Hi,
Thank you very much for this great plugin!
Can someone please explain to me how to retrain a classifier with qCANUPO?
And is it possible to improve the classifier by training the classes with different point clouds or does the .prm file only consider a single point cloud for training?
Hope you get my question.
Regards
HollowHeartNet
How to retrain a classifier with qCANUPO
-
- Posts: 7
- Joined: Wed Mar 17, 2021 11:39 pm
Re: How to retrain a classifier with qCANUPO
What do you mean by 'retrain' exactly? You want to update an existing classifier? (in this case it's not possible)
And you can definitely use as many points as you wish to train a classifier (once) but you'll have to regroup them as a single cloud at the time of the training. So you segment several areas / parts that represent your class, and then you merge them before doing the training.
And you can definitely use as many points as you wish to train a classifier (once) but you'll have to regroup them as a single cloud at the time of the training. So you segment several areas / parts that represent your class, and then you merge them before doing the training.
Daniel, CloudCompare admin