Mailman 3 January 2014 - scikit-image

Re: scikit-image paper
by Johannes Schönberger 18 Jan '14

18 Jan '14

Ping to all involved people. what's the status in this project?

2 1

3D Template matching
by Stéfan van der Walt 14 Jan '14

14 Jan '14

Hey everyone, Just a heads-up that 3D template matching is now available as a side-effect of https://github.com/scikit-image/scikit-image/pull/847 Thanks, Johannes! Stéfan

1 0

Re: Image filters
by Johannes Schönberger 12 Jan '14

12 Jan '14

+1 for skimage.effects On Jan 8, 2014, at 19:33, Stéfan van der Walt <stefan(a)sun.ac.za> wrote: > There are a lot of interesting algorithms related to "artsy" kind of > things, e.g. colorization, Picasso-fication, etc. Perhaps > ``skimage.effects``? > > -- > 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

incoherent face orientation in marching cubes
by Janto Dreijer 08 Jan '14

08 Jan '14

I am trying to create an STL file from the results of running marching cubes. Attached is the modified version of the marching cubes usage example from the documentation. It seems that skimage does not yield coherent face orientations. See attached image (the gaps are faces in the other direction). Using meshlab's "re-orient all faces coherently" command seems to help a bit... Is it just me or is this a bug in skimage? Regards Janto

3 3

Re: Image filters
by Johannes Schönberger 08 Jan '14

08 Jan '14

Yes, the code in the package is a bit over-complicated :-) It's probably easier to re-implement the rather simple functionality... maybe I'll have some spare time on a flight :-) The question is where this code should go. I don't think it's a good idea to place those "artistic" filters directly into skimage.filter, maybe better skimage.filter.effects... Am 06.01.2014 um 18:48 schrieb Ronnie Ghose <ronnie.ghose(a)gmail.com>: > the code is actually very simple - it shouldn't be too much work. quite a lot of random boilerplate though > > > On Mon, Jan 6, 2014 at 6:47 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> wrote: > You should know as well as anybody that reviewing the PR will be a lot of work! =) Again, if someone else volunteers, I'm happy for it to go ahead. But certainly if we invite the PR, it should be reviewed quickly. We don't want to invite someone and then have the PR languish in review. > > > On Tue, Jan 7, 2014 at 1:14 AM, Johannes Schönberger <jsch(a)demuc.de> wrote: > Yes, I definitely agree on all your points, but we could ask the author of the package if she/he is willing to do it? > > Johannes > > Am 05.01.2014 um 21:47 schrieb Ronnie Ghose <ronnie.ghose(a)gmail.com>: > > > + To Juan's Idea.. they look nice ... but I can't see intellectual applications? Maybe have a reference to them or the like, I don't know if we want to spend time maintaining them necessarily > > > > > > On Sun, Jan 5, 2014 at 9:45 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> wrote: > > I'm ambivalent. They are cool and would be nice for publicity, but let's not forget that it's scikit-image, and these filters have questionable scientific use (at least that I can think of — I've been shown more than once to lack imagination! =). In addition, the object-based implementation wouldn't fit very well with skimage's more functional interface. > > > > I wouldn't be opposed to including them but I also would not want to spend my time on them. > > > > Juan. > > > > > > > > On Mon, Jan 6, 2014 at 12:11 PM, Johannes Schönberger <jsch(a)demuc.de> wrote: > > Hi everyone, > > > > Interesting and those filters could be incorporated into scikit-image in my opinion: > > > > http://jason2506.github.io/imfilter/ > > > > What do others think? > > > > -- > > 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. > > > > > > -- > > 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. > > > -- > 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: Raw & Central Moments
by Johannes Schönberger 07 Jan '14

