Johannes Schönberger
dd45f15ced
Use explicit keyword for dtype
2012-08-30 09:40:54 +02:00
Johannes Schönberger
bbeaec6b3f
Add examples to doc string of clear_border
2012-08-30 09:36:28 +02:00
JDWarner
99238c44a5
sqrt(gradients) line removed
2012-08-29 17:07:13 -05:00
JDWarner
61320957eb
Changes based on PR review recommendations: input format, scaling, and bugfix.
...
In this new version, all instances of 'spectrum' have been replaced with 'channel'. The documentation also reflects this change, and the new multichannel kwarg used to indicate multichannel input is named appropriately.
New boolean multichannel kwarg added, which controls if the input has multiple channels or not. Input 'data' is now array_like for both gray-level and multichannel. This kwarg is needed mainly because a 3-D array could be either 3 spatial dimensions or a set of different 2-D channels.
New scaling kwarg added (may be removed in future), controlling if data scaling is applied to ALL channels or each channel individually, if multichannel=True. No effect for gray-level data.
Removed np.sqrt(gradients) in _compute_weights_3d(), which was a bug. Tests now pass consistently.
New method for maintaining shape from input to output, where dims = data.shape prior to np.atleast_3d(). A theoretical (70,100,1) array passed should now result in a (70,100,1) shaped output, for example.
Updated and fixed multispectral test script to work with new version. TODO: Additional test(s) likely needed to cover code branches from new kwargs.
2012-08-29 16:33:56 -05:00
JDWarner
682d0535cd
Added multispectral random walker test.
...
Since the multispectral path is equivalent except for gradient calcs,
only one test case is needed. This test is modeled on the 3-D
non-multispectral version. If deemed necessary, adding a 2-D case
would be simple.
2012-08-27 13:41:41 -05:00
Johannes Schönberger
1177cf1393
Move clear_border to segmentation package
2012-08-27 18:56:05 +02:00
JDWarner
feca48cc49
Added return_full_prob kwarg to solve call if pyamg not present.
2012-08-27 11:48:32 -05:00
JDWarner
8955dad32e
Multispectral modifications applied to random walker.
2012-08-27 11:42:30 -05:00
Emmanuelle Gouillart
06065f640a
PEP 8: _quickshift module
2012-08-27 16:23:11 +02:00
Emmanuelle Gouillart
0293403b68
DOC: example in slic segmentation docstring
2012-08-27 12:57:24 +02:00
Stefan van der Walt
8d54380dc0
Merge pull request #259 from emmanuelle/fix_random_walker
...
ENH: Fix random walker.
2012-08-27 02:40:56 -07:00
Emmanuelle Gouillart
9d29d5df78
PEP8: indentation in random_walker_segmentation
2012-08-27 11:26:11 +02:00
Emmanuelle Gouillart
28161eaee6
ENH: better handling of labels that need to be reordered
...
(this is now done automatically)
2012-08-24 15:09:46 +02:00
Johannes Schönberger
a08779e06a
Use predefined header files from Cython
2012-08-21 15:15:27 +02:00
Stefan van der Walt
1c3aeec2a8
BUG: Remove uses of xrange for py3 compatibility.
2012-08-20 15:25:40 -07:00
Andreas Mueller
6b1dab9f9a
MISC move felzenszwalb_cy.pyx to _felzenszwalb_cy.pyx, don't use xrange when not necessary
2012-08-20 22:30:58 +01:00
Andreas Mueller
fe2a4334fa
ENH addressed (hopefully all) of Tony's and Stefan's comments.
2012-08-20 20:22:06 +01:00
Andreas Mueller
37c0ffe072
cosmit: changed the names of x and y to r and c. that was no fun, i tell you.
2012-08-20 20:01:41 +01:00
Andreas Mueller
8d769a4cd9
ENH Felzenszwalbs segmentation somewhat faster
2012-08-10 10:48:38 +01:00
Andreas Mueller
312b03d1b1
ENH more speeeeed
2012-08-10 10:35:23 +01:00
Andreas Mueller
8f5337a2bf
ENH added some cdefs
2012-08-10 10:31:23 +01:00
Andreas Mueller
f88a29b091
ENH minor speedups.
