Mailman 3 April 2013 - scikit-image

Re: Image Feature Suggestion
by Stéfan van der Walt 19 Apr '13

19 Apr '13

On Thu, Apr 18, 2013 at 9:24 PM, Ronnie Ghose <ronnie.ghose(a)gmail.com> wrote: > So first off is there a sort of to-do list? https://github.com/scikit-image/scikit-image/wiki/Requested-features > Secondly, should scikit-image be > used in conjunction with OpenCV? I thought in general it would be standalone > like a few of the other scikits or numpy/scipy (standalone in the sense of > only using other python libraries as dependencies). There's no reason not to combine them. `scikit-image` aims to provide a highly Pythonic interface to commonly used image processing algorithms. Play around a bit with OpenCV, and you'll soon discover what that means :) Stéfan

1 0

Re: Cython memoryviews
by Stéfan van der Walt 19 Apr '13

19 Apr '13

On Fri, Apr 19, 2013 at 8:08 AM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> wrote: > I want to write a function that can take an array of any dimension. From > your response I'm guessing the best way is to write a wrapper in Python and > write a Cython function that takes a 1D array and a shape tuple? The example I gave does take an array of any dimension? Stéfan

1 0

Re: Cython memoryviews
by Juan Nunez-Iglesias 19 Apr '13

19 Apr '13

On Thu, Apr 18, 2013 at 7:37 PM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote: > Yes, the ::1 indicates the contiguous dimension, and can only be used > in the first and last places. > Thanks! I'm not sure I understand your last question correctly, I want to write a function that can take an array of any dimension. From your response I'm guessing the best way is to write a wrapper in Python and write a Cython function that takes a 1D array and a shape tuple?

1 0

Padding backend
by Josh Warner 19 Apr '13

19 Apr '13

It appears there are several scikit-image functions which have need of padding input. As far as I can see, these generally do so internally, with functions or routines that aren't exposed. From a quick search, the following functions in scikit-image all pad using their own internal methods: feature.match_template : mean value filter.lpi_filter : asymmetric zeros morphology.greyreconstruct : symmetric zeros or constant value morphology.watershed : pad child function for symmetric zeros transform.radon : symmetric zeros More broadly, there doesn't appear to be a publicly exposed API for padding NumPy image arrays in any packages I've come across. scipy.ndimage clearly has this functionality, via the common mode parameters, but it isn't exposed for general use. I've put some effort into making a pure Python 2d / 3d capable padding function for personal use, and most of the scipy.ndimage modes are available. Preliminary gist here wouldn't be hard to wrap into a PR (likely not optimal, but functional): https://gist.github.com/JDWarner/a26f77dae983fa1faba9 Would such functionality be desirable?

