3D surface modeling of a giant redwood trunk

General discussions of measurement techniques and the results of testing of techniques and equipment.

Moderators: edfrank, dbhguru

#81)  Re: 3D surface modeling of a giant redwood trunk

Postby fooman » Tue Feb 28, 2012 2:11 am

Hi Michael,

Sorry for not getting back to you earlier, but work has intervened.  I haven't been able to grab the software licence for Rapidform, as it is being used for commercial work right now.  Our training session is organised for next week, so I'll see what I can do then.

M.W.Taylor wrote:Matt,

I notice your Redwood Creek Giant mesh is a bit concave at the top ?

My volume solution for RCG is now about 1% from your RapidForm XOR now using a "best fit" strategy for the missing points (that always arrive in cloud maps).  I used a conservative linear interpolation between closest surface edge points. In my experience I have found that spline interpolation over-estimates tree surfaces so that was not used here.  



The concavity may be a function of the meshing algorithims and/or floating point accuracy on the model.  A linear interpolation betwen points is pretty much what a tet-mesh is.  More advanced analytical methods for stress analysis (e.g. finite element methods) will use polynominal (2nd order) interpolation of tet and quad meshes to obtain a more realistic model stiffness.


Also, you are measuring the volume and area of the mesh from RapidFormXOR, correct ?  In that case, your surface area should be a little greater because your mesh has 1/2 the points of the cloud set I am processing which has triangular facets too but more of them. I'll run the RapidForm mesh cloud through my custom code and see what happens.



Yes, all measurements are the inbuilt querying tools in the software.  The documentation does not discuss the methods used.


How much does RapidForm XOR cost ?  I get the feeling it's expensive.



At least in the tens of thousands.  A similar product we looked at as well was Geomagic.  The Rhino3d modeling software I mentioned up thread retails for ~1k, but I am not sure of the measuring capabilities of that software.  EDIT: a quick check shows it does measure volumes and surfaces.  And they do a pretty good evaluation offer: http://download.rhino3d.com/rhino/4.0/e ... /download/

Have you had a look at Matlab (or it's open source equivalent, GNU Octave?)  That sort of programming environment is made for hard core manipulation of large data sets.  There is a fairly well known Matlab method for calculating the volume of a point cloud, but I have yet to track down details, other than it is the summation of the volumes of the set of tetrahedrons created by the tet faces and a central point.  Actually pretty similar to what you are doing, but a 3d calculation, rather than a stack of 2d slices made from the area calculation of the triangles making up the slice.

Have a look at http://iso2mesh.sourceforge.net/cgi-bin/index.cgi?Home

Cheers,
Matt
User avatar
fooman
 
Posts: 53
Joined: Sun Dec 25, 2011 3:18 pm
Location: Wellington
Has Liked: 3 times
Has Been Liked: 29 times
Print view this post

#82)  Re: 3D surface modeling of a giant redwood trunk

Postby M.W.Taylor » Fri Mar 02, 2012 6:32 pm

Matt,

The Tet-Face volume summation is what MapSmart uses also. I am using it to test my ForestForm1.8 program. The Tet-Face system uses less calculations than the "orthongonal-quad-face" pie-wedge that I am volumizing in ForestForm1.8.  The Tet-Face summation adds up the volume just as you describe. A 3D central ray at the central base of the form (or mass centroid) to each vertex point on the trianglar mesh creates a network of tetrahedrons that can quickly and easily be fitted to most polystructures and volumized. If the top or bottom is open like a partial tree trunk point cloud, then it must be closed prior to Tet-Face summation. Both the Ortho-Quad-Face and the the Tet-Face break down when you get trees with multiple iteration and protrutring structures. The Tet-Face summatin in MapSmart adds up the space between the surface boundaries as though it were a continuous volume and thus over-estimates the size a little. I am not sure how Rapidform or Rhino3D deal with these polystructures? I assume they either just over-estimate or the software account for multiple surface boundaries. For $10,000+ I would assume RapidForm has this capability.  

I am currently re-coding the ForestForm(Version 2.0) to handle multiple iteration trees and any assorted complex polyform. I am also going to have the cloud load automatically into sheet2 instead of the old "cut and paste" technique. After doing that 1,000 times with 300k+ size point clouds,  I need a better system.  Also changing the ForestForm1.8 code to graph the quad-mesh as it is being interpolated and volumized. The final mesh will be rotatable like the older versions I posed. The rotating Mesh looks better than the point cloud. The lines show the surface curvatuve better than the raw point cloud in a 2d scatter plot.