07 Jan '14

Hi, instead of: > x_bar, y_bar = props.centroid it must be: > y_bar, x_bar = props.centroid Hope that helps? Johannes Am 06.01.2014 um 13:11 schrieb Jaidev Deshpande <deshpande.jaidev(a)gmail.com>: > Hi all, > > According to the wikipedia page on image moments, central moments and raw moments are related as in this link: http://en.wikipedia.org/wiki/Image_moment#Central_moments > > I’ve been trying to examine these relationships using the skimage.measure module. > > I created an ellipse image as per the measure region properties example (http://scikit-image.org/docs/dev/auto_examples/plot_regionprops.html#exampl…) and rotated it. > > Then I ran the following script: > > from skimage.measure import regionprops, moments, moments_central > from skimage.morphology import label > > props = regionprops(image) # where `image` is the rotated ellipse > x_bar, y_bar = props.centroid > > raw_moments = moments(image) > central_moments = moments_central(image, x_bar, y_bar) > > Now according to the wikipedia page, central_moments[2,0] should be equal to: > raw_moments[2,0] - x_bar * raw_moments[1,0] > > But that’s not working out. In fact, the value yielded by the above formula is much larger than what skimage says it should be. > > What am I missing? > > Thanks! > > -- > JD > > -- > 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 filters
by Juan Nunez-Iglesias 06 Jan '14

06 Jan '14

My preference would be to change the interface to be more functional (which fits better in skimage, despite a few OO interfaces currently in the library), so the code could change quite a bit from its present state. The simplicity might remain, but even tiny PRs often require a lot of formatting, new tests, and so forth. On Tue, Jan 7, 2014 at 10:48 AM, Ronnie Ghose <ronnie.ghose(a)gmail.com>wrote: > the code is actually very simple - it shouldn't be too much work. quite a > lot of random boilerplate though > > > On Mon, Jan 6, 2014 at 6:47 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com>wrote: > >> You should know as well as anybody that reviewing the PR will be a lot of >> work! =) Again, if someone else volunteers, I'm happy for it to go ahead. >> But certainly if we invite the PR, it should be reviewed quickly. We don't >> want to invite someone and then have the PR languish in review. >> >> >> On Tue, Jan 7, 2014 at 1:14 AM, Johannes Schönberger <jsch(a)demuc.de>wrote: >> >>> Yes, I definitely agree on all your points, but we could ask the author >>> of the package if she/he is willing to do it? >>> >>> Johannes >>> >>> Am 05.01.2014 um 21:47 schrieb Ronnie Ghose <ronnie.ghose(a)gmail.com>: >>> >>> > + To Juan's Idea.. they look nice ... but I can't see intellectual >>> applications? Maybe have a reference to them or the like, I don't know if >>> we want to spend time maintaining them necessarily >>> > >>> > >>> > On Sun, Jan 5, 2014 at 9:45 PM, Juan Nunez-Iglesias < >>> jni.soma(a)gmail.com> wrote: >>> > I'm ambivalent. They are cool and would be nice for publicity, but >>> let's not forget that it's scikit-image, and these filters have >>> questionable scientific use (at least that I can think of — I've been shown >>> more than once to lack imagination! =). In addition, the object-based >>> implementation wouldn't fit very well with skimage's more functional >>> interface. >>> > >>> > I wouldn't be opposed to including them but I also would not want to >>> spend my time on them. >>> > >>> > Juan. >>> > >>> > >>> > >>> > On Mon, Jan 6, 2014 at 12:11 PM, Johannes Schönberger <jsch(a)demuc.de> >>> wrote: >>> > Hi everyone, >>> > >>> > Interesting and those filters could be incorporated into scikit-image >>> in my opinion: >>> > >>> > http://jason2506.github.io/imfilter/ >>> > >>> > What do others think? >>> > >>> > -- >>> > 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. >>> > >>> > >>> > -- >>> > 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. >>> >> >> -- >> 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. >

