Mailman 3 April 2013 - scikit-image

Re: multi-page TIFF writer
by Hadrien Mary 12 Apr '13

12 Apr '13

Hi Juan, You should have a look at the excellent tiffile.py: http://www.lfd.uci.edu/~gohlke/code/tifffile.py.html -- Hadrien Mary Ph.D student in Biology Tournier-Gachet Team CNRS - LBCMCP - UMR 5088 Université de Toulouse - Bât. 4R3B1 118, route de Narbonne - 31062 Toulouse On Fri, Apr 12, 2013 at 2:06 PM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote: > Hi Juan > > On Fri, Apr 12, 2013 at 1:46 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> > wrote: > > Do any of you know about writing a numpy array to a multi-page TIFF file > in > > Python? Last time I checked, you could only read those with PIL, not > write. > > A quick Googling led me to Tiffany, but it doesn't seem to be > particularly > > numpy-friendly... > > The freeimage plugin has code for this already, although we still need > to implement an "imsave_collection" wrapper. Grep for > "write_multipage". > > Regards > 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 Stéfan van der Walt 12 Apr '13

12 Apr '13

Hi Juan On Fri, Apr 12, 2013 at 1:46 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> wrote: > Do any of you know about writing a numpy array to a multi-page TIFF file in > Python? Last time I checked, you could only read those with PIL, not write. > A quick Googling led me to Tiffany, but it doesn't seem to be particularly > numpy-friendly... The freeimage plugin has code for this already, although we still need to implement an "imsave_collection" wrapper. Grep for "write_multipage". Regards Stéfan

1 0

Re: Coordinate overlap in two labelled images
by Michael Aye 12 Apr '13

12 Apr '13

Great stuff, Juan, clear and concise. Thanks very much, it's very instructive! On Tuesday, April 2, 2013 9:33:11 PM UTC-7, Juan Nunez-Iglesias wrote: > > Suggestion number 3. I'll deal with each of three interpretations of your > question. Suppose your images are im1 and im2. Now, you might want 1) of > the regions in im1, the ones that intersect over at least 1 pixel with > regions of im2; 2) the regions that are part of regions in both im1 and im2 > (ie the intersection of regions in im1 and im2); or 3) the union of regions > of im1 and im2 that have at least one pixel intersection with the other > image. > > The last example assumes the union'ed regions don't touch each other. > > Then: > > In [1]: import scipy.ndimage as nd > > In [2]: im1 = np.array([[0, 0, 0, 0], [1, 1, 0, 2], [0, 0, 0, 2]]) > > In [3]: im1 > Out[3]: > array([[0, 0, 0, 0], > [1, 1, 0, 2], > [0, 0, 0, 2]]) > > In [4]: im2 = np.array([[0, 1, 0, 0], [0, 1, 0, 0], [0, 1, 0, 0]]) > > In [5]: im2 > Out[5]: > array([[0, 1, 0, 0], > [0, 1, 0, 0], > [0, 1, 0, 0]]) > > In [6]: intersect = im1.astype(bool) * im2.astype(bool) > > In [7]: intersect > Out[7]: > array([[False, False, False, False], > [False, True, False, False], > [False, False, False, False]], dtype=bool) > > In [8]: overlap_labels1 = np.unique(im1[intersect]) > > In [9]: selector = np.zeros(im1.max() + 1, int) > > In [10]: selector[overlap_labels1] = arange(1, len(overlap_labels1) + 1) > > In [11]: reg1 = selector[im1] > > In [12]: reg1 > Out[12]: > array([[0, 0, 0, 0], > [1, 1, 0, 0], > [0, 0, 0, 0]]) > > In [13]: reg2 = nd.label(intersect)[0] > > In [14]: reg2 > Out[14]: > array([[0, 0, 0, 0], > [0, 1, 0, 0], > [0, 0, 0, 0]], dtype=int32) > > In [15]: overlap_labels2 = np.unique(im2[intersect]) > > In [16]: selector2 = np.zeros(im2.max() + 1, int) > > In [17]: selector2[overlap_labels2] = 1 > > In [18]: nonzero_pixels = reg1 + selector2[im2] > > In [19]: reg3 = nd.label(nonzero_pixels)[0] > > In [20]: reg3 > Out[20]: > array([[0, 1, 0, 0], > [1, 1, 0, 0], > [0, 1, 0, 0]], dtype=int32) > > > > > > On Wed, Apr 3, 2013 at 5:52 AM, Stéfan van der Walt <ste...(a)sun.ac.za<javascript:> > > wrote: > >> Hi Stuart >> >> On Tue, Apr 2, 2013 at 7:04 PM, Stuart Mumford <little...(a)gmail.com<javascript:>> >> wrote: >> > I am using your (amazingly fast) label image routines and regionprops >> to extract the coordinates of the regions in two images. I need to filter >> these objects out if they do not overlap in both the images, then renumber >> all the remaining regions. At the moment I am doing something unspeakable >> with nested loops and sets which is pretty damn slow; I was hoping someone >> here would have a really clever idea on how to solve this. Failing that >> does anyone know how I can find the number of coordinates (output from >> regionprops) that overlap in two numpy arrays? >> >> You can use join_segmentations to find an intersection of your two >> label maps. Then, walk through the newly formed regions and check at >> each point whether the object exists in both images (that way, you >> don't have so many comparisons to do). >> >> >> http://scikit-image.org/docs/dev/auto_examples/plot_join_segmentations.html >> >> If that's not exactly what you had in mind, you can use this code to >> find all the connected regions in an image: >> >> https://github.com/stefanv/lulu/blob/master/lulu/base.pyx#L21 >> >> These regions are encoded sparsely, and testing their intersection is >> very fast (see the algorithm section of this paper: >> https://scholar.sun.ac.za/handle/10019.1/4883). >> >> I hope that helps! >> 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. >> >> >> >

