-
Notifications
You must be signed in to change notification settings - Fork 432
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Segmentation fault on filters.covariancefeatures #3848
Comments
Indeed, it looks like the OptimalKNN dimension ends up with 0's and the covariance features filter doesn't know how to handle these. Need to investigate why optimal neighborhoods is yielding a 0 and provide better error checking in the covariance filter. |
A workaround for now would be to use the assign filter to manually override OptimalKNN, e.g., to knn=8, where it is 0. This could be inserted into your pipeline between the two filters @digital-idiot. I have yet to determine why the updated knn and radius values are not being set (something with the entropy calculation, probably resulting in a NaN/inf). One thing we could do would be to at least fix the initialized values of knn and radius here: PDAL/filters/OptimalNeighborhoodFilter.cpp Lines 86 to 87 in 6cb54d8
It would make more sense to bind these to the min_k and distance to the min_kth neighbor. That way, if things fail for some reason, we at least have usable values. |
@chambbj Thank you for suggesting the workaround, it works well.
Do you have any suggestion how to debug this and pin point which points and why they are causing |
Workflow
filters.optimalneighborhood
filters.covariancefeatures
Pipeline
Encountered behavior
Expected behavior
Pipeline successfully generates processed point cloud or report appropriate error in the input dataset. The pipeline tested to be working till
filters.optimalneighborhood
stage. It fails onfilters.covariancefeatures
stage. Expermented with differentfeature_set
but outcome did not change.Reproduce
System/installation information:
Python environment information
The text was updated successfully, but these errors were encountered: