Hi,
I'm having a problem in CC canupo classifier. When I make the cloud point classification with two samples (grass and rocks) in a scene, the CC sends me a message like that: "some descriptors couldn't be computed (min scale may be too small).
I don't really understand what it means and it seems to me that's not so complicated, it's just that I'm a new user so don't know how to do it well.
Thanks in advance,
Bruna.
[qCANUPO] "Descriptors" problem
Re: [qCANUPO] "Descriptors" problem
In fact the descriptors of CANUPO are computed by looking at the cloud 'shape' around each core point. To do this a minimum number of points is required (for instance if the core point is isolated or if it has only one or two neighbors, it's impossible to guess anything).
Therefore if your cloud is not dense enough there may not be enough points at the smallest scales.
Mind that it is only a warning (since it's very likely that a few core points are isolated even in very dense point clouds). But if too many core points are isolated you won't get any reliable information.
Therefore if your cloud is not dense enough there may not be enough points at the smallest scales.
Mind that it is only a warning (since it's very likely that a few core points are isolated even in very dense point clouds). But if too many core points are isolated you won't get any reliable information.
Daniel, CloudCompare admin
-
- Posts: 21
- Joined: Wed Feb 19, 2014 10:28 am
Re: [qCANUPO] "Descriptors" problem
Ok, right.
I'll try again later cause I guess it was a density problem, indeed. It also does make sense to me, thanks for your explanation about the descriptors.
Bruna.
I'll try again later cause I guess it was a density problem, indeed. It also does make sense to me, thanks for your explanation about the descriptors.
Bruna.