2 1

multi-page TIFF writer
by Juan Nunez-Iglesias 12 Apr '13

12 Apr '13

Hi everyone, Do any of you know about *writing* a numpy array to a multi-page TIFF file in Python? Last time I checked, you could only read those with PIL, not write. A quick Googling led me to Tiffany<https://bitbucket.org/pydica/tiffany>, but it doesn't seem to be particularly numpy-friendly... Thanks! Juan.

1 0

Re: GSOC 2013 Proposal
by siddharth kherada 12 Apr '13

12 Apr '13

Thanks for the feedback. a) I want to point out that the domain transform filter is better than bilateral filter and can be used in various applications in which bilateral filter is used. b) Image denoising using non-local means method. Is this already implemented? http://inf.ufrgs.br/~eslgastal/AdaptiveManifolds/Gastal_Oliveira_SIGGRAPH20… c) Some Image Debluring techniques: 1) Deblur image using blind deconvolution 2) Deblur image using Lucy-Richardson method 3) Deblur image using regularized filter 4) Deblur image using Wiener filter What is your opinion about the above techniques? Best Regards, Siddharth On Tuesday, 9 April 2013 17:01:15 UTC+5:30, Stefan van der Walt wrote: > > Hi Siddharth > > On Tue, Apr 9, 2013 at 11:32 AM, siddharth kherada > <siddharth...(a)gmail.com <javascript:>> wrote: > > I am currently pursuing MS by Research in the field of image processing > at > > IIIT- Hyderabad, India. I am interested in working for Scikit-image this > > summer in a GSoC project. > > Thanks for getting in touch. Both these algorithms are cute, and > we'll consider PRs for them if they come in, but I don't think they > are major underlying parts of the library we'd want to spend a GSoC > on. If you are interested in this sort of thing, image deconvolution > may be more relevant. For GSoC, we'd like to implement algorithms > that will be useful across a wide array of disciplines, if possible. > > Also, we require that students who would like to participate in GSoC > first make other contributions to the project, so be sure to look at > the list of tickets and try to fix some of them, at a minimum. > > Regards > Stéfan >

2 3

Re: Annoying gstreamer bug
by Stéfan van der Walt 11 Apr '13

11 Apr '13

Hi Juan On Thu, Apr 11, 2013 at 10:28 AM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> wrote: > Is no one concerned that anyone writing a command-line tool that leverages > skimage.io will have their help menu overridden? (If they have GStreamer > installed.) This is a bug--I have it fixed in one of my branches, but looks like I forgot to make a PR. I'll see if I can track it down. I think what stopped me is that I realized that the gstreamer video reading is somewhat broken, and tried to get it up and running properly (which I never did). Stéfan

