Mailman 3 February 2014 - scikit-image

Re: Contributing in Image Compression Algorithms
by Stéfan van der Walt 08 Feb '14

08 Feb '14

On Sat, 08 Feb 2014 21:20:36 +1100, Juan Nunez-Iglesias wrote: > Sorry, I'm happy to do it, but I thought we were waiting for at least > another mentor to volunteer? Between you and myself we should be able to handle one student, I think. We'd need another mentor for two, for sure. StÃ©fan

1 0

Re: scikit-image paper
by Stéfan van der Walt 07 Feb '14

07 Feb '14

On Wed, Jan 29, 2014 at 12:43 AM, Juan Nunez-Iglesias <jni.soma(a)gmail.com> wrote: > Aaaaand, on cue: > http://blog.peerj.com/post/74820148558/peerj-partners-with-writelatex PeerJ is currently offering free publication (if you submit a pre-print first) until March 31st: http://blog.peerj.com/post/67476390580/free-publications-in-peerj This does look very appealing. I am sure we can hit that deadline (Johannes is done with his section, mine is nearing completion and we have two examples in the works). Stéfan

1 0

Re: image input
by vicky Liau 06 Feb '14

06 Feb '14

Hi Juan, The min of the image is -186.11397 and the max is 30790.521. That is why I got the error. Thanks for the reply and pointing out this page. I have more understandings on image data types for scikit-image. Best Regards, Vicky On Thursday, February 6, 2014 7:12:36 PM UTC-7, Juan Nunez-Iglesias wrote: > > Hi Vicky, > > What's the min and max of your "myimagearray"? As the error message says, > float images should only contain values between -1 and 1. ie, > myimagearray.max() should be less than 1, and myimagearray.min() should be > greater than -1. > > Here's the relevant part of the documentation describing data types: > http://scikit-image.org/docs/0.9.x/user_guide/data_types.html > > Hope that helps, > > Juan. > > > > On Fri, Feb 7, 2014 at 9:20 AM, vicky Liau <vickyt...(a)hotmail.com<javascript:> > > wrote: > >> Hi, >> I got an error message: Images of type float must be between -1 and 1, >> when I call: edge_sobel = sobel (myimagearray). >> My image dtype is float 32. >> I am not sure whether I am missing something. Should I convert my image >> into 8 bit integrate? >> Thanks! >> Vicky >> >> -- >> 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. >> > >

1 0

Re: image input
by Juan Nunez-Iglesias 06 Feb '14

06 Feb '14

Hi Vicky, What's the min and max of your "myimagearray"? As the error message says, float images should only contain values between -1 and 1. ie, myimagearray.max() should be less than 1, and myimagearray.min() should be greater than -1. Here's the relevant part of the documentation describing data types: http://scikit-image.org/docs/0.9.x/user_guide/data_types.html Hope that helps, Juan. On Fri, Feb 7, 2014 at 9:20 AM, vicky Liau <vickyting0910(a)hotmail.com>wrote: > Hi, > I got an error message: Images of type float must be between -1 and 1, > when I call: edge_sobel = sobel (myimagearray). > My image dtype is float 32. > I am not sure whether I am missing something. Should I convert my image > into 8 bit integrate? > Thanks! > Vicky > > -- > 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

GSoC 2014 (was: Contributing in Image Compression Algorithms)
by Stéfan van der Walt 06 Feb '14

06 Feb '14

On Thu, 06 Feb 2014 11:48:15 +1100, Juan Nunez-Iglesias wrote: > I'd be interested in mentoring, actually. But I only have vague ideas about > what contributions we're looking for. As Josh said, compression itself > would be more interesting to the scikit as a means to other ends, such as > denoising. Step 1: Apply as a mentoring organization before 14 February (I can help with the documentation) Step 2: If that fails, register under the umbrella of the PSF Step 3: Solicit proposals from students (at that time, students should already have become involved in the community and have made pull requests under discussion) Step 4: Choose one, maximum two excellent contributions. Step 5: Spend about 5 hours a week for two months mentoring students. The projects need to be large enough in scope to cover 2 months (a good example would be Ankit's feature detectors) and make a significant and needed contribution to the project. An example of such a project for this year would be to implement a registration submodule. The reason I am hesitant to take on more than one candidate is that we are still busy reviewing and refactoring Chintak and Marc's contributions from last year (inpainting and graph cuts). Regards StÃ©fan