1 0

Re: Image filters
by Ronnie Ghose 06 Jan '14

06 Jan '14

the code is actually very simple - it shouldn't be too much work. quite a lot of random boilerplate though On Mon, Jan 6, 2014 at 6:47 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com>wrote: > You should know as well as anybody that reviewing the PR will be a lot of > work! =) Again, if someone else volunteers, I'm happy for it to go ahead. > But certainly if we invite the PR, it should be reviewed quickly. We don't > want to invite someone and then have the PR languish in review. > > > On Tue, Jan 7, 2014 at 1:14 AM, Johannes Schönberger <jsch(a)demuc.de>wrote: > >> Yes, I definitely agree on all your points, but we could ask the author >> of the package if she/he is willing to do it? >> >> Johannes >> >> Am 05.01.2014 um 21:47 schrieb Ronnie Ghose <ronnie.ghose(a)gmail.com>: >> >> > + To Juan's Idea.. they look nice ... but I can't see intellectual >> applications? Maybe have a reference to them or the like, I don't know if >> we want to spend time maintaining them necessarily >> > >> > >> > On Sun, Jan 5, 2014 at 9:45 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> >> wrote: >> > I'm ambivalent. They are cool and would be nice for publicity, but >> let's not forget that it's scikit-image, and these filters have >> questionable scientific use (at least that I can think of — I've been shown >> more than once to lack imagination! =). In addition, the object-based >> implementation wouldn't fit very well with skimage's more functional >> interface. >> > >> > I wouldn't be opposed to including them but I also would not want to >> spend my time on them. >> > >> > Juan. >> > >> > >> > >> > On Mon, Jan 6, 2014 at 12:11 PM, Johannes Schönberger <jsch(a)demuc.de> >> wrote: >> > Hi everyone, >> > >> > Interesting and those filters could be incorporated into scikit-image >> in my opinion: >> > >> > http://jason2506.github.io/imfilter/ >> > >> > What do others think? >> > >> > -- >> > 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. >> > >> > >> > -- >> > 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. >> > > -- > 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: Image filters
by Johannes Schönberger 06 Jan '14

06 Jan '14

Yes, I definitely agree on all your points, but we could ask the author of the package if she/he is willing to do it? Johannes Am 05.01.2014 um 21:47 schrieb Ronnie Ghose <ronnie.ghose(a)gmail.com>: > + To Juan's Idea.. they look nice ... but I can't see intellectual applications? Maybe have a reference to them or the like, I don't know if we want to spend time maintaining them necessarily > > > On Sun, Jan 5, 2014 at 9:45 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> wrote: > I'm ambivalent. They are cool and would be nice for publicity, but let's not forget that it's scikit-image, and these filters have questionable scientific use (at least that I can think of — I've been shown more than once to lack imagination! =). In addition, the object-based implementation wouldn't fit very well with skimage's more functional interface. > > I wouldn't be opposed to including them but I also would not want to spend my time on them. > > Juan. > > > > On Mon, Jan 6, 2014 at 12:11 PM, Johannes Schönberger <jsch(a)demuc.de> wrote: > Hi everyone, > > Interesting and those filters could be incorporated into scikit-image in my opinion: > > http://jason2506.github.io/imfilter/ > > What do others think? > > -- > 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. > > > -- > 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

Raw & Central Moments
by Jaidev Deshpande 06 Jan '14

06 Jan '14

Hi all, According to the wikipedia page on image moments, central moments and raw moments are related as in this link: <http://en.wikipedia.org/wiki/Image_moment#Central_moments> http://en.wikipedia.org/wiki/Image_moment#Central_moments I’ve been trying to examine these relationships using the skimage.measure module. I created an ellipse image as per the measure region properties example (<http://scikit-image.org/docs/dev/auto_examples/plot_regionprops.html#exampl…> http://scikit-image.org/docs/dev/auto_examples/plot_regionprops.html#exampl…) and rotated it. Then I ran the following script: from skimage.measure import regionprops, moments, moments_central from skimage.morphology import label props = regionprops(image) # where `image` is the rotated ellipse x_bar, y_bar = props.centroid raw_moments = moments(image) central_moments = moments_central(image, x_bar, y_bar) Now according to the wikipedia page, central_moments[2,0] should be equal to: raw_moments[2,0] - x_bar * raw_moments[1,0] But that’s not working out. In fact, the value yielded by the above formula is much larger than what skimage says it should be. What am I missing? Thanks! -- JD

1 0