Matterport 3D Tour Pros  |  Google Street View Pros  |  Real Estate Agents  |  3rd Party Service Providers
Should I buy a Camera? Quiz How Much to Start 3D Tour Business? Calculate How Much Can I Make as a Pro? Calculate Win More Listings? Calculate
View Map Full Screen
Last 30 Days: 80,778 Page Views | 11,611 Visitors | 82 New Members
WGAN Knowledge Base | Total Posts: 49,613  |  Total Topics: 7,063
4,100 WGAN Members in 114 Countries
We Get Around Network Forum
WGAN Training Academy (Learn more)
WGAN-TV Channel
How to Make Money Shooting GSV with
Discover Africa Group Director Andre Van Kets
AdviceAirbusAirplaneScanningTips

Airbus 3307423

ThePhilOnMatt private msg quote post Address this user
Hello everybody,
I am planing to scan the inside of a commercial airplane in a few weeks and my client doesn't want to have odd objects on trays or having overhead bins opened to prevent misalignment issues. They want it as clean as possible. Here is a picture of what I will be scanning.




This shot was photoshopped a lot and there are EXIT signs and other type of markers but would that be sufficient ?

Should I start praying for the "Misalignment God" or build a Voodoo doll ?

Every row will have a scan point on each aisle(2).

Tips and tricks anyone ?

Philippe
Post 1 IP   flag post
justinv private msg quote post Address this user
You have to make sure every scan is placed in the correct spot as you scan. That many rows of seats, the software is going to get confused and have incorrect placement. If you build off the scan in wrong location, it is going to be hard to find the one misplaced scan.
Post 2 IP   flag post
Standard
Member
Home3D private msg quote post Address this user
The biggest issue you face is that there are rows and rows of identical seats and windows and bins. This is murder for alignment. The other issue is that, yes, you have to do each and every row, because if you don't, there will be holes in the dollhouse floor wherever the camera has not been able to "see". So be sure you select your tripod positions for the maximum visibility of the floor between the rows of seats.

I modeled the Directors Guild of America theaters in Los Angeles. DGAevent.com Theater One has about 600 seats. At first I thought I could do every other row, but there were holes in the "between" floors and also alignment problems due to the highly repetitive space. In my case the walls were also fairly far away which gave the IR less to register to. You don't have that issue here.

Anyway, sounds like a cool project. I look forward to seeing the end result if you can share it.
Kevin Dole
Post 3 IP   flag post
justinv private msg quote post Address this user
What expectation did you set for the client?

I can see parallax being a problem like it is with stair rails! It will be too close to be perfect that many times.
Post 4 IP   flag post
ThePhilOnMatt private msg quote post Address this user
Quote:
Originally Posted by justinv
What expectation did you set for the client?

I can see parallax being a problem like it is with stair rails! It will be too close to be perfect that many times.


Yes they are aware of technical limitation due to being to close like the inside of a car. I just hope everything will go as smooth as possible. They will have different scenarios with different props but that would be like the plan B. Thanks for your feedback.
Post 5 IP   flag post
WGAN Forum Founder DanSmigrod private msg quote post Address this user
@ThePhilOnMatt

I have not scanned a project so challenging for all the reasons given by @justinv @Home3D above.

That said, if I was doing this project, aside from managing the expectations of the Client, I would:

1. I would scan the entire aircraft TWICE in the SAME model
2. First, I would scan with the Matterport Camera set at the lowest tripod level. When the camera is closest to the ground it seems to have more success in knowing where it is (in a symmetrical environment.) Since you are going to disable all these scans in Workshop, the intent here is: 1) to make it easier for the eye-level scans to know where to attach themselves; and to eliminate the back holes under the seats as described above. If you don't have scan data, it means the ability to "fly through" the black hole. Yikes!
3. Set the Camera in the aisle. Turn the camera so that it is at a 45 degree angle to the row. The purpose is to avoid stitching errors on the chair because the stitching will occur within the empty row instead of seat backs.
4. After each scan, confirm that the scan is in the right location. If not, delete the scan and try again
5. Now, set the camera about a foot above the seats and do the second level of scans
6. In Workshop, disable all the lowest level scans (where the tripod was set to the lowest level)

Even a DSLR Pro using a fish-eye lens would be challenged with this project because of closeness of the camera to the subject.

Please do share the results as well as what technique that you tried to accomplish this challenging project.

Best,

Dan

P.S. I could imagine (potentially) succeeding with a Ricoh Theta V + CUPIX. The Ricoh Theta V has the advantage of just two lenses. So, you could have the sides of the camera face the front and back of the aircraft. That said, CUPIX might be as confused as Matterport in figuring out which 360º photo sphere gets attached where. If the final use was Google Street View, you could use Ricoh Theta V + Panoskin Pro to moderate. Plus, if you are shooting 360º photos, these can be edited in post to "clean up" any bad stitching.
Post 6 IP   flag post
ron0987 private msg quote post Address this user
What you might consider is every third scan offset from the isle to a seat area and when you offset on both sides kind of similar to example, this way your breaking up some of the consistence and then you can shut off unneeded scans.
Just a though to give the software something different, you also might consider using the target Matterport has to give it a reference point.
Post 7 IP   flag post
WGAN Forum Founder DanSmigrod private msg quote post Address this user
Quote:
Originally Posted by @ron0987
Just a though to give the software something different, you also might consider using the target Matterport has to give it a reference point.


Ah! Matterport AprilTags in the lower level of scans that you are going to hide anyway!

Dan
Post 8 IP   flag post
Standard
Member
3dVuz private msg quote post Address this user
Hoping for the best for you....

BUT

I have $100 says you’re gonna be crying by row 10
Post 9 IP   flag post
ThePhilOnMatt private msg quote post Address this user
Thank you guys for all your comments and ideas. It's really appreciated. Will post when the finished project if sucessfull. If not I will wire transfer 100$ to @3dVuz.
Post 10 IP   flag post
Gladsmuir private msg quote post Address this user
Yes, good luck with the job which is certainly a challenge.
Camera height for the final scans is a real issue.
My experience with this was a Boeing 757 Business Class cabin for Aer Lingus shot on a DSLR for publication to Google Street View.
As you will see https://goo.gl/WZoK2G we went for a lens height just above seat level.
A challenge was the seat back TV screens which were scrolling through five different images. As we were shooting in segments we had to keep waiting for the correct image.
So a pre-shoot request to the client to coordinate any monitors might save additional issues on the day.
It will be an interesting model!
Post 11 IP   flag post
PedroAvilez private msg quote post Address this user
Hi Phil
I have scan a A330 last year.
Pls find enclosed some pics.
We have faced several dificulties (alignment problems, distortions due to heat, etc).
But at the end the result was really nice.
Any doubts pls contact me by PM
Best regards
Pedro


Post 12 IP   flag post
50190 12 12
Log in or sign up to compose a reply.
destitute