[ros-users] How to compare two point clouds?

Raymond Sheh rsheh at cse.unsw.edu.au
Fri Feb 18 14:11:03 UTC 2011

Hi Christan,

I guess it really depends what the purpose of your comparison is. What 
are you trying to do?

ICP and other related alignment techniques are useful if you're trying 
to fit data to a model (eg. a new scan into an existing map, a possible 
object into a scene, etc.) and generally also provides transform 
information but if you're trying to do other things with your comparison 
perhaps a different (and possibly simpler) technique may be more suited.

- Raymond

On 18/02/2011 7:02 PM, Christian Mueller wrote:
> Hello everybody,
> I have general question about 3D point cloud perception, since I am not
> really familiar in processing point clouds. I would like to measure the
> similarity of two point clouds ... a kind of fitness score between a
> reference point cloud and another point cloud.
> Currently I am checking the literature, however I couldn't find a
> "basic" algorithm. The only common way I found, is the alignment of both
> points clouds by their center of mass. Afterwards one of the point cloud
> is rotated till a best fit is achieved with the other point could.
> Is there maybe some other approach or technique?
> Thanks for any comment.
> Christian
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users

More information about the ros-users mailing list