Rolling object to scan

Please feel free to ask other members for help with certain projects
Post Reply
Jacob
I have made <0 posts
I have made <0 posts
Posts: 5
Joined: Mon Dec 09, 2019 10:36 am
Full Name: Jakub Bies
Company Details: Scanning
Company Position Title: Technican
Country: Poland
Has thanked: 2 times
Been thanked: 1 time

Rolling object to scan

Post by Jacob »

Hello everyone!

I would like to ask for an advice. Maybe someone has an experience with such projects and can help :)

I'm wondering about technique to use for scanning crane onboard the ship (crane height: around 70m). There is no access on the crane, so whole the job must be done from shore and ship deck. Ship will be alongside during job.

I would like to ask what kind of scanning technique do you recommend?

I have an idea to do it by laser scanning on the shore + photogrammetry using drone.

But there is one issue. Due to wind and waves ship rolling. And I’m wondering if the 3D data will be accurate? And how separated data from 3D scanner and photogrammetry will align each other? Of course rolling is not significant but still exist.

Do you have any advice how to compensate influence of ship rolling? Especially for 3D scanner because part of scans probably will be made from shore and part from ship deck.

Thank you in advance and wish all the best! :D

User avatar
Mike Annear
V.I.P Member
V.I.P Member
Posts: 541
Joined: Mon Jun 08, 2009 5:45 am
Full Name: Mike Annear
Company Details: Mike Annear
Company Position Title: 3D Modeler
Country: Australia
Linkedin Profile: Yes
Location: On an island off Brisbane ...
Has thanked: 7 times
Been thanked: 28 times
Contact:

Re: Rolling object to scan

Post by Mike Annear »

That is a tricky scanning situation !
I was involved with scanning Oil Rigs a few years ago, which sounds similar.
This is how I would do it ...

First of all, you need a decent scanner with long range and the ability to do a "Detail Window" scan.
A leica P-40 / P-50 would be perfect.

Do ALL of your scans from onboard the ship, with the compensator turned OFF. This means that if the ship rolls it wont affect your scans.
When you do a scan do it in Two passes without bumping or moving the tripod.
Do one overall 360 degree scan of the boat deck etc, then do a detail window scan of the crane at a higher resolution. The 360 degree scans will help with registration, the detail scans will pick up the top of the crane (hopefully)
Do lots of scans from the boat deck with plenty of overlap, the crane itself wont be much help as far as registration goes.
It is not ideal as far as registration goes, but it should work.

Important - Before you do your registration - Trim all the scans so only the ship and the crane are in the pointcloud, nothing from the shore.
If you are using Cyclone there is a function in the Scanworld for making the trimmed pointcloud the "Default Cloud". You need to do this or Cyclone will use the full untrimmed pointcloud.

Good Luck !
Regards,
Mike.
Mike Annear
www.mikeannear.com
QLD, Australia.

jaredmatthews
V.I.P Member
V.I.P Member
Posts: 150
Joined: Sun Sep 13, 2015 7:19 pm
Full Name: Jared
Company Details: Sova Surveys Limited
Company Position Title: Director
Country: UK
Skype Name: jaredmatthews
Linkedin Profile: Yes
Has thanked: 3 times
Been thanked: 1 time
Contact:

Re: Rolling object to scan

Post by jaredmatthews »

I would go with what Mike said above, and also add that I would suggest using targets (checkerboards or sphere, but preferably spheres) on the ship to help with registration.

Cloud to Cloud registration is generally good enough these days, but in a tricky scan like this, id have a failsafe backup with targets.

User avatar
Daniel Wujanz
I have made 80-90 posts
I have made 80-90 posts
Posts: 88
Joined: Fri Aug 24, 2018 11:26 am
Full Name: Daniel Wujanz
Company Details: technet GmbH
Company Position Title: 3D Laser Scanning Specialist
Country: Germany
Linkedin Profile: Yes
Location: Berlin
Has thanked: 32 times
Been thanked: 35 times
Contact:

Re: Rolling object to scan

Post by Daniel Wujanz »

Dear Jacob,

I have played around with this problem for about four years. In simple terms you're facing the inverse case of mobile mapping. A consequence of the ships' movement during scanning is, that the resulting point cloud will be geometrically distorted. Using cloud to cloud or another registration technique for static scans will NOT solve the problem. What you can do is the following:

- Perform a static scan from the shore / dock
- Measure the object's movement (so called kinematic parameters) during the scan, with e.g. three tracking total stations, a stereoscopic tracking system or a funky combination of other sensors. It is important that all sensors are in the same coordinate system AND are synchronised
- Rectify the point cloud (apply kinematic parameters) by correcting individual points

While I was at uni we went through three different stages in order to solve this issue. The outcome are two Master's thesis in English language that I can send you and / or the two following articles:
I've attached three images:

One that show a static scans of ships in motion. One with the latest evolution of our system and the corrected point cloud.

To the best of my knowledge there is no off-the-shelf solution to this problem. If you have further questions just send me a PM.

Cheers

Daniel Wujanz
You do not have the required permissions to view the files attached to this post.

jaredmatthews
V.I.P Member
V.I.P Member
Posts: 150
Joined: Sun Sep 13, 2015 7:19 pm
Full Name: Jared
Company Details: Sova Surveys Limited
Company Position Title: Director
Country: UK
Skype Name: jaredmatthews
Linkedin Profile: Yes
Has thanked: 3 times
Been thanked: 1 time
Contact:

Re: Rolling object to scan

Post by jaredmatthews »

Amazing work Daniel!

User avatar
Daniel Wujanz
I have made 80-90 posts
I have made 80-90 posts
Posts: 88
Joined: Fri Aug 24, 2018 11:26 am
Full Name: Daniel Wujanz
Company Details: technet GmbH
Company Position Title: 3D Laser Scanning Specialist
Country: Germany
Linkedin Profile: Yes
Location: Berlin
Has thanked: 32 times
Been thanked: 35 times
Contact:

Re: Rolling object to scan

Post by Daniel Wujanz »

Thanks Jared - the hole project was heaps of fun and very challenging at the very same time!

Unfortunately I left uni before we could go really big. We wanted to monitor container ships before and after loading. These beasts can ultimately fall apart if the weight distribution is wrong.

If you're interested, there are fascinating articles by insurance companies that summarise why ships sink:

https://www.agcs.allianz.com/content/da ... w-2018.pdf

Cheers

Daniel

P.S. here's my favourite quote from this report: "Friday is the most frequent day for shipping losses over the past decade (175) with Saturday (143) the safest day at sea. "

Jacob
I have made <0 posts
I have made <0 posts
Posts: 5
Joined: Mon Dec 09, 2019 10:36 am
Full Name: Jakub Bies
Company Details: Scanning
Company Position Title: Technican
Country: Poland
Has thanked: 2 times
Been thanked: 1 time

Re: Rolling object to scan

Post by Jacob »

Hello!

Thank you Daniel, Mike and Jared for answers! :D That's good knowledge to be analyzed.

Our scanning is delayed and for now we don't know when exactly is going to be done. But I'm grateful for your posts.

,,To the best of my knowledge there is no off-the-shelf solution to this problem. If you have further questions just send me a PM." - thank you Daniel! When I find time to analize lectures you have sent (during this time unfortunately I didn't have) and I have any questions I will feel free to write you a PM ;)

Post Reply

Return to “Request Help With Projects”