Jump to content
J. Wallace

Does anyone make use of mesh from drone data?

Recommended Posts

I'm starting a new project and have received some drone data which I've used for creating a site model as well as high definition images. 

My question for people is whether (and how) one would make use of some mesh data. I've attached a screen shot and some information, the file is huge and an interesting reference but I'm not sure if it has any value besides that. You can see that once imported, the saved file size leaps up to 994 mb in size (this is the mesh only, no other data in the file, site model in a separate file).  Love to hear peoples comments. Thanks

mesh1.png

Intial image1.jpg

mesh.jpg

Edited by J. Wallace
  • Like 2

Share this post


Link to post

Hello J.:

 

You might want to try using the Modify->Mesh Simplification menu item, which will let you reduce the heavy load from the mesh without hurting quality too much (it is a smart function).

 

  • Like 3

Share this post


Link to post

Thanks for showing this.  I have no info/experience, but rather another question:

 

Your middle image shows overlay of contours.  Are those somehow generated from the mesh?  Or applied from some other source? Are they extractable and therefore useful data for DTM?

 

Looking forward to where this thread might go.

 

-B

Share this post


Link to post
1 hour ago, Benson Shaw said:

Your middle image shows overlay of contours.  Are those somehow generated from the mesh?  Or applied from some other source? Are they extractable and therefore useful data for DTM?

 

 

 

 

1

Hi @Benson Shaw. The contours are from a site model which we created with shapefile contour data (from the drone, the shape file comes with other support files). Super easy to create, I think it took but a few moments. The high def image is from the drone, other than for observation purposes I haven't figured out how to use the mesh.  It's a monster file that is best not mixed with the project data.  I'm curious to hear how other folks use this type of information. 

Share this post


Link to post

Hi @J. Wallace I use this kind of data all the time.

I rarely pull in the mesh though, as I don't find it all that useful. The mesh will have been generated from a point cloud in the first place, and that's what you want to use.

I make a new file and then import the point cloud in to 3 layers. One layer with 100% points imported, one with 50% and one with 20%. Called Hi, Med and Low Detail.

Then I use a DLVP to import it into my working file, so I can switch between levels of detail as required.

 

Point clouds are much more useful than the mesh, as you can snap stake objects on to the points as need be to create a native VW site model with as much resolution that you need. Too many stakes makes messy contours.

 

As for the mesh, it'll either have a huge texture mapped to it (with the weirdest mapping and UV's you can imagine) or it'll be coloured with vertex colours.

You should be able to get a point cloud with point colours anyway.

 

But which ever way you choose to go, I would definately recommend having the drone data in a seperate VWX and viewporting it in. Let me know how you get along.

 

While not all of this video is entirely relevant, some of this video may help (it was prepared for another client a while back)

 

 

  • Like 3

Share this post


Link to post

Here is another interesting blogpost about similar topic 

 

 

Its relatively easy to make a detailed 3d building in VW but harder to do the surroundings. I have played with Twinmotion and its quite easy to make a realistic natural landscape but it would be fabulous to get the surroundings imported from drone photo as a textured 3d mesh. This video shows a workflow for lumion :

 

https://lumion.com/blog.html?post=175009050560

  • Like 2

Share this post


Link to post
4 hours ago, RussU said:

Hi @J. Wallace I use this kind of data all the time.

I rarely pull in the mesh though, as I don't find it all that useful. The mesh will have been generated from a point cloud in the first place, and that's what you want to use.

I make a new file and then import the point cloud in to 3 layers. One layer with 100% points imported, one with 50% and one with 20%. Called Hi, Med and Low Detail.

Then I use a DLVP to import it into my working file, so I can switch between levels of detail as required.

 

Point clouds are much more useful than the mesh, as you can snap stake objects on to the points as need be to create a native VW site model with as much resolution that you need. Too many stakes makes messy contours.

 

As for the mesh, it'll either have a huge texture mapped to it (with the weirdest mapping and UV's you can imagine) or it'll be coloured with vertex colours.

You should be able to get a point cloud with point colours anyway.

 

But which ever way you choose to go, I would definately recommend having the drone data in a seperate VWX and viewporting it in. Let me know how you get along.

 

While not all of this video is entirely relevant, some of this video may help (it was prepared for another client a while back)

 

 

7

Thanks very much @RussU. That is super helpful information.

What your describing here looks like your taking cloud point data to make a site model, by snapping stake objects to portions of this.

In my case I get a set of shapefiles that contain 1' contours. I really like your approach and if I had a drone along with additional software, it's something that I could do myself.

Love the video you put together and thank you again for sharing this.

  • Like 1

Share this post


Link to post

Dear Jamie,

You might also consider exploring Sarah Barrett's Marionette network - presumably after you've got rid of the bits of the mesh you don't want included which is a hard thing to achieve I'd imagine. I haven't tried it but it should work like the Model>Point Clouds>Extract 3D loci from Point Clouds which is great.

 

 

 

Cheers,

Peter

 

 

  • Like 2

Share this post


Link to post
22 hours ago, Peter Neufeld said:

Dear Jamie,

You might also consider exploring Sarah Barrett's Marionette network - presumably after you've got rid of the bits of the mesh you don't want included which is a hard thing to achieve I'd imagine. I haven't tried it but it should work like the Model>Point Clouds>Extract 3D loci from Point Clouds which is great.

 

 

 

Cheers,

Peter

 

 

Thanks very much @Peter Neufeld I'll check that out

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

 

7150 Riverwood Drive, Columbia, Maryland 21046, USA   |   Contact Us:   410-290-5114

 

© 2018 Vectorworks, Inc. All Rights Reserved. Vectorworks, Inc. is part of the Nemetschek Group.

×