Geek asks and answers

How to determine the difference between two clouds of a 3D point?

I have two clouds of 3D points using the Cloud Cloud Library . One of them is a cloud reference point (let's call it A ), and the other with deformation (let's call it B ). Both point clouds are taken from objects that somehow do not have any or very small objects on the surface, except for the edges. These point clouds A and B are also aligned.

  • I want to know whether there is any algorithm that can detect the missing cloud of Bed and .
  • How to build a 3D-image high-resolution missing part of Bed and .

Helps to evaluate.

+3
source share
3 answers

I do not understand these things, so these are mainly ideas, not solutions, and I could be wrong.

But my naive approach would be logical operations / constructive solid geometry based on two meshes (see also the question from the gamedev ). If you calculate A-B, you get the grid (s) containing everything that is in A but not in B - or, in other words: the missing part B.

There are two questions with this approach:

  • Logical operations are complex due to floating point inaccuracies and special cases.
  • Your grids are noisy, so their surfaces will basically not match, even outside the missing region.

"" . , A - .

, , , clounds (, ), (, ). .

, OpenGL DirectX.

+1

" ", PCL.

:

octree- ( ) .

+2

( Kinect, AFAIK , ), . , ( Kinect , ), , , , . , , B, NaNs (, fooobar.com/questions/2040443/...) voila, B, A.

Although I know this approach is used, but I never used it myself and never played with Kinect. I think due to noise and small vibrations to the ground, the created mask can also be too noisy, especially at the edges and “silhouette” points of the scene, and it is there that image processing tools applied to depth masks help.

0
source

More articles:


All Articles