1 0

Re: Annoying gstreamer bug
by Juan Nunez-Iglesias 11 Apr '13

11 Apr '13

Is no one concerned that anyone writing a command-line tool that leverages skimage.io will have their help menu overridden? (If they have GStreamer installed.) On Wed, Apr 10, 2013 at 12:23 AM, Juan Nunez-Iglesias <jni.soma(a)gmail.com>wrote: > Hi All, > > Consider the following code, saved in a file called test.py: > > import gst > > I will bet that few among you would predict this output: > > $ python test.py -h > Usage: > test.py [OPTION...] - GStreamer initialization > > Help Options: > -h, --help Show help options > --help-all Show all help options > --help-gst Show GStreamer Options > > > It seems that someone forgot an > > if __name__ == '__main__' > guard somewhere! > > Now, this would be none of our concern except for the following corollary: > > $ python -c 'from skimage.io.collection import alphanumeric_key' -h > Usage: > -c [OPTION...] - GStreamer initialization > > Help Options: > -h, --help Show help options > --help-all Show all help options > --help-gst Show GStreamer Options > > > Now, I don't know how to fix this, because it make sense that `io` would > import `video` (although `from .video import *` is sloppy). But any > suggestions are very very welcome. > > Juan. >

1 0

Re: Histogram backprojection
by Tony Yu 11 Apr '13

11 Apr '13

On Wed, Apr 10, 2013 at 9:12 PM, abid rahman <abidrahman2(a)gmail.com> wrote: > > But it makes another question : If scipy.ndimage is there and scikit is > using scipy, what is the need of scikit-image ? scipy.ndimage itself is a > good library. > Hi Abid, You should think of `scikit-image` as a complement to `scipy.ndimage`: `scikit-image` adds additional capabilities to what is provided by `scipy.ndimage`. In fact, the maintainers of `scipy.ndimage`<https://github.com/scipy/scipy/blob/master/doc/MAINTAINERS.rst.txt> are Stefan van der Walt, the creator of `scikit-image`, and Thouis (Ray) Jones, a frequent contributor. Part of the rationale for the development of scikits (not just `scikit-image`) was a faster, independent development cycle. This has allowed the SciPy Stack (I think that's the tentative name for the entire scientific python ecosystem, not just `scipy` the package) to grow much faster than before. Cheers, -Tony > Regards > > > On Thu, Apr 11, 2013 at 6:08 AM, Josh Warner <silvertrumpet999(a)gmail.com>wrote: > >> I believe we generally leverage the convolution tools in `scipy.ndimage`. >> On a high level, the scikit-image project is an extension toolkit for SciPy >> so we generally* don't reinvent the wheel. >> >> * there are always exceptions. >> >> >> On Wednesday, April 10, 2013 6:13:52 AM UTC-5, abid rahman wrote: >>> >>> Doesn't scikit-image has a 2d convolution function like cv2.filter2D in >>> OpenCV which enables convolution of an image (or any numpy array) with >>> custom kernel ? >>> >>> On Tuesday, 9 April 2013 22:10:26 UTC+5:30, abid rahman wrote: >>>> >>>> Hi, >>>> >>>> I would like to add a feature on histogram backprojection as proposed >>>> by *Swain & Michael J.* in their paper *"Indexing via color histograms" >>>> *.. Earlier, I had written a simple article on that ( it was focussed >>>> for OpenCV) in my blog : http://opencvpython.blogspot.** >>>> in/2013/03/histograms-4-back-**projection.html.<http://opencvpython.blogspot.in/2013/03/histograms-4-back-projection.html.> >>>> >>>> I can have some simple modification to it to integrate it into >>>> scikit-image. So I would like to know the opinion of senior developers on >>>> this proposal. >>>> >>>> Regards >>>> Abid K. >>>> >>> -- >> 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. >> >> >> > > > > -- > Abid K. > opencvpython.blogspot.com > > -- > 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: Histogram backprojection
by abid rahman 11 Apr '13

11 Apr '13

Hi, I ran into a problem with scikit installation in windows 7. I am using python 2.7 32 bit. python setup.py install shows error : Unable to find vcvarsall.bat. Any idea what it is? I googled and tried a lot of hacks, but still the same error. Any idea what it is? Meanwhile, I will switch into Fedora for backprojection work !!! Regards On Tuesday, 9 April 2013 22:26:52 UTC+5:30, Stefan van der Walt wrote: > > On Tue, Apr 9, 2013 at 6:51 PM, abid rahman <abidr...(a)gmail.com<javascript:>> > wrote: > >> > > http://opencvpython.blogspot.in/2013/03/histograms-4-back-projection.html. > > > I got hold of it now, thanks. That looks very interesting, as does > > http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.132.5942 > > We look forward to your pull request! Let us know if you need any help. > > Regards > Stéfan >

2 1

Montréal scikit-image overview (in French)
by Stéfan van der Walt 10 Apr '13

10 Apr '13

Hi all Marianne Corvellec gave a very nice overview of scikit-image at the Montréal Python user group: http://youtu.be/fqCu6kkPNxU Have a look! Stéfan

1 0