c6d0acfdab
Initial commit of adapthist modifications.
2012-12-07 17:22:33 -06:00
4f68f1cd36
doc: fix and clarify peak_local_max docstring
2012-12-03 20:23:01 -06:00
50c0399aa7
Merge pull request #379 from cpsaltis/fix_rotate
...
Fix resize argument handling in rotate.
2012-12-03 14:48:52 -08:00
68c2ab205e
BUG: WindowsError not defined under Linux.
2012-12-03 08:49:35 -08:00
298abe1ac0
Add test for resize option in rotate
2012-12-03 15:54:26 +02:00
6953785eab
added rgb2lab test
2012-12-02 22:51:16 -08:00
bf28a38f6e
ENH: Make black the default boundary color for mark_boundaries.
...
This improves visibility in light-colored image regions. Note also that the new behavior matches the behavior from the 0.7 release.
2012-12-02 16:02:27 -05:00
2e1729a9fb
Cleaned up adapthist func to use LAB for all RGB->required rest result change
2012-12-01 23:46:52 -06:00
22f94d8707
Fix resize argument handling in rotate.
...
It used to have exactly the opposite effect that it supposed to.
2012-12-01 12:30:40 +02:00
b5d1e9f209
FIX quickshift nosetest with different xyz color conversion
2012-11-30 15:36:01 +01:00
43eb33d26b
flake8 on colorconv.py and removed unneeded .copy()
2012-11-30 01:30:59 -08:00
5108d8c346
Fixed RGB<->XYZ conversion.
...
Now correctly follows the algorithm from EasyRGB.
Also fixed a XYZ<->Lab conversion issue by adding
a .copy().
2012-11-29 12:06:01 -08:00
b906082034
Add testing functions for join and relabel
2012-11-28 12:48:13 +11:00
dc377970f5
Add functions to join and relabel segmentations
...
The join of two segmentations is the segmentation in which two voxels
are in the same segment if and only if they are in the same segment in
both input segmentations.
2012-11-28 12:22:14 +11:00
98ce475113
style: initialize out at beginning of function
...
* remove API compatibility comment
* remove unnecessary `del maskim`
2012-11-25 23:09:06 -06:00
45a91aa9c5
doc: deprecated string and docstring formatting fixes in is_local_maximum
2012-11-25 22:56:45 -06:00
d2262227d9
BUG - rank_order now imported.
...
Also changed (peak.py):
* Standardized documentation as requested.
* Removed `threshold` as an optional kwarg.
* Removed extra line break incorrect by PEP8 standards.
Also changed (watershed.py):
* Added @deprecated decorator and import statement
2012-11-24 17:24:14 -06:00
38f699471e
TST Remove unneeded test
...
Per stefanv's comment on github, this was testing functionality higher
up in skimage.io
2012-11-24 19:04:55 +00:00
10ce9b89ab
RFCT Use skimage.utils.convert for type conversion
2012-11-24 19:04:16 +00:00
6ebc43bad8
ENH Add imread io plugin
...
This relies on imread (https://github.com/luispedro/imread ) which
supports a few microscopy formats otherwise hard to read (LSM and
MetaMorph STK) as well as basic ones.
2012-11-24 19:04:01 +00:00
e7288f9140
BUG - Dtype conversion to Boolean in a new code path not included in current tests.
2012-11-20 01:17:42 -06:00
da5a2a8221
BUG - Now parsing all unique labels != 0 properly, and set threshold_rel=0 in is_local_maximum() wrapper.
2012-11-20 01:11:31 -06:00
233b368195
BUG - Fixed a bug where the image was not properly masked for each unique label.
2012-11-20 00:45:01 -06:00
33c9514062
BUG - Fixed a bug where, when indices = False, output was not Boolean.
2012-11-20 00:36:12 -06:00
226eaaff92
FEAT - Unified peak finder backend
...
Detailed changelist:
* Fully documented new API for peak_local_max()
* Removed commented-out old code for is_local_maximum()
* Added "See also" and "Notes" to is_local_maximum() redirecting to peak_local_max()
2012-11-20 00:12:47 -06:00
63b5c5a4a0
FEAT - combined API from is_local_maximum() into peak_local_max()
...
is_local_maximum() is a wrapper function for peak_local_max()
is_local_maximum() runs much faster (~20% of prior runtime, nearly = to peak_local_max())
All tests in .feature and .morphology subpackages pass as written with these changes.
Todo:
* Fully document API
* remove commented-out old algorithm in is_local_maximum()
* add new tests for full coverage of new, more complex peak_local_max()
2012-11-19 23:38:58 -06:00
13c61b9694
Merge pull request #361 from ahojnnes/hough-peaks
...
Add hough transform peak detection
2012-11-17 07:30:08 -08:00
9d653fec78
BUG: Clean up imports
2012-11-15 23:53:47 -05:00
0d411e21dc
STY: Refactor parts of LineTool to CanvasToolBase
2012-11-15 22:04:02 -05:00
d219a76f58
BUG: Fix blitting behavior in canvastools
2012-11-15 21:56:51 -05:00
5be5813070
STY: Reuse ThickLineTool in LineProfile
2012-11-15 00:16:19 -05:00
2ff3e78c52
BUG: Move callback connection to parent class
2012-11-15 00:16:19 -05:00
4f6f25efe9
ENH: Add canvastools subpckages with base class and line tools
2012-11-15 00:16:19 -05:00
80a9a5aba5
Apply numpy reference style to rank filter descriptions
2012-11-14 08:33:27 +01:00
70f2b67ef7
fix: ref and test_otsu doc
2012-11-14 08:00:59 +01:00
698d6ae13f
fix:< Py_ssize_t > 0 and & pop
2012-11-13 17:07:21 +01:00
411d4686d5
fix:wrap lines in tests
2012-11-13 17:02:08 +01:00
f133158792
autopep8 for plot_* and rank sources
2012-11-13 09:37:58 +01:00
4a0b22aff5
Use common spelling for greylevel
2012-11-12 20:07:02 +01:00
eaf48a71e9
Use numpy reference style
2012-11-12 19:31:06 +01:00
a376a07c3b
PEP8 fixes in doc string examples
2012-11-12 19:29:00 +01:00
fce2ae3f5f
Fix spelling of X-bit
2012-11-12 19:25:46 +01:00
f724b97f90
Fix typos and spelling of readme
2012-11-12 19:24:02 +01:00
da391089cb
add: local Otsu synthetic example
2012-11-12 14:48:38 +01:00
db2e066f4c
add: test otsu
2012-11-12 14:40:49 +01:00
add437a9f2
add: test entropy 8- & 16-bit
2012-11-12 10:39:12 +01:00
c9f1b67c7b
fix: check pop==0 for some kernels
2012-11-12 10:05:52 +01:00
fd82ac9959
add new tests
2012-11-12 09:54:10 +01:00
a09ec78977
remove local demo from git
2012-11-12 09:34:33 +01:00
08b748c379
local Otsu returns now the threshold values
2012-11-11 14:39:09 +01:00
Steven Silvester