Mailman 3 November 2012 - scikit-image

Re: Error importing feature: Issue with match_template function?
by Marianne Corvellec 30 Nov '12

30 Nov '12

Hi again, Oh, sorry! I skipped 'Rebuild and then try again?' which you wrote earlier... And I should have read the README more carefully. ;) I love READMEs... <3 It's all fine now. Thank you so much for your help! Marianne On Friday, November 30, 2012 9:37:22 PM UTC+1, Stefan van der Walt wrote: > > Hi Marianne > > It looks like your extension modules are not compiled. Did the build > complete successfully? > > Stéfan > On Nov 30, 2012 12:21 PM, "Marianne Corvellec" <marianne....(a)ens-lyon.org<javascript:>> > wrote: >

2 1

Re: Error importing feature: Issue with match_template function?
by Stéfan van der Walt 30 Nov '12

30 Nov '12

Hi Marianne It looks like your extension modules are not compiled. Did the build complete successfully? Stéfan On Nov 30, 2012 12:21 PM, "Marianne Corvellec" < marianne.corvellec(a)ens-lyon.org> wrote:

1 0

Re: Error importing feature: Issue with match_template function?
by Marianne Corvellec 30 Nov '12

30 Nov '12

Hi Stéfan, Thanks for the reply. Okay. The official repo is upstream for me. So $ git fetch upstream $ git reset --hard upstream/master HEAD is now at a77923f Merge pull request #373 from luispedro/imread_io_plugin But I'm getting the same import error. Actually, I'm a little confused. When I first try to import feature, I get an error related to the import of filter: >>> from skimage import feature Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/[...]/skimage/feature/__init__.py", line 5, in <module> from .template import match_template File "/home/[...]/skimage/feature/template.py", line 4, in <module> from . import _template File "_template.pyx", line 37, in init skimage.feature._template (skimage/feature/_template.c:4073) File "/home/[...]/skimage/transform/__init__.py", line 1, in <module> from .hough_transform import * File "/home/[...]/skimage/transform/hough_transform.py", line 8, in <module> from skimage import measure, morphology File "/home/[...]/skimage/measure/__init__.py", line 2, in <module> from ._regionprops import regionprops, perimeter File "/home/[...]/skimage/measure/_regionprops.py", line 6, in <module> from skimage.morphology import convex_hull_image File "/home/[...]/skimage/morphology/__init__.py", line 6, in <module> from .watershed import watershed, is_local_maximum File "/home/[...]/skimage/morphology/watershed.py", line 30, in <module> from ..filter import rank_order File "/home/[...]/skimage/filter/__init__.py", line 7, in <module> from ._denoise import denoise_bilateral ImportError: No module named _denoise And on the second try or later, the error is indeed the same as before the fetch and reset: >>> from skimage import feature Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/[...]/skimage/feature/__init__.py", line 5, in <module> from .template import match_template File "/home/[...]/skimage/feature/template.py", line 4, in <module> from . import _template ImportError: cannot import name _template (I displayed the error messages from the Python interpreter just because they are shorter than those from IPython.) ... Marianne On Friday, November 30, 2012 1:56:01 AM UTC+1, Stefan van der Walt wrote: > > Hi Marianne > > On Thu, Nov 29, 2012 at 9:01 AM, Marianne Corvellec > <marianne....(a)ens-lyon.org <javascript:>> wrote: > > I am having an Import Error when I try to import feature, namely: > > > > In [28]: from skimage import feature > > Could you please try the following? > > $ get fetch origin > $ git reset --hard origin/master > > Rebuild and then try again? You may have to replace "origin" with > "upstream" or whatever you called the official repo. > > Stéfan >

1 0

Re: Making algorithms at least 3D, preferably nD
by Emmanuelle Gouillart 30 Nov '12

30 Nov '12

Hi Juan, +1 for having 3-D support whenever it's easy to implement. I'm processing mostly 3-D images (from X-ray tomography), so I have a strong interest in algorithms compatible with 3-D images. Adding a paragraph to the contribute section sounds like a good idea. Cheers, Emmanuelle On Fri, Nov 30, 2012 at 03:45:49PM +1100, Juan Nunez-Iglesias wrote: > Hey Guys, > I mentioned this briefly at SciPy, but I would like to reiterate: a lot of > data is 3D images these days, and more and more data is being generated > that is multi-channel, 3D+t. Therefore, it would be awesome if > scikit-image started making more of an effort to support these. In the > best case, the dimension of the underlying array can be abstracted away � > see [1]here for example, the functions juicy_center (which extracts the > centre of an array, along all dimensions), surfaces (grabs the "border" > arrays along each dimension), hollowed (zeroes-out the centre), and more. > Otherwise, writing a 3D function that gracefully degrades to 2D when one > of the dimensions is 1 is also possible. > In general, the amount of additional effort to make code 3-, 4- or n- > dimensional is relatively low when you write the algorithm initially, > relative to refactoring a whole bunch of functions later. I'll try to > fiddle with whichever code I need, but in the meantime, what do you think > about adding a paragraph or a sentence about this issue in the > scikit-image [2]contribute section, so that people at least have this in > mind when they are thinking of writing something new? > Thanks, > Juan.

