I'm trying to recreate the view of a LIDAR scan in the woods (reference in attachment). For the test, I loaded several tree types as Packed Geometry and distributed them across the plane. Next I want to add various bumps, rocks and branches for detail.

The most obvious solution for lidar simulation was to use POP Solver. This decision seemed like a very long time in the simulation to me, but maybe I'm doing something wrong. I also ran into a problem: Packed geometry is not defined as a collider object by particles. It is important for me to maintain high detail of objects

Are there more elegant solutions to this problem?

I came up with the idea to do point-scattering on each geometry, then pack the points in the final scene and do something like backface-culling for the object (conditional LIDAR). Q: how to do this for Packed geometry?

How about you unpack the geometry, place some high resolution spheres, keep only their points and shoot those into the scene with a Ray SOP?

Konstantin Magnus
How about you unpack the geometry, place some high resolution spheres, keep only their points and shoot those into the scene with a Ray SOP?

Perfect! Thank you so much

Here is an example file if someone needs it.

Animated LIDAR scan:

Konstantin Magnus
Animated LIDAR scan:

Looks really cool! I'm playing around with forest scanning simulation now and I have a question: Is it possible to build Explicit Constraints between generated points to use it for POP Grains dynamics node?

Here is a tutorial on creating the Lidar 3D scanning look: https://youtu.be/IbZRDFSz9hQ [youtu.be]