Cannot catch MemoryError

[dcox1776]

Description:

When running marching_cubes on a stack of large images I encounter MemoryError, Exception ignored in 'skimage.measure._marching_cubes_lewiner_cy.Cell._increase_size_faces'

The error occurs at:
venv\Lib\site-packages\skimage\measure_marching_cubes_lewiner.py", line 181, in _marching_cubes_lewiner
vertices, faces, normals, values = func(volume, level, L,

I have enclosed my call to marching_cubes inside a try except. However, rather than throwing an exception that I can catch, the marching_cubes seems to be stuck in a loop and keeps outputting the following lines:

MemoryError:
MemoryError
Exception ignored in: 'skimage.measure._marching_cubes_lewiner_cy.Cell._increase_size_faces'
Traceback (most recent call last):
File "d:\Development\myproject\venv\Lib\site-packages\skimage\measure_marching_cubes_lewiner.py", line 181, in _marching_cubes_lewiner
vertices, faces, normals, values = func(volume, level, L,
^^^^^^^^^^^^^^^^^^^^^^

Ideally, marching_cubes will manage memory better and not generate these errors. From what I can tell it uses a simple doubling to reserve memory. If this is not possible, then at least throw an exception that can be caught rather than uncontrollably sending these messages.

Way to reproduce:

There is no easy way to reproduce this error. You need a stack of a large number of very large images. For me, I encounter the problem with 1000 images that are 3287 x 2215 16-bit grayscale images running marching_cubes with a step size of 1. My machine has an nvidia 4070ti with 12GB and 128GB of primary memory.

Version information:

>>> import sys; print(sys.version)
3.11.2 (tags/v3.11.2:878ead1, Feb  7 2023, 16:38:35) [MSC v.1934 64 bit (AMD64)]
>>> import platform; print(platform.platform())
Windows-10-10.0.22631-SP0
>>> import skimage; print(f'scikit-image version: {skimage.__version__}')
scikit-image version: 0.22.0
>>> import numpy; print(f'numpy version{numpy.__version__}')
numpy version1.26.4