2 1

Making algorithms at least 3D, preferably nD
by Juan Nunez-Iglesias 30 Nov '12

30 Nov '12

Hey Guys, I mentioned this briefly at SciPy, but I would like to reiterate: a lot of data is 3D images these days, and more and more data is being generated that is multi-channel, 3D+t. Therefore, it would be awesome if scikit-image started making more of an effort to support these. In the best case, the dimension of the underlying array can be abstracted away — see here<https://github.com/jni/ray/blob/develop/ray/morpho.py#L328>for example, the functions juicy_center (which extracts the centre of an array, along all dimensions), surfaces (grabs the "border" arrays along each dimension), hollowed (zeroes-out the centre), and more. Otherwise, writing a 3D function that gracefully degrades to 2D when one of the dimensions is 1 is also possible. In general, the amount of additional effort to make code 3-, 4- or n- dimensional is relatively low when you write the algorithm initially, relative to refactoring a whole bunch of functions later. I'll try to fiddle with whichever code I need, but in the meantime, what do you think about adding a paragraph or a sentence about this issue in the scikit-image contribute <http://scikit-image.org/docs/dev/contribute.html> section, so that people at least have this in mind when they are thinking of writing something new? Thanks, Juan.

1 0

"No definition found"
by Juan Nunez-Iglesias 30 Nov '12

30 Nov '12

I'm getting this message when trying to find the definition header for skimage.segmentation.slic: In [12]: %pdef segmentation.slic No definition header found for segmentation.slic In [15]: %pdef segmentation.join_segmentations segmentation.join_segmentations(s1, s2) Anyone know why this is? Is this a general feature of Cython functions?

1 0

Re: Error importing feature: Issue with match_template function?
by Stéfan van der Walt 30 Nov '12

30 Nov '12

Hi Marianne On Thu, Nov 29, 2012 at 9:01 AM, Marianne Corvellec <marianne.corvellec(a)ens-lyon.org> wrote: > I am having an Import Error when I try to import feature, namely: > > In [28]: from skimage import feature Could you please try the following? $ get fetch origin $ git reset --hard origin/master Rebuild and then try again? You may have to replace "origin" with "upstream" or whatever you called the official repo. Stéfan

1 0

Help with find_boundaries or watershed
by Sigmund 27 Nov '12

27 Nov '12

Moin moin, I'm using the find_boundaries function for separating labeled regions (separated by the watershed). Afterward I set the as True returned pixels to 0. This works quit well but since my objects (labeled regions) can be small (couple pixels) and seldom stick together I would like to limit the find_boundaries to connected regions. Is there a easy way to only find the boundaries not connected to 0. Or tell the watershed function to leave a space between separate objects? Example: import numpy as np from scipy import ndimage as nd from skimage.segmentation.boundaries import find_boundaries a = np.zeros((15,15)) a[4:9,4:9]=1 a[6:10,7:12]=2 a[1:3,1:3]=3 print "a" print a.astype(int) a[find_boundaries(a)]=0 print a.astype(int) Output: [[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 3 3 0 0 0 0 0 0 0 0 0 0 0 0] [0 3 3 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 1 1 1 1 1 0 0 0 0 0 0] [0 0 0 0 1 1 1 1 1 0 0 0 0 0 0] [0 0 0 0 1 1 1 2 2 2 2 2 0 0 0] [0 0 0 0 1 1 1 2 2 2 2 2 0 0 0] [0 0 0 0 1 1 1 2 2 2 2 2 0 0 0] [0 0 0 0 0 0 0 2 2 2 2 2 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]] After find_boundaries: [[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 3 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 1 1 1 1 0 0 0 0 0 0] [0 0 0 0 0 1 1 0 0 0 0 0 0 0 0] [0 0 0 0 0 1 1 0 2 2 2 2 0 0 0] [0 0 0 0 0 1 1 0 2 2 2 2 0 0 0] [0 0 0 0 0 0 0 0 2 2 2 2 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]] As you can see, the region 3 is almost gone. Thanks Siggi

2 1

Re: Jet colormap to grayscale
by Guillaume CALMETTES 26 Nov '12

26 Nov '12

Hi Stefan, Works perfectly. However, I'm not sure why but I had to initialize the jet colormap object to create the _lut array. I replaced: lut = plt.cm.jet._lut[..., :3] By: jet = plt.cm.jet jet._init() lut = jet._lut[..., :3] Thanks a lot. Guillaume On Nov 25, 2012, at 8:25 PM, Stéfan van der Walt <stefan(a)sun.ac.za> wrote: > Hi Guillaume > > On Thu, Nov 22, 2012 at 8:08 PM, Guillaume CALMETTES > <guillaume.calmettes(a)gmail.com> wrote: >> Do someone has any trick to convert a "jet image" to grayscale? > > Here's a snippet that does it by brute force lookup: > > https://gist.github.com/4146612 > > Hope that helps! > Stéfan > > -- > >

1 0

Jet colormap to grayscale
by Guillaume CALMETTES 26 Nov '12

26 Nov '12

Hello, I am looking to convert an image saved with the jet colormap to grayscale. I tried to play a bit with the color module of skimage, but for the moment I couldn't manage to to map the jet colormap to linearly go from dark to light shades of gray. The main problem when applying a basic rgb2gray conversion is that the min/max values (blue and red) converge to the same dark value in the grayscale image, while the mid-range values (yellow/green) become the brightest. Do someone has any trick to convert a "jet image" to grayscale? Thanks a lot Guillaume PS: I have attached a picture as an example if you're willing to play with it ;)

3 3