scene's Adaptive Point size and Gap filling to help with the densities.
User avatar
jcoco3
Global Moderator
Global Moderator
Posts: 1724
Joined: Sun Mar 04, 2012 5:43 pm
12
Full Name: Jonathan Coco
Company Details: Consultant
Company Position Title: Owner
Country: USA
Linkedin Profile: No
Has thanked: 70 times
Been thanked: 157 times

Re: Outdoor scan for ortho pictures

Post by jcoco3 »

Hi Vincent,

That really is a nice looking building.
I was thinking of setting up te scanner on 180° and using 1/4 with 3X quality.
Please don't scan just 180°, I seriously recommend performing a full 360° scan as the difference in time is negligible and you may need the rest of the data later for registration or modeling purposes.

As far as the resolution goes... After looking at the building from street view I can see that there are some fairly small details very high up, so I am leaning more towards many 1/2 resolution scans or maybe even full 1:1. Really it depends on what final resolution your ortho photos will need to be for the smallest detail you are trying to portray, but keep in mind that some of the smaller details may not be perceivable even in a high resolution ortho. When attempting to render such a high resolution ortho you can end up creating images that are unsaleable on most computers due to their extreme size. That may not be problem here, but I think it is worth mentioning.

You could scan with many more lower resolution scans, but since you cannot control exactly how the scans will overlap it is difficult to estimate the final density on distant surfaces. I prefer to scan at the resolution needed to obtain the appropriate density(for the smallest surface detail) over more than half the overlap distance, so the overlap will exceed the requirement. I hope that makes sense. I attached a crude density calculator based off the scanner's stated point spacing at 30 feet. It may help you estimate what the point spacing will be at a given distance, but it doesn't take into account surface geometry or angle of incidence, etc. You just input the distance then a maximum and minimum spacing and it highlights the correlating answers below. Use at your own risk :?

As David said the adaptive point size will help with the higher resolution orthos, because as you render higher resolution orthos the gaps between the points become more apparent giving the images a transparent look. You may even have to manually set the point size so high that it looks terribly coarse on your monitor, but it will look less transparent and sharper in the ortho.

Good luck :)
{attachment deleted}
myvisuals
V.I.P Member
V.I.P Member
Posts: 122
Joined: Tue Oct 07, 2014 12:05 pm
9
Full Name: Vincent Impens
Company Details: RealVisuals
Company Position Title: Managing Partner
Country: Belgium
Linkedin Profile: Yes

Re: Outdoor scan for ortho pictures

Post by myvisuals »

Thank you guys for the good advice... let's hope ;)
User avatar
jcoco3
Global Moderator
Global Moderator
Posts: 1724
Joined: Sun Mar 04, 2012 5:43 pm
12
Full Name: Jonathan Coco
Company Details: Consultant
Company Position Title: Owner
Country: USA
Linkedin Profile: No
Has thanked: 70 times
Been thanked: 157 times

Re: Outdoor scan for ortho pictures

Post by jcoco3 »

Gents, For all those that are using the Point Density Calculator I created, please download this revision:
{attachment deleted}
A recent post on another thread prompted me to create a metric version, but after it was complete I noticed that the results are quite different(quite massive at the lower end of the resolution scale!) To create these spread sheets I used the numbers reported by the Faro scanner verbatim, but I think there are some rounding issues, unless of course my spreadsheet math has a bust. Since the scanner's real units are in meters, my hunch is that the reported metric numbers are more accurate, but it will take empirical results to determine what is going on. The new spreadsheet has tabs for FEET, METERS, and the Delta showing the difference in the two tabs results. Maybe somebody with better math skills can make sense of all this and let me know what I did wrong? I am really hoping that somebody points out something stupid that I did in the spreadsheet otherwise I think the numbers on the scanner might be out of wack a little :?
brianpgreen
V.I.P Member
V.I.P Member
Posts: 324
Joined: Fri Oct 25, 2013 8:46 pm
10
Full Name: Brian Green
Company Details: Green Forensics-Azure Green Consultants
Company Position Title: Owner
Country: United States
Skype Name: brian.green.forensics
Linkedin Profile: Yes
Has thanked: 1 time
Been thanked: 1 time
Contact:

Re: Outdoor scan for ortho pictures

Post by brianpgreen »

Found it. You're dividing by 30 rather than 10 in column C of the METERS tab. Deltas are small after the change. I don't claim to have better math skills than you do nor would I claim you did something stupid. Just the hazards of CTRL C/CTRL V.

Thanks for the density calculator. I've been using it quite a bit.
User avatar
jcoco3
Global Moderator
Global Moderator
Posts: 1724
Joined: Sun Mar 04, 2012 5:43 pm
12
Full Name: Jonathan Coco
Company Details: Consultant
Company Position Title: Owner
Country: USA
Linkedin Profile: No
Has thanked: 70 times
Been thanked: 157 times

Re: Outdoor scan for ortho pictures

Post by jcoco3 »

Nice, thanks for picking up after me. I can fix it and repost in the morning.
So much for ever doubting anyone else mathematical skills again.
brianpgreen
V.I.P Member
V.I.P Member
Posts: 324
Joined: Fri Oct 25, 2013 8:46 pm
10
Full Name: Brian Green
Company Details: Green Forensics-Azure Green Consultants
Company Position Title: Owner
Country: United States
Skype Name: brian.green.forensics
Linkedin Profile: Yes
Has thanked: 1 time
Been thanked: 1 time
Contact:

Re: Outdoor scan for ortho pictures

Post by brianpgreen »

I'm finding that the point spacing vertically matches the published spec very closely but that horizontally it's a little different. Should this be the case?
User avatar
jcoco3
Global Moderator
Global Moderator
Posts: 1724
Joined: Sun Mar 04, 2012 5:43 pm
12
Full Name: Jonathan Coco
Company Details: Consultant
Company Position Title: Owner
Country: USA
Linkedin Profile: No
Has thanked: 70 times
Been thanked: 157 times

Re: Outdoor scan for ortho pictures

Post by jcoco3 »

I'm finding that the point spacing vertically matches the published spec very closely but that horizontally it's a little different.
Interesting...I have no idea :shock: Maybe there is a slight difference in the vertical step versus the horizontal step, but the specs say they are supposed to be the same size:
SpecsheetDeflectionUnit.PNG
I fixed that error you found in my spreadsheet, and I found another typo error that I corrected. Now that all my sloppy errors have been corrected the deltas are very small as you indicated.

Here is the revision:
Faro_PointDensityCalculatorR2.xlsx
You do not have the required permissions to view the files attached to this post.
Post Reply

Return to “FARO”