integer wraparound in Cascade test case

[grlee77]

Description:

Testing with the recent NumPy 1.24rc1 pre-release revealed a potential issue with the XML file used for our test_detector_astronaut test case.

Specifically there is a warning about:
"DeprecationWarning: NumPy will stop allowing conversion of out-of-bound Python integers to integer arrays. The conversion of -67130709 to uint32 will fail in the future."

It seems that the data in the test XML file has negative values that have been getting silently wrapped to positive ones on cast to uint32.

Way to reproduce:

install NumPy>=1.24rc1 and run the test case:

`pytest feature/tests/test_cascade.py`

(note, the warning may soon be suppressed by changes in #6637 in which case the catch_warnings context would have to be removed from the test case to see it)

Traceback or output:

From the CI log [here](https://github.com/scikit-image/scikit-image/actions/runs/3567086648/jobs/5994344865)

FAILED feature/tests/test_cascade.py::test_detector_astronaut - DeprecationWarning: NumPy will stop allowing conversion of out-of-bound Python integers to integer arrays.  The conversion of -67130709 to uint32 will fail in the future.
For the old behavior, usually:
    np.array(value).astype(dtype)`
will give the desired result (the cast overflows).

Version information:

3.8.15 | packaged by conda-forge | (default, Nov 22 2022, 08:49:35) 
[GCC 10.4.0]
Linux-5.14.0-1054-oem-x86_64-with-glibc2.10
scikit-image version: 0.20.0.dev0+git20221128.f25ae8557
numpy version: 1.24.0rc1

[github-actions]

Hey, there hasn't been any activity on this issue for more than 180 days. For now, we have marked it as "dormant" until there is some new activity. You are welcome to reach out to people by mentioning them here or on our forum if you need more feedback! If you think that this issue is no longer relevant, you may close it by yourself; otherwise, we may do it at some point (either way, it will be done manually). In any case, thank you for your contributions so far!