Mailman 3 June 2013 - scikit-image

Re: SciPy sprint for scikit-image
by Marianne Corvellec 24 Jun '13

24 Jun '13

Hello, Yay! I am super excited at SciPy 2013 as well. :D Is providing answers on StackOverFlow (which make use of scikit-image) still {under consideration, considered {a priority, an important thing to do}}? I have some interest for 2d -> 3d stuff; the video stuff sounds great -- I recently contributed (1 line :p) to PiTiVi (http://www.pitivi.org/) so I'm in the mood. ;) Otherwise, I'm all for improving the docs -- you can count on me. This sprint looks promising! :) Can't wait to see some of you in person, Marianne On Sunday, June 23, 2013 12:20:59 AM UTC-4, Stefan van der Walt wrote: > > On Sun, Jun 23, 2013 at 6:04 AM, Tony Yu <tsy...(a)gmail.com <javascript:>> > wrote: > > Sorry, I've been out drinking tonight. There may have been a bit of > > repetition in that last paragraph ;) > > As with most late-night posting, it's often only the second one that > gives you away :) > > Cheers! > Stéfan >

3 2

Inconsistent projection centre for the Radon transform
by jostein.floystad＠gmail.com 22 Jun '13

22 Jun '13

Hi everyone, I recently found out that scikit-image has a bug (and has had for a long time, it seems) in the Radon transform module (radon/iradon, not related to the discreete versions). The problem manifests itself as images being shifted by going through a forward and inverse Radon transform. In fact, the following code would shift an image N pixels both horizontally and vertically: # image = np.array(...) for i in range(N): sinogram = radon(image) image = iradon(sinogram) These issues can also easily lead to poorer reconstruction quality than what the data in the sinogram allows. The purpose of this message is two-fold: 1. To ask for review on my PR addressing this bug: https://github.com/scikit-image/scikit-image/pull/596 . I realize that this PR is probably not the easiest to review; I guess this reflects that it was tricky for me to get right. However, to me this just makes a review seem all the more important. 2. Alert any users that have based results on this code to check them carefully for systematic errors. If the radon_transform module has been used for tomography simulations, it is quite likely that the results are affected. In the case of using the module for reconstructing tomography data obtained otherwise, the results may be affected and they may not; this will depend on the specifics of the situation. Jostein

2 1

Re: SciPy sprint for scikit-image
by Stéfan van der Walt 22 Jun '13

22 Jun '13

On Sun, Jun 23, 2013 at 6:04 AM, Tony Yu <tsyu80(a)gmail.com> wrote: > Sorry, I've been out drinking tonight. There may have been a bit of > repetition in that last paragraph ;) As with most late-night posting, it's often only the second one that gives you away :) Cheers! Stéfan

1 0

Re: SciPy sprint for scikit-image
by Stéfan van der Walt 22 Jun '13

22 Jun '13

Hey, Tony On Sun, Jun 23, 2013 at 6:01 AM, Tony Yu <tsyu80(a)gmail.com> wrote: > So it looks like there will be a pretty good collection of scikit-image > contributors at SciPy 2013. (I'm looking at you Stéfan van der Walt, Josh > Warner, Marianne Corvellec, James Bergstra... others?) And any lurkers on > the list of whatever level should join in for sure. Very exciting! We've got such a fantastic team and great momentum going at the moment, and I can't wait to working with all of you in person again. > We should have a pretty good sprint this year. Now might be a good time to > propose some ideas and see what we can knock out in a day or two. Let's keep track of the sprint ideas on the wiki page: https://github.com/scikit-image/scikit-image/wiki/SciPy2013-Sprint > I think it'd be great to expand the docs and maybe try to improve the docs > (with start-to-finish type examples). Also, so improvements to the IO > infrastructure and video IO. All good ideas! The video I/O especially needs a lot of love. The CV backend works quite well, but the GStreamer backend is fundamentally broken. Also, it would be nice to figure out how to do real-time video display in the viewer module. I'm in the middle of a 30-hour itinerary, so I'll be out of touch for the next day or two. Stéfan

