2f05992148
Expand on skimage support for 3D processing
2015-01-22 13:01:04 +11:00
64f00bb860
Add note on 5D images
2015-01-22 12:49:28 +11:00
6eb8b0d079
Add table summarising image types
2015-01-22 12:47:44 +11:00
738953b946
image_unwrapped created as ndarray always
...
In unwrap_phase the image_unwrapped variable is created as an
ndarray even when image is a masked array.
2015-01-21 19:37:23 -06:00
110ac37dc4
Add doctest skip directives to timing calls
2015-01-22 11:57:49 +11:00
ea989e2456
Fix a few typos
2015-01-22 11:38:37 +11:00
1439341461
Remove unused cmorph extension
...
The C morphology module has been superseded by wrapping scipy.ndimage.
2015-01-22 11:38:37 +11:00
658a83d46c
Don't use backticks to refer to functions
...
See the "See Also" section in the NumPy Docstring Conventions
document:
https://github.com/numpy/numpy/blob/master/doc/HOWTO_DOCUMENT.rst.txt
2015-01-22 11:38:37 +11:00
2d0a3eff33
Expand explanation of _invert_selem
2015-01-22 11:38:37 +11:00
9eed552ee7
Replace em-dash with regular dash in comment
2015-01-22 11:38:37 +11:00
7935e51596
Clarify and fix docstrings
2015-01-22 11:38:37 +11:00
ac5c0c30a9
Fix incorrect call to mark_boundaries in test
2015-01-22 11:38:37 +11:00
9b9730b058
Specify mode for mark_boundary tests
2015-01-22 11:38:37 +11:00
6586d1eb3f
Test subpixel mark_boundaries
2015-01-22 11:38:37 +11:00
307d4b6137
Simplify code for coloring boundaries
2015-01-22 11:38:37 +11:00
242dbf9b2b
Don't outline boundaries by default
...
mark_boundaries used to draw outlines by default. This was necessary
partly because of the asymmetric way boundaries were being drawn. Now,
boundaries are either thick or drawn around background or drawn with
subpixel precision, so I suggest that outlines are less needed.
2015-01-22 11:38:37 +11:00
cb2002c367
Allow custom background_label in mark_boundaries
2015-01-22 11:38:36 +11:00
257502c099
Prevent unnecessary array copy for RGB images
2015-01-22 11:38:36 +11:00
28591cc3ee
Allow different modes in mark_boundaries
2015-01-22 11:38:36 +11:00
a406f8dd2a
Make subpixel find_boundary doctest more readable
2015-01-22 11:38:36 +11:00
a55e3346f7
Set up correct type for find_boundaries test
2015-01-22 11:38:36 +11:00
fb893c277d
Add correct subpixel boundary estimation.
2015-01-22 11:38:36 +11:00
9c78fce4cd
Remove unused deprecation warning
2015-01-22 11:38:36 +11:00
32a960db0d
Fix 'outer' mode of find_boundaries
2015-01-22 11:38:36 +11:00
ff1631c717
Update find_boundaries with new modes
2015-01-22 11:38:36 +11:00
d8555142da
Update binary morphology tests, which no longer warn
2015-01-22 11:38:36 +11:00
d538abdb97
Fix order kwarg for some versions of numpy.copy
...
It seems it isn't always a valid kwarg. This bypasses that problem
by calling `np.array` directly (which is just what `np.copy` does
under the hood anyway).
2015-01-22 11:38:36 +11:00
b1891dc24e
Make find_boundaries symmetric
...
Fixes #738
2015-01-22 11:38:35 +11:00
f321fefb8b
Fix morpho open/close edge bugs by padding image
2015-01-22 11:38:35 +11:00
92433b2961
Fix and test 0-end bug in crop
2015-01-22 11:38:35 +11:00
893c3be289
Import crop into util package namespace
2015-01-22 11:38:35 +11:00
c96fdb9338
Fix automatic copy even when copy=False in crop
2015-01-22 11:38:35 +11:00
d7138f1a51
Fix incorrect crop tests
2015-01-22 11:38:35 +11:00
eaa2aec3cc
Add test functions for crop
2015-01-22 11:38:35 +11:00
fb7ed1d253
Add crop function, inverse to pad
2015-01-22 11:29:09 +11:00
97770e9fa5
Minor update to binary morphology docstrings
2015-01-22 11:29:09 +11:00
04536cc7df
Update morphology dtypes tests
...
Previously, we were testing that any dtype would get converted to uint8
and then correctly processed. Now, since we are using ndimage, we are
directly processing all dtypes. I've updated the tests accordingly.
2015-01-22 11:29:08 +11:00
1fa6904c04
Workaround for strange ndimage selem inversion
2015-01-22 11:29:08 +11:00
8ffed5e424
Wrap scipy.ndimage binary morphology functions
2015-01-22 11:29:08 +11:00
56764f8209
Wrap scipy.ndimage grey morphology functions
2015-01-22 11:29:08 +11:00
8eebb17bac
No SegFault when unwrapping a 3D fully masked array
...
The unwrap_phase function no longer causes a Segmentation fault when passed
a 3D array in which all elements are masked.
Closes #1347
2015-01-21 16:45:23 -06:00
d3024b5c73
No SegFault when unwrapping a 2D fully masked array
...
The unwrap_phase function no longer causes a Segmentation fault when passed
a 2D array in which all elements are masked.
2015-01-21 16:36:58 -06:00
de01ea56ae
unwrap_phase returns masked array when image is masked
...
With commit e3b84ed
2015-01-21 16:23:50 -06:00
eef3734bfe
Remove end of line whitespace in test_unwrap.py
2015-01-21 16:03:18 -06:00
8ebe073ad5
BUG: unwrap_phase on compressed masked array
...
unwrap_phase works on masked arrays with a compressed mask (no elements masked)
Closes #1346
2015-01-21 16:01:13 -06:00
9b16b35d96
Merge pull request #1318 from blink1073/pyqt5
...
Update viewer for PyQt5 compatibility
2015-01-21 13:27:40 -08:00
acf68c6d7c
Add new parameter to catch exceptions during RANSAC
2015-01-21 14:10:09 -05:00
91c697c5f7
Handle special case of 1 pixel image
2015-01-21 14:10:09 -05:00
5dbb6e3257
Improve conditioning of silimarity transform design matrix
2015-01-21 14:10:09 -05:00
983919404f
Improve conditioning of homography design matrix
2015-01-21 14:10:09 -05:00
Juan Nunez-Iglesias