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

User avatar
M.W.Taylor
Site Admin
Posts: 530
Joined: Sun May 29, 2011 11:45 am

Re: 3D surface modeling of a giant redwood trunk

Post by M.W.Taylor » Fri Jan 13, 2012 1:30 am

edfrank wrote:Mike,

I have downloaded and installed MeshLab, so at least one person as done it so far.

Ed
Excellent !

I am attaching a few of the first trees I have scanned. If the tree forms lack texture or color try

Tools-Options- Increase Pixel Size from 1 to 2 or even 3

Also, if you want to change the Axis of rotation...goto Filters-Normals,Curvatures,Orientation then Transforms:Move, Translate Center... Then point and click with mouse the part of the shape you want for the new axis of roation. This is a very useful command which allows views from any perspective desired. As you can see, these clouds are partial and need to be filled in some more. This is only the beginning Ed,

Michael
Last edited by M.W.Taylor on Fri Jan 13, 2012 1:42 am, edited 2 times in total.

User avatar
M.W.Taylor
Site Admin
Posts: 530
Joined: Sun May 29, 2011 11:45 am

Re: 3D surface modeling of a giant redwood trunk

Post by M.W.Taylor » Fri Jan 13, 2012 1:33 am

Ed,

I tried to attach a few of the tree scan graphics but the ENTS blog site does not like the PLY extension. It's an ASCII file with X,Y,Z postition + R,G,B color code, nothing more. If you can't turn the filter off, give me your email and I will send the PLY files to you that way. thanks !

Michael

User avatar
Rand
Posts: 1217
Joined: Tue Mar 16, 2010 9:25 pm

Re: 3D surface modeling of a giant redwood trunk

Post by Rand » Fri Jan 13, 2012 5:55 pm

numero dos.........................

User avatar
M.W.Taylor
Site Admin
Posts: 530
Joined: Sun May 29, 2011 11:45 am

Re: 3D surface modeling of a giant redwood trunk

Post by M.W.Taylor » Fri Jan 13, 2012 10:55 pm

Ok,

Here we go. These were the 1st trunks I have scanned so there will be room for improvement. Don't forget to re-center axis to get the best views...Normals,Curvatures,Orientations Then Transform: Move, Translate, Center. Click mouse pointer on new axis to reset. Also Tools-Options pixel size from 1 to 2 sometimes looks better, depending on the background noise.

I will first attach the raw file without noise filtering.then filtered for noise to show the processing needed. I still need to clean the filtered version up it up a little. You can run these cloud sets through AutoCad and have only discrete structures (i.e. tree trunks and forms with a surface) filtered automatically out of the noise. The 1st. cloud map is the lower trunk of Terex Titan. I will get a MUCH better scan of this tree soon. This 1st. scan was more of a calibration run.

If you have any questions or problems viewing the graphics, let me know,

I may be away for a few days as I attempt to get better quality tree cloud maps.

Michael Taylor
WNTS VP
http://www.landmarktrees.net
Last edited by M.W.Taylor on Sat Jan 28, 2012 2:52 pm, edited 1 time in total.

User avatar
M.W.Taylor
Site Admin
Posts: 530
Joined: Sun May 29, 2011 11:45 am

Re: 3D surface modeling of a giant redwood trunk

Post by M.W.Taylor » Fri Jan 13, 2012 11:49 pm

The next attachments are from a large redwood trunk with my friend Mike standing underneath. I attached the raw version + filtered version. For filtered version, increase pixel size from 1 to 2 for best color. As I stated previously, you can see Mike's head, eyes, nose, mouth, chin, maroon sweatshirt, silver camera and blue jean. This was also a calibration run. I should be able to improve cloud density and color with practice.

Michael
Last edited by M.W.Taylor on Sat Jan 28, 2012 2:55 pm, edited 1 time in total.

