Hello everyone!
I have a custom shape component and my goal is to generate it from the given point on the scene. This point is located inside some environment (room) and I need the result as a point cloud as precise to the scanned room as possible. The room can have any shape, both convex and not.
I have tried doing ray casts from the center, building voxel grid and traversing with BFS, but these techniques are not precise, angles are often cut and BFS can take too much time.
I also saw SLAM algorithm and OpenCV doing this for real-life robot scanners. Can you give some hints? It sounds like an easy task, replicating the given room shape, but appears to be not so trivial.
Thanks!