I am very interested in the Rhino3D. The price is right. If it does everything MeshLab does (but with good documentation) and everything my custom VBA spreadsheet does for volumizing and surface area ( but more user friendly) then I am interested in getting the Rhino3D software. RapidForm is too expensive. If I got a grant for it or money was no issue then perhaps I would get that instead. Otherwise I will get by with MeshLab and ForestForm. The biggest problem I see with MeshLab is the lack of documentation.

Matt, Thank you again for your suggestions and testing my code. You have been very helpful. I found a rather serious bug in the last Forest Form attachment so I am going to delete and reload with the more stable version. Also going to code the Tet-Mesh volumizer alternative. It is only a minor change to the VBA to add this so why not ?  It should be MUCH faster for volume solving and surface area calculation due to fewer processes.

Michael Taylor
Attachments
forestform1.8.xls
forestform1.8
(8.46 MiB) Downloaded 167 times
Last edited by M.W.Taylor on Sat Mar 03, 2012 12:42 am, edited 1 time in total.
User avatar
M.W.Taylor
 
Posts: 252
Joined: Sun May 29, 2011 12:45 pm
Location: Northern California
Has Liked: 12 times
Has Been Liked: 190 times
Print view this post

#83)  Re: 3D surface modeling of a giant redwood trunk

Postby edfrank » Fri Mar 02, 2012 9:40 pm

Michael,

Please delete the version with the bug, and post the new revised version as a separate post so as not to confuse people, or have them miss your revised effort.

Ed
"I love science and it pains me to think that so many are terrified of the subject or feel that choosing science means you cannot also choose compassion, or the arts, or be awe by nature. Science is not meant to cure us of mystery, but to reinvent and revigorate it." by Robert M. Sapolsky
User avatar
edfrank
Site Admin
 
Posts: 4141
{ IMAGES }: 0
Joined: Sun Mar 07, 2010 6:46 pm
Location: Reynoldsville, Pennsylvania, USA
Has Liked: 864 times
Has Been Liked: 635 times
Blog: View Blog (3)
Print view this post

#84)  Re: 3D surface modeling of a giant redwood trunk

Postby M.W.Taylor » Sat Mar 03, 2012 12:44 am

Ed,

I just uploaded the latest bug free version of the volume solver. ForestForm1.8. All previous versions should be considered obsolete.  

Michael
User avatar
M.W.Taylor
 
Posts: 252
Joined: Sun May 29, 2011 12:45 pm
Location: Northern California
Has Liked: 12 times
Has Been Liked: 190 times
Print view this post

#85)  Re: 3D surface modeling of a giant redwood trunk

Postby fooman » Sun Mar 04, 2012 7:12 pm

Hi Michael,

Have a look at http://www.gom.com/3d-software/gom-inspect.html

Freeware mesh manipulation tool (developed by industrial 3d measurement company).

Cheers,
matt
User avatar
fooman
 
Posts: 53
Joined: Sun Dec 25, 2011 3:18 pm
Location: Wellington
Has Liked: 3 times
Has Been Liked: 29 times
Print view this post

#86)  Re: 3D surface modeling of a giant redwood trunk

Postby M.W.Taylor » Mon Mar 05, 2012 4:14 am

Matt,

I am going to download that and give it a try. Is that an evaluation version that blows up afer 30 days ?  Or permanently free ?  Just curious. I don't mind buying software. But RapidForm is just out of my price range.  Thanks again for the link,

Michael
User avatar
M.W.Taylor
 
Posts: 252
Joined: Sun May 29, 2011 12:45 pm
Location: Northern California
Has Liked: 12 times
Has Been Liked: 190 times
Print view this post

#87)  Re: 3D surface modeling of a giant redwood trunk

Postby fooman » Mon Mar 05, 2012 5:31 pm

Hi Michael,

Free forever, apparently. The tutor on the course recmmended it. I won't be able to do much more on that course - i'm about to get my appendix removed.

Cheers,
Matt
User avatar
fooman
 
Posts: 53
Joined: Sun Dec 25, 2011 3:18 pm
Location: Wellington
Has Liked: 3 times
Has Been Liked: 29 times
Print view this post

Previous

Return to Measurement and Dendromorphometry

Who is online

Users browsing this forum: No registered users and 1 guest

cron