1 0

Re: SciPy sprint for scikit-image
by Tony Yu 22 Jun '13

22 Jun '13

On Sat, Jun 22, 2013 at 11:01 PM, Tony Yu <tsyu80(a)gmail.com> wrote: > Hi everyone, > > So it looks like there will be a pretty good collection of scikit-image > contributors at SciPy 2013. (I'm looking at you Stéfan van der Walt, Josh > Warner, Marianne Corvellec, James Bergstra... others?) And any lurkers on > the list of whatever level should join in for sure. > > We should have a pretty good sprint this year. Now might be a good time to > propose some ideas and see what we can knock out in a day or two. > > I think it'd be great to expand the docs and maybe try to improve the docs > (with start-to-finish type examples). Also, so improvements to the IO > infrastructure and video IO. > > Cheers! > -Tony > Sorry, I've been out drinking tonight. There may have been a bit of repetition in that last paragraph ;)

1 0

SciPy sprint for scikit-image
by Tony Yu 22 Jun '13

22 Jun '13

Hi everyone, So it looks like there will be a pretty good collection of scikit-image contributors at SciPy 2013. (I'm looking at you Stéfan van der Walt, Josh Warner, Marianne Corvellec, James Bergstra... others?) And any lurkers on the list of whatever level should join in for sure. We should have a pretty good sprint this year. Now might be a good time to propose some ideas and see what we can knock out in a day or two. I think it'd be great to expand the docs and maybe try to improve the docs (with start-to-finish type examples). Also, so improvements to the IO infrastructure and video IO. Cheers! -Tony

1 0

Re: how to highlight/shade a segment in an image.
by Brickle Macho 20 Jun '13

20 Jun '13

On 21/06/13 7:03 AM, Tony Yu wrote: > > Short version, given a label, a lable-image and a image, how do I > shade/tint the label area. > > > You could try out the label2rgb PR: > > https://github.com/scikit-image/scikit-image/pull/485 Thanks. Look interesting. How do I try/pull/incorporate label2rgb code? Michael. --

2 1

Re: Heap and Eikonal eqn
by Zachary Pincus 19 Jun '13

19 Jun '13

> Hi Stefan, > > Thanks I'll check it out. Though, I will need to store 3 values in the heap. I see that BinaryHeap takes just a float value. I'll try and see if it can take a list perhaps. No, it can't I'm afraid -- the heap stores a flat C-style float* array internally. Your best option is to use the heap cython code as a starting point and either add two more flat arrays of floats, or change it to an array of pointers to some other data structure. If you don't want to hack on cython code just now, you could instead maintain three heaps (one for each value). This is obviously suboptimal, but at least it'll work with only constant-time slowdown so you can use it as a proof of concept. Zach

1 0

Re: Heap and Eikonal eqn
by Chintak Sheth 19 Jun '13

19 Jun '13

Hi Stefan, Thanks I'll check it out. Though, I will need to store 3 values in the heap. I see that BinaryHeap takes just a float value. I'll try and see if it can take a list perhaps. Chintak

1 0

Re: How to get full size image which is transformed by warp function
by Yongda Chen 19 Jun '13

19 Jun '13

Hi Stefan, Thank you very much for your suggestion. I am trying to figure out how to add in a certain amount of translation for different transforms. Thanks Yongda On Tuesday, June 18, 2013 12:24:46 AM UTC-7, Stefan van der Walt wrote: > > On Mon, Jun 17, 2013 at 8:01 PM, Yongda Chen <yongd...(a)gmail.com<javascript:>> > wrote: > > I am trying to do geometric transformation with warp in scikit-image, > but I > > cannot get the full size transformed image which has different image > size to > > original image. Even after setting up output_shape parameter, it still > lost > > part of image which is transformed to negative coordinates. > > I'd represent the rotation as a transformation matrix, and then also > add in a certain amount of translation to account for the negative > coordinates. I believe this is what Johannes also recommended, and > what is coded up in `skimage.transform.rotate`. > > Stéfan >

1 0