Hello,
I’m trying to remove the leaves from this tree point cloud but I’m out of ideas and I was trying maybe to do a bounding box and remove the points nearest to the bounding box but I’m not able or by density? I thought by colour but it doesn’t seem the case. Any ideas or some help of how to do so please?
thank you in advance
This should probably be done through point color intensity, but I only know how to do that through python. If you could share the file I am sure someone (or me in a few days) can help you out.
Please show the gh.
Hi!!! Thank you for your response I tried a little bit by myself this is the file with the point cloud file internalised.
albero GHprova 1.gh (14.1 KB)
I have a feeling that you are heading North instead of South with this puzzle of yours. I mean: IF you want to create “trees” (and a zillion other things) the only proper way is via code (and Recursion). Like:
So if you have plans to learn how to code … notify for an entry level example on that matter.
Hello and thank you for your responce. The fact is that I don’t want to create trees. I need to take the point cloud of a survey use the point cloud and clean it from the leaves and get only the branches. Then from that I will have to create the mesh.
Alexia,
The file you shared references your pointcloud directory but no actual pointcloud data is in the file.
Internalize the pointcloud or upload the .e57 separately - especially when you’re using various plugins some of us don’t have installed.
Note:
Rhino 8 grasshopper has pointcloud options now that make it easier to skip volvox and grab the color information you need for this.
If you’re staying with Rhino 7, then take a look at this as well:
Thank you for letting me know that it was not internalized. I did try now to upoload but the point cloud is too big to upload there is a limit of 20mb of files. how can I do it so I can get some help?
You could reduce and upload part of your pointcloud to make it work with the 20 MB limit. Alternatively, you could just temporarily share via a drive link or WeTransfer.
In the meantime, maybe look at this:
cloud.gh (10.8 KB)
In summary:
I used an example pointcloud I had:
similarity component (with a tolerance) so I get rid of points with a certain value similar to the specified color:I suppose you can replace the information with your cloud and with the equivalent volvox components.
Note:
I don’t have volvox installed so I grabbed the python component from one example in the link I had shared. Though the ‘leaves’ are getting removed, this is simply treating the pointcloud as points. I believe Volvox has an option to cull the actual pointcloud, which sounds to me like what you actually want (remove points without losing the original colors).
Lastly: Do you have to use grasshopper for this? Open source CloudCompare, which volvox communicates with anyway, can help you split colors much faster and export an .e57, before bringing into Rhino - just wondering.
If you manage to reduce/share the original cloud and it’s not ridiculously big, then I might install volvox to look at it more specifically (to cull clouds in volvox I think you need a more specific approach using ‘keys’ defined from the cloud itself). Alternatively, if you try it with Rhino 8, then the process can be more straightforward and you won’t need the python script nor a plugin.
Good luck!
Thank you so so much for your reply, I’m trying my best and yes it has to be done in grasshopper. The only think for volvox the name of the file has to have no empty spaces if not it doesn’t work at least for me. I’m trying to do also the cull cloud but I’m not able. If you can help me that would be great!
I’m trying to find the density of the point cloud. Please help me!
But today I will have also a look at your python script.
-maybe I can get an idea or anything… I can’t use Rhino 8… I have rhino 7
alberoe57.zip
I see you are familiar with python me not much I found a python python script online from a phd research, is there any possibility to turn this into a Grasshopper script?
So sorry if I ask too much
source-rc1.zip (66.9 KB)
@Alexia_Charalambous your file has restrictions so I requested permission to download - you should have gotten an email notification.
So sorry it’s the stupid stuff of protection of the university, of course you can download.
I did try also this approach to create boxes from the cloud with octree with a minimum of 100 or 10 then tried to eliminate the points less than 500 but it doesn’t seem to work either
I cannot:
Maybe use a WeTransfer link instead 
Now it should work.
Thanks!
Not sure what to tell you for ‘finding density’ other than use 3D point proximity?
I’d have to navigate Volvox a little more but I can’t at the moment, apologies.
The only automatic ‘key’ it finds is ‘intensity’ so I look at it similar to CloudCompare to play with values:
Likewise, not enough - I did check that phd stuff, though, (this?) pretty interesting.
Again, no time to integrate it properly into GH, maybe somebody more knowledgeable will 
@Holo
@Quan_Li
@MateuszZwierzycki
But I asked AI in case you wish to delve into it yourself (check your inbox).
Meanwhile, check these playful ways (using volvox components to retain pointcloud + color information):
cloud2.gh (21.3 KB)
'Hope it helps!
Obviously, you’ll have to specify your file directory again.
Best
You should check this video! This guy explain pretty well Volvox, you can jump to min 25, i hope this can help you!
I did see this video and it was very helpful thank you!