1 0

Re: scikit-image paper
by Juan Nunez-Iglesias 06 Feb '14

06 Feb '14

On Thu, Feb 6, 2014 at 4:05 PM, Tony Yu <tsyu80(a)gmail.com> wrote: > (there's a reason I left academia!) > And here I thought it was greed! =P I also think an author on a paper should contribute to the writing > The criterion for inclusion that is usually brandished about, and that I tend to agree with, is "*significant intellectual contribution*", where "significant" allows for some wiggle room. Even though that wiggle room is sometimes abused in academia, as you bitterly recall, I don't think there is any question about whether your contribution was significant.

1 0

Re: Contributing in Image Compression Algorithms
by Jenil Shah 05 Feb '14

05 Feb '14

Hi all, How about implementing Wavelet based techniques? Like Image Processing using Wavelet Transform. Decompose the image into wavelet domain Alter the wavelet coefficients as per the application such as denoising, compression, edge enhancement, etc. Reconstruct the image with the altered wavelet coefficients. I am looking forward to contribute.Will this help scikit-image? Suggestions needed. On Thursday, February 6, 2014 6:18:15 AM UTC+5:30, Juan Nunez-Iglesias wrote: > > I'd be interested in mentoring, actually. But I only have vague ideas > about what contributions we're looking for. As Josh said, compression > itself would be more interesting to the scikit as a means to other ends, > such as denoising. > > Fixing up the watershed would be useful! ;) > > > > On Thu, Feb 6, 2014 at 5:39 AM, MIT SHAH <mitk...(a)gmail.com <javascript:>>wrote: > >> Mentoring Organizations can submit applications from 3rd feb to 14th feb. >> Student Application period is from 10th to 21st march. >> And yes, I am very much interested in working with scikit-image this >> summer. Actually, I came to know about scikit image from my senior Chintak >> Sheth, who worked with you last time. Since last 8-9 months I am learning >> and working on Image Processing. So, it will be my pleasure if I get a >> chance to work with scikit-image & contribute my best. >> >> >> On Sunday, February 2, 2014 2:32:30 PM UTC+5:30, MIT SHAH wrote: >> >>> Hi, >>> I am working on Image Compression Algorithms. I want to contribute it in >>> scikit-image. >>> Can I contribute it ? >>> I am asking this, because I didn't find Image Compression as Requested >>> features on the following link : https://github.com/scikit- >>> image/scikit-image/wiki/Requested-features >>> Please guide me. >>> >> -- >> 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. >> > >

1 0

Re: Contributing in Image Compression Algorithms
by Kakul Chandra 05 Feb '14

05 Feb '14

Hello. I am undergraduate working on Newspaper Text Extraction. I would like to contribute to scikit-image. Is scikit-image applying for mentoring organization in GSoC14? On Thursday, February 6, 2014 6:18:15 AM UTC+5:30, Juan Nunez-Iglesias wrote: > > I'd be interested in mentoring, actually. But I only have vague ideas > about what contributions we're looking for. As Josh said, compression > itself would be more interesting to the scikit as a means to other ends, > such as denoising. > > Fixing up the watershed would be useful! ;) > > > > On Thu, Feb 6, 2014 at 5:39 AM, MIT SHAH <mitk...(a)gmail.com <javascript:>>wrote: > >> Mentoring Organizations can submit applications from 3rd feb to 14th feb. >> Student Application period is from 10th to 21st march. >> And yes, I am very much interested in working with scikit-image this >> summer. Actually, I came to know about scikit image from my senior Chintak >> Sheth, who worked with you last time. Since last 8-9 months I am learning >> and working on Image Processing. So, it will be my pleasure if I get a >> chance to work with scikit-image & contribute my best. >> >> >> On Sunday, February 2, 2014 2:32:30 PM UTC+5:30, MIT SHAH wrote: >> >>> Hi, >>> I am working on Image Compression Algorithms. I want to contribute it in >>> scikit-image. >>> Can I contribute it ? >>> I am asking this, because I didn't find Image Compression as Requested >>> features on the following link : https://github.com/scikit- >>> image/scikit-image/wiki/Requested-features >>> Please guide me. >>> >> -- >> 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. >> > >