2012-08-10 10:03:20 +01:00
Andreas Mueller
f560346309
Trying to avoid name collisions.
2012-08-05 21:31:24 +01:00
Andreas Mueller
f421587aa4
ENH renamed "felzenszwalb_segmentation" to "felzenszwalb", remove debug output from slic
2012-08-05 21:10:29 +01:00
Andreas Mueller
73dd46019b
FIX width/height trouble, add non-regression test
2012-08-05 21:06:37 +01:00
Andreas Mueller
e0034e384a
Minor fixes addressing @tonysyu's comments.
2012-08-05 20:30:43 +01:00
Andreas Mueller
530f2c31c4
pep8
2012-08-05 20:25:34 +01:00
Andreas Mueller
8098718036
Fixup tests, add test for slic.
2012-08-05 14:33:30 +01:00
Andreas Mueller
b1eb4265ed
Post-processing for felzenszwalbs algorithm.
2012-08-05 13:52:15 +01:00
Andreas Mueller
e65d6f6635
Added smoothing option to quickshift
2012-08-04 20:27:31 +01:00
Andreas Mueller
10934cfaff
Fixed bug in SLIC smoothing
2012-08-04 20:27:18 +01:00
Andreas Mueller
d770fc714d
quickshift: convert to lab in function, some comments in code.
2012-08-04 19:56:22 +01:00
Andreas Mueller
1746d2c38c
Some scaling issues to be closer to reference implementation
2012-08-04 18:03:03 +01:00
Andreas Mueller
b6059b5672
Put RGB2Lab into slic as it seems to be essential.
2012-08-03 12:00:39 +01:00
Andreas Mueller
cd1007a0bc
Rename km_segmentation to slic. They have a PAMI paper now so I guess we should use their name.
2012-08-03 11:57:02 +01:00
Andreas Mueller
026b6b1df0
Fix initialization in km_segmentation, prettier examples
2012-08-03 11:43:03 +01:00
Andreas Mueller
d9a22d867b
Documentation, example for km_segmentation
2012-08-03 11:37:12 +01:00
Andreas Mueller
1c69adb817
MISC some simplifications, minor speedup
2012-08-03 11:37:11 +01:00
Andreas Mueller
501a6db8ad
ENH speedup, means and image use pointers
2012-08-03 11:37:11 +01:00
Andreas Mueller
8f24366790
starting cython implementation of km_segmentation
2012-08-03 11:37:11 +01:00
Andreas Mueller
7b646ad7ea
fix trying to make this implementation more like slic
2012-08-03 11:37:11 +01:00
Andreas Mueller
49bc44c6e9
FIXed test to work with the fixed "hashing" of colors
2012-08-03 11:37:11 +01:00
Andreas Mueller
ccfb89b957
FIX Tried to address @stefanv's comments on the PR.
2012-08-03 11:37:11 +01:00
Andreas Mueller
05cc863f3f
First draft for numpy based km_segmentation
2012-08-03 11:37:11 +01:00
Andreas Mueller
d2e226fe59
ENH tests for Felzenszwalbs segmentation, fixed off-by-one error
2012-08-03 11:37:11 +01:00
Andreas Mueller
a7c98cb67a
Remove felzenszwalb_segmentation_gray again since it just complicates the interface.
2012-08-03 11:37:11 +01:00
Andreas Mueller
f0a7212c4f
ENH Rename parameters in quickshift, add "ratio"
2012-08-03 11:37:11 +01:00
Andreas Mueller
08df2a5103
enh: minor simplifications
2012-08-03 11:37:11 +01:00
Andreas Mueller
8fa5427afc
FIX build problem and cython problem resolved.
2012-08-03 11:37:11 +01:00
Andreas Mueller
ce26467ad4
ENH: make quickshift more tolerant to input type, just convert to float. Also keep track of random seed for reproducable tests.
...
Finally, do a unique on the output and add testing.
2012-08-03 11:37:11 +01:00
Johannes Schönberger