scrawk / hull-delaunay-voronoi Goto Github PK

Hello,

I'd like to ask if it is possible to create irregular / non-convex meshes via Hull-Delaunay-Voronoi? By using Delaunay3D I can already create rectangular meshes. I would also like to create non-convex meshes such as U shape. I tried the ExampleVoronoi3D code with the vertex3 array that I defined. But it did not work or I did something wrong. I would be glad if you can help me with that.

Cheers,
Beril

Greetings,

How do I go about adding points to the delaunay triangulation incrementally with your implementation?

Thanks,
Georgii

Greetings,

Can I ask why did you extend the 3rd dimension in the generate() method under DelaunayTriangulation3?

What will happen if just generate the convex hull using random points in the 3rd dimension?

Thanks,
Georgii

I'm trying to port this project to java and I cannot find any implementation details on the recalculateBounds method. Could you explain, what does it do?

Thanks for the excellent library! I was wondering if it was currently possible to tag or ID vertices fed to a VoronoiMesh2 instance to enable edges and regions to determine which raw vertex they relate to/were generated from.

I noticed the Vertex2 class has an Id property but this doesn't seem to propagate.

I need to create a convex hull of a high poly 3D object in my scene. I can access its vertices, but its type is UnityEngine.Vector3[]. ConvexHull3.Generate() requires an input of type Vertex3[] in my case. I can convert each Vector3 in my vertex array to Vertex3, but that would take too much time to achieve in real time. Considering both Vertex3 and Vector3 are essentially the same thing, supporting Vector3 in ConvexHull3 would be ideal.

Since Delaunay is calculated and is the dual graph of the voronoi graph, is there a way to use it to get all adjacent cells to a given cell?

The lost cells issue is still present. See my comment at #6 (comment) for more information. I'd vote to reopen that issue and continue the conversation there since it has the all the context about it. Sorry to be the bearer of bad news, because otherwise this library seems really advanced!

Awesome library! This is really cool and I haven't found another like it.

Is there any way to generate a 3DVoronoi using voronoi sites using this library? Or is there a way to convert from sites to the input that VoronoiMesh3 accepts?

Hello！
How to add constraints on your delaunay code?Thank U!

Greetings,

I have a java background, and I'm looking at the UpdateAdjacency method with some confusion.
If you're resetting the tags to 0 and 1 at the beginning, won't the lv[j].Tag will be always be 0, and line 272 will force the method to exit?

Thanks,
Georgii

Hello!

I am trying to use DelaunayTriangulation3 to generate a mesh from point cloud data ( a set of vectors3) in Unity. I was able to have this result on the first picture where the left part is the desired result, and the right side is the generated mesh. I limited cell.radius < 2 in RegionsToMeshes() function. Nevertheless, as you may see it can not generate the floor. I notice that it is creating triangles between far points, like from the bottom (floor) to the high points(to the top of trees). How can we eliminate these undesired triangles?
.

In order to debug, I created 6 points that I want to be rendered in the end as in the figure, on the right side. You can see the corners as cubes on the left side and the generated triangles. Even though I do not remove any simplex and add all cells I am confused why only these triangles are rendered.

Could you help me how I can tackle these issues to render all possible triangles in which the vertexes are close to each other?

Cheers,
Beril

I'm trying to make free Unity LightProbe placement extension and stuck with finding adjacent probes for a given one. Unity Light Probes uses Delaunay tetrahedralisation to find adjacent probes as I could understand (https://www.gdcvault.com/play/1015522/Light-Probe-Interpolation-Using-Tetrahedral) so I tried using your code to find matching probes from a list of all probes in a group. It worked almost like it should! But in some cases some cells gets lost for some reason.
I attached the script I used for testing and a GIF representing the issue. I'll appreciate if you could point out where to look for solution. Then I'll be able to release some free lightprobe placement tool :)

LightProbeGroupDelaunau.txt