1 0

Re: scikit-image paper
by Tony Yu 05 Feb '14

05 Feb '14

On Wed, Feb 5, 2014 at 10:49 PM, Juan Nunez-Iglesias <jni.soma(a)gmail.com>wrote: > I'm not positive about this new criterion. My gut feeling is that it > should be a bit nuanced on either side. For example, even if Tony doesn't > write a single line for the paper, I still think he should be an author. > Conversely, I don't think someone who has not thus far contributed a single > line of code, issue, or tutorial should be able to jump in, write a few > paragraphs, and be an author on the paper. > As much as I love scikit-image, I hate writing academic papers (there's a reason I left academia!). I would feel a bit guilty having my name on the paper without contributing any text to it; I'm not sure yet if that guilt is strong enough to make me contribute to the writing (again, because I hate writing papers---I still love scikit-image). In short, I would be fine if I were left off the authors list (assuming I haven't guilted myself into writing a section of the paper)---it would great to be listed, but I also think an author on a paper should contribute to the writing (... which, sadly, isn't always the case in academia... actually you don't even have to contribute that much to the research either... sorry I may still have some post-academia hang ups:P). Best, -Tony P.S. Sorry I haven't contributed much to the paper discussion. > Having said that, I do agree that "core contributors" is too nebulous, so, > for example, I'd be perfectly comfortable having François on the author > list. > > Juan. > > > > On Thu, Feb 6, 2014 at 2:27 AM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote: > >> On Wed, Feb 5, 2014 at 4:25 PM, Stéfan van der Walt <stefan(a)sun.ac.za> >> wrote: >> > On Wed, Feb 5, 2014 at 4:01 PM, Johannes Schönberger <jsch(a)demuc.de> >> wrote: >> >> I also thought about contributing one example covering feature >> detection, extraction, matching and some simple image stitching using >> projective transformations... your thoughts? Is this too long? >> >> By the way, you can make really nice panoramas if you export the >> transformed images to TIFF and stitch them together using "enblend". >> Since blending is not really in our scope, it's probably fine to use >> that as the final component, if you want a particularly shiny result. >> >> 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. >

1 0

Re: scikit-image paper
by Juan Nunez-Iglesias 05 Feb '14

05 Feb '14

I'm not positive about this new criterion. My gut feeling is that it should be a bit nuanced on either side. For example, even if Tony doesn't write a single line for the paper, I still think he should be an author. Conversely, I don't think someone who has not thus far contributed a single line of code, issue, or tutorial should be able to jump in, write a few paragraphs, and be an author on the paper. Having said that, I do agree that "core contributors" is too nebulous, so, for example, I'd be perfectly comfortable having François on the author list. Juan. On Thu, Feb 6, 2014 at 2:27 AM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote: > On Wed, Feb 5, 2014 at 4:25 PM, Stéfan van der Walt <stefan(a)sun.ac.za> > wrote: > > On Wed, Feb 5, 2014 at 4:01 PM, Johannes Schönberger <jsch(a)demuc.de> > wrote: > >> I also thought about contributing one example covering feature > detection, extraction, matching and some simple image stitching using > projective transformations... your thoughts? Is this too long? > > By the way, you can make really nice panoramas if you export the > transformed images to TIFF and stitch them together using "enblend". > Since blending is not really in our scope, it's probably fine to use > that as the final component, if you want a particularly shiny result. > > 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