2 2

Re: Image Feature Suggestion
by Josh Warner 18 Apr '13

18 Apr '13

I've found that within the universe of SciPy +extensions (like the various SciKits, really anything that likes NumPy arrays) it's fairly easy to get things to play nice without much trouble. Interlacing algorithms from there with OpenCV, SimpleITK, VTK, and the like are more of a chore. They all exist in their own worlds as far as internal datatypes, syntax structure, etc. Nowhere near as friendly to hack together (it can be done, with order(s) of magnitude more effort). However, certain packages in the SciPy universe use these, usually wrapped such that the user never sees it. E.g. Mayavi uses VTK heavily, and Scikit-Image now has a SimpleITK plugin for io - but they're wrapped to users only have to deal with Python/NumPy. In my opinion there is definite value in ports, though they should probably take somewhat lower priority than new/novel work. On Thursday, April 18, 2013 2:24:29 PM UTC-5, Ronnie Ghose wrote: > > So first off is there a sort of to-do list? Secondly, should scikit-image > be used in conjunction with OpenCV? I thought in general it would be > standalone like a few of the other scikits or numpy/scipy (standalone in > the sense of only using other python libraries as dependencies). > > > On Thu, Apr 18, 2013 at 12:50 PM, Zachary Pincus <zachary...(a)yale.edu<javascript:> > > wrote: > >> > Would this be a useful thing to add then? I thought it was interesting >> you made references to OpenCV rather then internal options. >> >> Yeah, but no need to reinvent the wheel. OpenCV is really focused on >> algorithms for cameras-moving-around-in-the-world, which big chunks of your >> problem are. So if you want to port some of that code over to skimage, that >> would be a cool and useful contribution! But if you just wanted to solve >> your problem ASAP, then OpenCV might be your best resource... >> >> Zach >> >> PS. Fun story: a while ago I got a sparse optical-flow algorithm (using >> cross-correlation) running on a GPU. It was totally non-portable (pre-CUDA, >> custom-shader code), and it didn't actually work precisely right, but I >> think that it's something that shouldn't be too much trouble to actually >> make happen these days... >> >> >> > On Thu, Apr 18, 2013 at 8:19 AM, Stéfan van der Walt <ste...(a)sun.ac.za<javascript:>> >> wrote: >> > On Wed, Apr 17, 2013 at 11:16 PM, Zachary Pincus >> > <zachary...(a)yale.edu <javascript:>> wrote: >> > > Identifying features to track between frames is a pretty classic task >> in computer vision. Things that come to mind would include the simple >> Harris corner detector (in skimage), or more recent things like SIFT >> features and its variants like SURF etc. (DAISY is in skimage) might be of >> use. All of these tools should be in OpenCV too, IIRC. (Time to spend a >> while on wikipedia reading about these things and related, if it's all >> totally unfamiliar.) >> > >> > Having binary features (there are many) and Star Features (also known >> > as "Consensus") in skimage would be very valuable. >> > >> > > Alternately, you could try optical flow based techniques to densely >> track local pixel patches between frames. >> > >> > That's also a fun application, and a good way to learn Cython! >> > >> > Stéfan >> > >> > -- >> > You received this message because you are subscribed to the Google >> Groups "scikit-image" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> an email to scikit-image...(a)googlegroups.com <javascript:>. >> > For more options, visit https://groups.google.com/groups/opt_out. >> > >> > >> > >> > >> > -- >> > You received this message because you are subscribed to the Google >> Groups "scikit-image" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> an email to scikit-image...(a)googlegroups.com <javascript:>. >> > For more options, visit https://groups.google.com/groups/opt_out. >> > >> > >> >> -- >> You received this message because you are subscribed to the Google Groups >> "scikit-image" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to scikit-image...(a)googlegroups.com <javascript:>. >> For more options, visit https://groups.google.com/groups/opt_out. >> >> >> >

2 1

Re: Image Feature Suggestion
by Zachary Pincus 18 Apr '13

18 Apr '13

> Would this be a useful thing to add then? I thought it was interesting you made references to OpenCV rather then internal options. Yeah, but no need to reinvent the wheel. OpenCV is really focused on algorithms for cameras-moving-around-in-the-world, which big chunks of your problem are. So if you want to port some of that code over to skimage, that would be a cool and useful contribution! But if you just wanted to solve your problem ASAP, then OpenCV might be your best resource... Zach PS. Fun story: a while ago I got a sparse optical-flow algorithm (using cross-correlation) running on a GPU. It was totally non-portable (pre-CUDA, custom-shader code), and it didn't actually work precisely right, but I think that it's something that shouldn't be too much trouble to actually make happen these days... > On Thu, Apr 18, 2013 at 8:19 AM, Stéfan van der Walt <stefan(a)sun.ac.za> wrote: > On Wed, Apr 17, 2013 at 11:16 PM, Zachary Pincus > <zachary.pincus(a)yale.edu> wrote: > > Identifying features to track between frames is a pretty classic task in computer vision. Things that come to mind would include the simple Harris corner detector (in skimage), or more recent things like SIFT features and its variants like SURF etc. (DAISY is in skimage) might be of use. All of these tools should be in OpenCV too, IIRC. (Time to spend a while on wikipedia reading about these things and related, if it's all totally unfamiliar.) > > Having binary features (there are many) and Star Features (also known > as "Consensus") in skimage would be very valuable. > > > Alternately, you could try optical flow based techniques to densely track local pixel patches between frames. > > That's also a fun application, and a good way to learn Cython! > > Stéfan > > -- > You received this message because you are subscribed to the Google Groups "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > > > > -- > You received this message because you are subscribed to the Google Groups "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > >

