Here is a quick test-case, for the project I am working on, I have 800k vertices from Lidar data on a terrain that I need to convert to a mesh. I want to visualise it in Blender and would like to smooth it at render time, for which the triangles pose a problem as they take a big performance hit.
Is it possible to get rid of the triangulation that is caused by the Delaunay Mesh component?
I am assuming that the quad remesh component is the best solution for this as it will also get rid of non-flat quads? Or is there an alternative solution to convert a point file (CSV) into a quad mesh?
You won’t want quads that fold in their center (which will happen unless you want to lose the shape of that). The display and shading will look pretty bad.
Instead of creating a Delaunay mesh and trying to get rid of triangles you can create a quad mesh from sorted points, using a mesh plane on the XY plane.
Yes indeed, which is why I am asking
Then again, I don’t want there to be too much of a deviation, which is why I think Quad Remesh may not be the best approach.
At least creating a point grid is more accurate than the Heightfield command…
I’ll give that a try, thanks. Will have to check whether the quads remain flat (not folded).
In general, does remeshing create flat quads in Rhino/ GH?
Yes of course, I understand that much
I was looking for a way to minimise deviation while getting a quad mesh. Anyways, the remesher works well, but is prone to causing five sided poles in the mesh. Is the algorithm based on instant meshes?
In the end, the drape command surface modelling ended up working better with GH, there’s no mesh project. Urbano has a project to terrain, but that’s pretty slow compared to surface project, so…
The author was supposedly working on integrating Quadwild, an open-source remesher. This addon already supports Instant Meshes and Blender’s Quadriflow.