User avatar
Rand
Posts: 1217
Joined: Tue Mar 16, 2010 9:25 pm

Re: 3D surface modeling of a giant redwood trunk

Post by Rand » Fri Jan 13, 2012 11:50 pm

Uh-yuppa..seems to be working:
Screen shot 2012-01-13 at 11.45.20 PM.jpg

User avatar
M.W.Taylor
Site Admin
Posts: 530
Joined: Sun May 29, 2011 11:45 am

Re: 3D surface modeling of a giant redwood trunk

Post by M.W.Taylor » Sat Jan 14, 2012 12:00 am

One more for tonight. This is the familiar oak tree with the background noise mostly filtered out and axis centered to tree form.
Last edited by M.W.Taylor on Sat Jan 28, 2012 2:56 pm, edited 1 time in total.

User avatar
M.W.Taylor
Site Admin
Posts: 530
Joined: Sun May 29, 2011 11:45 am

Re: 3D surface modeling of a giant redwood trunk

Post by M.W.Taylor » Sat Jan 14, 2012 12:04 am

Rand wrote:Uh-yuppa..seems to be working:
Screen shot 2012-01-13 at 11.45.20 PM.jpg

Rand, did you add a mesh layer ? Where you able to get the colorized pixels to display ?

Michael T.

User avatar
M.W.Taylor
Site Admin
Posts: 530
Joined: Sun May 29, 2011 11:45 am

Re: 3D surface modeling of a giant redwood trunk

Post by M.W.Taylor » Fri Jan 20, 2012 5:50 pm

The attached are some recent trunk scans. The cloud set are 100-250k in size so you will need MeshLab to view. I am attaching some jpg images first of Forest Glen Giant, Redwood Creek Giant, Melkor and Stratosphere Giant.

For all these new cloud maps, the entire lower trunk is modeled in 3-D. These cloud sets are the original scan without the noise filtered. You can see many of the background trees in these scans. The best details came out in Forest Glen Giant, an 8' dbh ponderosa that died recently. You can see where the bark is peeling off high up the trunk. The plates and color detail came out nicely. For Melkor you can see the lower burl and fire scarring quite well. With Redwood Creek Giant I captured about 250k cloud points in the first 15 feet of trunk. You can see the moss backed North side and ground detalis near the root flares. This cloud is plenty dense enough for an accurate volume calculation.

I hope to soon have an example of the Visual Basic code that calculates the irregular volumes under these scanned surfaces. I'll post an open source code version when finished.

Michael Taylor
WNTS VP
Attachments
Stratosphere Giant - 17'+ dbh
Stratosphere Giant - 17'+ dbh
Stratosphere Giant -372' tall
Stratosphere Giant -372' tall
Redwood Creek Giant - 17' dbh
Redwood Creek Giant - 17' dbh
Redwood Creek Giant - 362 feet tall
Redwood Creek Giant - 362 feet tall
Melkor is definitely not a double
Melkor is definitely not a double
Melkor 2nd largest coast redwood
Melkor 2nd largest coast redwood
forest glen giant now dead
forest glen giant now dead
perfect specimen pinus ponderosa
perfect specimen pinus ponderosa
8' dbh Forest Glen Giant
8' dbh Forest Glen Giant
Last edited by M.W.Taylor on Sat Jan 21, 2012 4:23 am, edited 1 time in total.

User avatar
Rand
Posts: 1217
Joined: Tue Mar 16, 2010 9:25 pm

Re: 3D surface modeling of a giant redwood trunk

Post by Rand » Fri Jan 20, 2012 7:00 pm

Michael,

I went to the menu bar and picked 'Filters'>'Remeshing, Simplification, and Reconstruction'>'Alpha Complex/Shape'. Then to actually see the surface you have to change the display mode to show the surfaces. Go to the toolbar and look for the icons that look like hexagonal extrusions, these control which geometry the program displays (Points, wireframe, flat line etc).

I haven't got into the different remeshing options, I just picked the first one, which seems to be a lucky guess that works okay. I image better results can be obtained by, as usual, reading the documentation.

Edit: On the other hand having the link to the online documentation working would be nice too. <Sigh> oh well, google turns up some youtube tutorial videos.

Post Reply

Return to “Measurement and Dendromorphometry”