2 1

Re: Image Feature Suggestion
by Ronnie Ghose 18 Apr '13

18 Apr '13

Would this be a useful thing to add then? I thought it was interesting you made references to OpenCV rather then internal options. Thanks, Ronnie On Thu, Apr 18, 2013 at 8:19 AM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote: > On Wed, Apr 17, 2013 at 11:16 PM, Zachary Pincus > <zachary.pincus(a)yale.edu> wrote: > > Identifying features to track between frames is a pretty classic task in > computer vision. Things that come to mind would include the simple Harris > corner detector (in skimage), or more recent things like SIFT features and > its variants like SURF etc. (DAISY is in skimage) might be of use. All of > these tools should be in OpenCV too, IIRC. (Time to spend a while on > wikipedia reading about these things and related, if it's all totally > unfamiliar.) > > Having binary features (there are many) and Star Features (also known > as "Consensus") in skimage would be very valuable. > > > Alternately, you could try optical flow based techniques to densely > track local pixel patches between frames. > > That's also a fun application, and a good way to learn Cython! > > Stéfan > > -- > You received this message because you are subscribed to the Google Groups > "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > >

1 0

Re: multi-page TIFF writer
by Zachary Pincus 18 Apr '13

18 Apr '13

OK, unittest added: https://github.com/scikit-image/scikit-image/pull/508 Also, BTW, I get the below test failure when I have matplotlib but no PIL installed. I think test_imread_url or the test right before it must be making some incorrect assumptions. (If I re-run the tests after this failure, then also all of the tifffile tests fail, so it seems that some test isn't doing proper teardown of the image loading plugins?) Zach ERROR: test_io.test_imread_url ---------------------------------------------------------------------- Traceback (most recent call last): File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/nose-1.1.2-py2.7.egg/nose/case.py", line 197, in runTest self.test(*self.arg) File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/scikit_image-0.9dev-py2.7-macosx-10.6-intel.egg/skimage/io/tests/test_io.py", line 27, in test_imread_url image = io.imread(image_url) File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/scikit_image-0.9dev-py2.7-macosx-10.6-intel.egg/skimage/io/_io.py", line 136, in imread img = call_plugin('imread', f.name, plugin=plugin, **plugin_args) File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/scikit_image-0.9dev-py2.7-macosx-10.6-intel.egg/skimage/io/_plugins/plugin.py", line 100, in call return func(*args, **kwargs) File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/matplotlib/pyplot.py", line 1750, in imread return _imread(*args, **kwargs) File "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/matplotlib/image.py", line 1193, in imread raise ValueError('Only know how to handle extensions: %s; with PIL installed matplotlib can handle more images' % handlers.keys()) ValueError: Only know how to handle extensions: ['png']; with PIL installed matplotlib can handle more images On Apr 18, 2013, at 8:47 AM, Juan Nunez-Iglesias wrote: > I think Zach knows better than I what the source of the error was, so he should probably handle it. ;) > > > On Thu, Apr 18, 2013 at 12:52 PM, Stéfan van der Walt <stefan(a)sun.ac.za> wrote: > Hi Juan > > > On Apr 18, 2013 2:35 AM, "Juan Nunez-Iglesias" <jni.soma(a)gmail.com> wrote: > > Yep, I've confirmed that your PR solved the problem for me! So I'll spare you that backtrace, Stefan. ;) > > Great! Now, for some extra brownie points and our eternal gratitude: a unit test? > > Stéfan > > -- > You received this message because you are subscribed to the Google Groups "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > > > > -- > You received this message because you are subscribed to the Google Groups "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > >

2 4

Re: multi-page TIFF writer
by Zachary Pincus 18 Apr '13

18 Apr '13

> On Thu, Apr 18, 2013 at 5:21 PM, Zachary Pincus <zachary.pincus(a)yale.edu> wrote: >> I don't think I know what I did (if anything) to provoke this... > > Nope! Not your fault. In these situations, doing another push or > close/open or rebase will all re-trigger the tests. Oh great, thanks. Is there a clean / recommended way to do a nop push / close/open / rebase? My git is limited... sorry for the trouble here!

2 1

Re: multi-page TIFF writer
by Zachary Pincus 18 Apr '13

18 Apr '13

>> Also, BTW, I get the below test failure when I have matplotlib but no PIL installed. I think test_imread_url or the test right before it must be making some incorrect assumptions. (If I re-run the tests after this failure, then also all of the tifffile tests fail, so it seems that some test isn't doing proper teardown of the image loading plugins?) > > That's quite possible--can you see how to fix that test as well? No idea -- that's not setup or teardown code from freeimage... I assume I've noticed this only because everyone else has matplotlib + PIL installed, which I do not. I don't have time to go deeper into the plumbing right now, I'm afraid.

1 0