Registration issue with indoor scans

Discuss all Faro related software issues here.
Scott
V.I.P Member
V.I.P Member
Posts: 498
Joined: Tue Mar 29, 2011 7:39 pm
Full Name: Scott Page
Company Details: Scott Page Design- Architectural service
Company Position Title: Freelance
Country: USA
Linkedin Profile: No

Re: Registration issue with indoor scans

Post by Scott » Mon Aug 28, 2017 9:13 pm

sshemro wrote:
Mon Aug 28, 2017 7:38 pm
3DForensics wrote:
Mon Aug 28, 2017 7:13 pm
Manually placing scans close together gives the C2C algorithm a good starting point and best chance of success. Top Down View is a means of helping with initial scan positions and is what I would call a "lead in" for C2C.
Eugene
Exactly. My point in this instance was that if he manually aligns the improperly placed scans, and then runs top down registration, he'll lose the positioning of all his scans. Better to not do top down at all I would think, and just use C2C to not risk throwing scans all over the place.
Time = money as the cliché goes, so if you get a tight point cloud quickly from Top-down > C2C registration, then you've made it profitable. I seldom use spheres anymore -except when survey control is required. C2C works very well, once you understand it. Correspondence view adjustments are much easier to see from a large high-res monitor, than from a laptop. Good luck!

User avatar
blazaj
V.I.P Member
V.I.P Member
Posts: 415
Joined: Tue Jan 26, 2010 1:07 pm
Full Name: Jiri Blazek
Company Details: Blazek Win Precision Systems
Company Position Title: 3D documentation
Country: Czech
Skype Name: blazaj
Linkedin Profile: No
Location: 50.2379311N, 12.8408917E

Re: Registration issue with indoor scans

Post by blazaj » Tue Aug 29, 2017 7:34 am

IMO TV+C2C is perfect for smaller blocks. Individual floors of a building for example. To register blocks together I prefer targets. It doesn't have to be spheres, carefully placed checkerboards and planar surfaces will do the job as well. It dramatically reduces computing time, allows for adjustments, and mis-alignments are easier to spot. It adds some extra time to place targets on site, but much less than using targets all the way or registering the whole thing without any - relying on C2C only.

User avatar
sshemro
I have made 30-40 posts
I have made 30-40 posts
Posts: 33
Joined: Fri Mar 24, 2017 1:24 pm
Full Name: Skyler Shemro
Company Details: Rand Construction
Company Position Title: BIM Specialist
Country: United States
Linkedin Profile: Yes
Location: Alexandria, VA

Re: Registration issue with indoor scans

Post by sshemro » Tue Aug 29, 2017 1:41 pm

Scott wrote:
sshemro wrote:
Mon Aug 28, 2017 7:38 pm
3DForensics wrote:
Mon Aug 28, 2017 7:13 pm
Manually placing scans close together gives the C2C algorithm a good starting point and best chance of success. Top Down View is a means of helping with initial scan positions and is what I would call a "lead in" for C2C.
Eugene
Exactly. My point in this instance was that if he manually aligns the improperly placed scans, and then runs top down registration, he'll lose the positioning of all his scans. Better to not do top down at all I would think, and just use C2C to not risk throwing scans all over the place.
Time = money as the cliché goes, so if you get a tight point cloud quickly from Top-down > C2C registration, then you've made it profitable. I seldom use spheres anymore -except when survey control is required. C2C works very well, once you understand it. Correspondence view adjustments are much easier to see from a large high-res monitor, than from a laptop. Good luck!
I agree doing interior scans TD and C2C is the more efficient option. But he already used targets so I don't understand the point of negating the targets by using top down at this point

Sent from my SM-G930T using Tapatalk


donCPS
I have made <0 posts
I have made <0 posts
Posts: 8
Joined: Wed Mar 01, 2017 9:39 pm
Full Name: Don
Company Details: ACME
Company Position Title: Designer
Country: USA

Re: Registration issue with indoor scans

Post by donCPS » Tue Aug 29, 2017 2:07 pm

Reporting back to say the issue has been resolved.

The problem was 2 scans at the very beginning of the "leg" were corrupted AFTER assigning targets. It must have happened halfway through registration, as it was after I registered these scans, but before I viewed the project point cloud.

I deleted the corrupt scans, imported the raw scans of the back in, processed and placed them, and everything lined back up the way it was supposed to.

Thanks for your help!

Return to “FARO SCENE”