Plugin to remove "light pencils" or banding (sample attached)

[Jacob Keller]

Dear List,

I was wondering whether anyone had written a plugin or knows of any
software to take away the banding pattern or "light pencil" effect in
images similar to the one attached hereto?

Jacob

--
*******************************************
Jacob Pearson Keller, PhD
Postdoctoral Associate
HHMI Janelia Farms Research Campus
email: [hidden email]
*******************************************

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[John Oreopoulos]

Dear Jacob,

Perhaps this feature (which is already built into ImageJ) would be useful? Note the part about suppression of lines/bands along one direction:

http://rsbweb.nih.gov/ij/plugins/fft-filter.html

Sincerely,

John Oreopoulos
Staff Scientist
Spectral Applied Research
Richmond Hill, Ontario
Canada
www.spectral.ca



On 2013-01-16, at 1:55 PM, Jacob Keller wrote:

> Dear List,
>
> I was wondering whether anyone had written a plugin or knows of any
> software to take away the banding pattern or "light pencil" effect in
> images similar to the one attached hereto?
>
> Jacob
>
> --
> *******************************************
> Jacob Pearson Keller, PhD
> Postdoctoral Associate
> HHMI Janelia Farms Research Campus
> email: [hidden email]
> *******************************************
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> <SPC00_TM00001_ANG000_CM0_CHN00_PH0_c0032_selection.tif>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Jacob Keller]

Okay, actually, it seems that there is a great algorithm for just this
purpose known as VSNR from Pierre Weiss' group in Toulouse, but I am
not sure whether it has been implemented in imagej.

JPK

On Wed, Jan 16, 2013 at 2:06 PM, John Oreopoulos
<[hidden email]> wrote:

> Dear Jacob,
>
> Perhaps this feature (which is already built into ImageJ) would be useful? Note the part about suppression of lines/bands along one direction:
>
> http://rsbweb.nih.gov/ij/plugins/fft-filter.html
>
> Sincerely,
>
> John Oreopoulos
> Staff Scientist
> Spectral Applied Research
> Richmond Hill, Ontario
> Canada
> www.spectral.ca
>
>
>
> On 2013-01-16, at 1:55 PM, Jacob Keller wrote:
>
>> Dear List,
>>
>> I was wondering whether anyone had written a plugin or knows of any
>> software to take away the banding pattern or "light pencil" effect in
>> images similar to the one attached hereto?
>>
>> Jacob
>>
>> --
>> *******************************************
>> Jacob Pearson Keller, PhD
>> Postdoctoral Associate
>> HHMI Janelia Farms Research Campus
>> email: [hidden email]
>> *******************************************
>>
>> --
>> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>> <SPC00_TM00001_ANG000_CM0_CHN00_PH0_c0032_selection.tif>
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
*******************************************
Jacob Pearson Keller, PhD
Postdoctoral Associate
HHMI Janelia Farms Research Campus
email: [hidden email]
*******************************************

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[dscho]

Hi Jacob (& Jérôme),

On Wed, 16 Jan 2013, Jacob Keller wrote:

> On Wed, Jan 16, 2013 at 2:06 PM, John Oreopoulos
> <[hidden email]> wrote:
>
> > On 2013-01-16, at 1:55 PM, Jacob Keller wrote:
> >
> >> I was wondering whether anyone had written a plugin or knows of any
> >> software to take away the banding pattern or "light pencil" effect in
> >> images similar to the one attached hereto?
> >
> > Perhaps this feature (which is already built into ImageJ) would be useful? Note the part about suppression of lines/bands along one direction:
> >
> > http://rsbweb.nih.gov/ij/plugins/fft-filter.html
>
> Okay, actually, it seems that there is a great algorithm for just this
> purpose known as VSNR from Pierre Weiss' group in Toulouse,

Thanks for sending me on a chase to find this paper. I found it! Yay! For
everybody else being interested in this topic:

http://www.math.univ-toulouse.fr/~weiss/Publis/IEEEIP_VSNR_Final.pdf

> but I am not sure whether it has been implemented in imagej.

Since the Acknowledgments section of the mentioned paper include this
sentence:

        They thank Beat Münch for providing a Fiji implementation of the
        Wavelet-FFT algorithm and test images.

I assume that it has been implemented in ImageJ.

It is too sad that the implementation has not been made public. In my
view, keeping implementations in scientific articles secret hampers
science greatly by making it hard for other scientists to enhance upon
results, or even simply to provide the peer review that is so critical for
good science.

See also http://developer.imagej.net/why-open-software-vital-science

Jérôme, would you please reconsider sharing the implementation (preferably
with source code, so as to encourage work with and on the algorithm)?

Thank you very much,
Johannes

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Jacob Keller]

>
> Since the Acknowledgments section of the mentioned paper include this
> sentence:
>
>         They thank Beat Münch for providing a Fiji implementation of the
>         Wavelet-FFT algorithm and test images.
>
> I assume that it has been implemented in ImageJ.
>
> It is too sad that the implementation has not been made public. In my
> view, keeping implementations in scientific articles secret hampers
> science greatly by making it hard for other scientists to enhance upon
> results, or even simply to provide the peer review that is so critical for
> good science.
>
> See also http://developer.imagej.net/why-open-software-vital-science
>
> Jérôme, would you please reconsider sharing the implementation (preferably
> with source code, so as to encourage work with and on the algorithm)?
>
> Thank you very much,
> Johannes

Well, in all fairness, they did provide the matlab and c code on their
webpage, so those who have matlab could run the algorithm. Is there any
easy way to get it into imagej though?

Also, I think the acknowledgements section you quoted was talking about an
alternative method to which they compared their algorithm, so probably not
a proof that there was an imagej implementation.

Thanks very much for your help, and also for the encouragement of openness!

Jacob



--
*******************************************
Jacob Pearson Keller, PhD
Postdoctoral Associate
HHMI Janelia Farms Research Campus
email: [hidden email]
*******************************************

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[dscho]

Dear Jérôme,

On Thu, 17 Jan 2013, Jerome FEHRENBACH wrote:

> I hope the implementation you found on Pierre Weiss' webpage is
> sufficient for you. Let us know if you can have it run successfully or
> not.  If anyway you are interested in the present (preliminary) FIJI
> plugin, we would be glad to share it with you and to have your feedback
> as a user.

Please let me reiterate my encouraging you to release even the preliminary
Fiji plugin publicly. You will get more useful feedback by more
scientists, including competent programmers.

Ciao,
Johannes

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html