Exact center of pipe from scan
-
- I have made 40-50 posts
- Posts: 40
- Joined: Thu Aug 31, 2017 4:38 pm
- 6
- Full Name: none
- Company Details: none
- Company Position Title: none
- Country: usa
Exact center of pipe from scan
What would be a good workflow for finding the exact center of a scanned concrete pipe? The Pipe is supposed to be 42" and I would like to sample different points along the interior for dimensioning. I am expecting the pipe to be out of round but I need a way to locate the exact center. Thoughts?
- Mike Annear
- V.I.P Member
- Posts: 800
- Joined: Mon Jun 08, 2009 5:45 am
- 14
- Full Name: Mike Annear
- Company Details: The Yard Brisbane
- Company Position Title: Manager - 3D Design and Scanning
- Country: Australia
- Linkedin Profile: Yes
- Location: Hemmant, Queensland, Australia
- Has thanked: 122 times
- Been thanked: 196 times
- Contact:
Re: Exact center of pipe from scan
I had to model an entire roller coaster a few years ago, I fitted cylinders to the point cloud in Rhino 3D then used those to extract a centre line.
I then used the Pipe command in Rhino to create a curved pipe along the centerline.
I then used the Pipe command in Rhino to create a curved pipe along the centerline.
You do not have the required permissions to view the files attached to this post.
-
- I have made 100> posts
- Posts: 108
- Joined: Thu Sep 21, 2017 9:43 pm
- 6
- Full Name: Iain Took
- Company Details: WMP
- Company Position Title: Dogsbody
- Country: United Kingdom
- Linkedin Profile: No
- Has thanked: 3 times
- Been thanked: 6 times
Re: Exact center of pipe from scan
The arc (3 points on arc) tool will give you A centre.
Might be a bit tedious though.
create object - from pick points - arc... (To create arc)
And
Tools - info - object info to see the centre coordinate
All in cyclone. Mike's way looks much easier.
Might be a bit tedious though.
create object - from pick points - arc... (To create arc)
And
Tools - info - object info to see the centre coordinate
All in cyclone. Mike's way looks much easier.
- frassolutions
- I have made <0 posts
- Posts: 9
- Joined: Thu Feb 14, 2019 2:36 pm
- 5
- Full Name: Jean-Simon Fraser
- Company Details: FRAS Solutions
- Company Position Title: CEO
- Country: Canada
- Skype Name: jsfraser2
- Linkedin Profile: Yes
- Contact:
Re: Exact center of pipe from scan
I second mike's method of fitting many cylinders.
There are other ways too. Fitting cylinders is a simple and effective solution, but it's not the most accurate one.
In Polyworks, I first cut the pipe with a plane along the length to extract the path. The pipe can be a little bent, so while doing so I make sure that the plane is properly aligned to get the curvature, if any. The goal is to cut in the center as much as possible and create a cross-section. The cross-section can then be used to create a polyline.
The polyline is then used to create a serie of sections along it. While doing that, it is smoothed just a little, so the result is reliable even if the surface is noisy. Once the sections are done, you can then fit a circle or even an ellipse, which will be much more accurate than a circle if the pipe is out of roundness. All sections will be perpendicular to the pipe axis and the center will be more accurate in 2D because you are extracting it at a specific location, not making an average with a cylinder.
Of course, this is tedious work unless it is automated with a macro script.
There are other ways too. Fitting cylinders is a simple and effective solution, but it's not the most accurate one.
In Polyworks, I first cut the pipe with a plane along the length to extract the path. The pipe can be a little bent, so while doing so I make sure that the plane is properly aligned to get the curvature, if any. The goal is to cut in the center as much as possible and create a cross-section. The cross-section can then be used to create a polyline.
The polyline is then used to create a serie of sections along it. While doing that, it is smoothed just a little, so the result is reliable even if the surface is noisy. Once the sections are done, you can then fit a circle or even an ellipse, which will be much more accurate than a circle if the pipe is out of roundness. All sections will be perpendicular to the pipe axis and the center will be more accurate in 2D because you are extracting it at a specific location, not making an average with a cylinder.
Of course, this is tedious work unless it is automated with a macro script.
https://frassolutions.com
Automation of 3D measurements and portable 3D scanning services
Automation of 3D measurements and portable 3D scanning services
- gsisman
- V.I.P Member
- Posts: 898
- Joined: Fri Oct 07, 2016 1:51 pm
- 7
- Full Name: Steve Long
- Company Details: Montgomery County DOT _ MD
- Company Position Title: Land Survey Supervisor
- Country: United States
- Skype Name: gsisman1
- Linkedin Profile: Yes
- Has thanked: 767 times
- Been thanked: 148 times
Re: Exact center of pipe from scan
Wow! (Roller Coaster) This is interesting- and an open challenge I'm sure for the Scan software coders-programmers. Seems like working on a mesh of the structures (with a hunk of a machine)and using some algorythm to identify a similar more geometrically regular matching spline figure (Complex Geometric solid library) and extruding it along that calculated splined-fit path is a good routine to include. Mike has done it manually in his workflow, so it should be programmable.......frassolutions wrote: ↑Thu Apr 18, 2019 4:01 pm I second mike's method of fitting many cylinders.
There are other ways too. Fitting cylinders is a simple and effective solution, but it's not the most accurate one.
In Polyworks, I first cut the pipe with a plane along the length to extract the path. The pipe can be a little bent, so while doing so I make sure that the plane is properly aligned to get the curvature, if any. The goal is to cut in the center as much as possible and create a cross-section. The cross-section can then be used to create a polyline.
The polyline is then used to create a serie of sections along it. While doing that, it is smoothed just a little, so the result is reliable even if the surface is noisy. Once the sections are done, you can then fit a circle or even an ellipse, which will be much more accurate than a circle if the pipe is out of roundness. All sections will be perpendicular to the pipe axis and the center will be more accurate in 2D because you are extracting it at a specific location, not making an average with a cylinder.
Of course, this is tedious work unless it is automated with a macro script.
If you are going for TRUE asbuilt "Capture Reality" models and not estimations
- Mike Annear
- V.I.P Member
- Posts: 800
- Joined: Mon Jun 08, 2009 5:45 am
- 14
- Full Name: Mike Annear
- Company Details: The Yard Brisbane
- Company Position Title: Manager - 3D Design and Scanning
- Country: Australia
- Linkedin Profile: Yes
- Location: Hemmant, Queensland, Australia
- Has thanked: 122 times
- Been thanked: 196 times
- Contact:
Re: Exact center of pipe from scan
Hi, the centreline extraction was quite accurate, the 3D pipe model was checked against the pointcloud using Rhino's Point Set deviation tool.
I would just like to add that the 3D model was just a by product of the real purpose of the scan which was to analyze the spatial relationship of the 3 rails.
The analysis was done using "Grasshopper", the graphical algorithm editor that integrates with Rhino 3D (it is now built in to version 6.0)
Most of the work that I do is for Surveying companies, the whole project was based on strict control and the results thoroughly checked by traditional survey methods.
Cheers,
Mike.
I would just like to add that the 3D model was just a by product of the real purpose of the scan which was to analyze the spatial relationship of the 3 rails.
The analysis was done using "Grasshopper", the graphical algorithm editor that integrates with Rhino 3D (it is now built in to version 6.0)
Most of the work that I do is for Surveying companies, the whole project was based on strict control and the results thoroughly checked by traditional survey methods.
Cheers,
Mike.
You do not have the required permissions to view the files attached to this post.