CloudCompare forum

If you do a lot of computations on the same cloud with various parameters / classifiers, CC will try to store the last one (I believe) and then try to use it again the next time. But if the parameters / classifer have changed you'll need to recompute everything. I guess you should simply click on ok ;)

thanks you so much for the response!! :D

anyway, is there any difference of the result between the one i classify using cloud compare and "classify" tool in canupo suite?

If you speak of the original standalone tool then it's the exact same algorithm (but a parallelized version).

does canupo consider the orientation of the data? i mean, can it differ oblique and transverse plane that have the exact same density?

Sorry, I'm not sure to understand your question. But Canupo considers the local shape of a cloud whatever the orientation. Therefore I'm not sure that it can differentiate planar features with different orientations.

However, the qFacet plugin can be used to filter the points based on their (normal) orientation:
http://www.cloudcompare.org/doc/wiki/in ... ive_filter

Simple answer: no it does not :) But for planes I used to compute normals then Dip/Dip directions to separates vertical plans from horizontal ones ;)

Hello,

Sorry if the question has already been answered sooner in the thread :

Are there any news about a public release of qCanupo source code ?

Thanks
Thomas

Hi,

actually in the original command line version, for each classified point, we computed the slope (or dip) using the largest scale of the classifier. It was really useful to automatically filter out afterwards data for near horizontal or vertical structures (building, roads, river banks vs river bed...). I don't remember why we didn't implemented it in the CC version ?

Dimitri

Don't remember either :D.

We would have to add options in the dialog and create one or two new scalar fields...

is this slope used in the PCA process ? i mean with this information, can canupo really differ between horizontal and tilted plane ( like football field vs road ) ? or is it hopeless to differ two objects that are both totally look 2d eventhough the orientation is different?
do you think it is possible to classify this point cloud?