I was recently working with the deformation of some models, I needed to do something like the Mesh Deform Modifier, but I wasn’t getting good results with it.
So I tried a custom implementation of something very similar to that modifier, in which I used Inverse Distance Weighting to calculate the “weight” of the cage deform for each vertex. The results were pretty good, and I’m wondering if it’s used somewhere in the codebase, for example it would be a good option for the “Transfer attributes” operation.
A related question: Can someone give me some hints about how the Mesh Deform and Lattice modifiers are implemented?
Not sure exactly what the Lattice modifier is based on, but that sort of thing is described in CG textbooks and graphics papers from the 80ies like this:
I don’t think we use inverse distance weighting specifically. There are alternatives like this too that may have a better balance between simplicity of implementation and quality: https://www.cs.wustl.edu/~taoju/research/meanvalue.pdf