You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm encountering problems with filters.csf and filters.covariancefeatures.
A little bit of context : I had a scan separate in 8 tiles (but for this example, let's say we only have to, it will be easier). First, I run a first pipeline on these tiles to add the OriginId dimension for each point to identify which point correspond to which scan :
But there is a problem, the pipeline is ok on the wam-ob00000.las file but not on wam-ob00004.las, on this file the pipeline crash and stop.
After some researches I think the filers.covariancefeatures is responsible of the bug, but I really don't know why it is not working on the second file. I checked the pdal info output for wam-ob00004.las and nothing weird.
Looks like a couple of issue going on here. First, the eigen decomposition to obtain covariance features is failing. I wonder if a larger radius could help to get around this (or use fixed knn rather than radius). Second, we are not cleanly exiting all threads when the exception is thrown about the eigenvalues.
What does the solver.info() check ? Because if the solver.info() is not equal to "Success", I will see the message Cannot perform eigen decomposition in my console right ? So I wonder if the eigen decomposition gives "corrupted" values or something like that.
And for the threads who not exit cleanly, do you know which exception is thrown ?
Hi everyone,
I'm encountering problems with filters.csf and filters.covariancefeatures.
A little bit of context : I had a scan separate in 8 tiles (but for this example, let's say we only have to, it will be easier). First, I run a first pipeline on these tiles to add the OriginId dimension for each point to identify which point correspond to which scan :
I get these two files in output :
Then, I want to process this pipeline on these two files :
But there is a problem, the pipeline is ok on the
wam-ob00000.las
file but not onwam-ob00004.las
, on this file the pipeline crash and stop.After some researches I think the filers.covariancefeatures is responsible of the bug, but I really don't know why it is not working on the second file. I checked the
pdal info
output forwam-ob00004.las
and nothing weird.Did someone have an idea ?
System/installation information:
PDAL version :
ver :
Conda details :
The text was updated successfully, but these errors were encountered: