Closed EwingKang closed 1 year ago
I've been doing some bug tracing. Some related values are printed when pdb caught the assertion:
(Pdb) overlapped_size
[(-1, 22), (57, 22), (280, 22)]
(Pdb) unique
array([ -1, 57, 280], dtype=int32)
(Pdb) region
array([[ 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57],
[ 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57],
[ -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1],
[ -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1],
[280, 280, 280, 280, 280, 280, 280, 280, 280, 280, 280],
[280, 280, 280, 280, 280, 280, 280, 280, 280, 280, 280]],
dtype=int32)
It seems like ID: -1 is taken into account when there are multiple nearby groups. I'll try adding some patch to work around the issue.
The PR has been merged.
Describe the bug AssersionError with this specific image
Input Image Edit :removed copyrighted image, contact me privately (Note: I use the tiff version directly from the scanner, which explains the discrepancy between command and logs, I'm able to recreate this error with this uploaded PNG version)
Full Traceback
Command You Execute
PYTHONPATH=$(pwd) LOG_LEVEL=debug python3 oemer/ete.py ../sample_img/IMG_0005.tif --save-cache