Weiner Filter

[ashamim]

Hello Every one

Do we have any plugin for Weiner Filter in ImageJ so far. I really appreciate your help.

Regards

Ahsan

[Mario Faretta]

Dear Ashan,
accordingly to a previous discussion in the list I suggest you to have a
look at the "FFTJ and DeconvolutionJ" plugin by Nick Linnenbrügger.
A Kalman filter plugin also exists.
Hope it helps
Mario

Il 8/18/2011 12:59 PM, ashamim ha scritto:

> Hello Every one
>
> Do we have any plugin for Weiner Filter in ImageJ so far. I really
> appreciate your help.
>
> Regards
>
> Ahsan
>
> -----
> Ahsan
> --
> View this message in context: http://imagej.588099.n2.nabble.com/Weiner-Filter-tp6698937p6698937.html
> Sent from the ImageJ mailing list archive at Nabble.com.

--
---PLEASE Note the change in telephone number---

--
Mario Faretta
Department of Experimental Oncology
European Institute of Oncology
c/o IFOM-IEO Campus for Oncogenomics
via Adamello 16 20139 Milan Italy
Phone: ++39-0294375027
email: [hidden email]
http://www.ifom-ieo-campus.it

[Arne Seitz]

..just for clarification: I guess you are looking for a Wiener and not a Weiner filter.

Cheers Arne

> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
> ashamim
> Sent: jeudi 18 août 2011 12:59
> To: [hidden email]
> Subject: Weiner Filter
>
> Hello Every one
>
> Do we have any plugin for Weiner Filter in ImageJ so far. I really
> appreciate your help.
>
> Regards
>
> Ahsan
>
> -----
> Ahsan
> --
> View this message in context:
> http://imagej.588099.n2.nabble.com/Weiner-Filter-tp6698937p6698937.html
> Sent from the ImageJ mailing list archive at Nabble.com.

[ashamim]

Thanks Mario ... That is one useful plugin for me.

Regards

[ashamim]

@Ame Thanks for the correction. Yup its Wiener Filter.

[ashamim]

Sorry Arne not Ame. My bad again  

[Herbie-6]

Ahsan,

as I've writen on this list about a year ago (you may search The list
archive), there is no such things as _the_ Wiener filter
BTW, it is named after Norbert Wiener, a mathematician at the MIT.

The reason is that the (linear) Wiener optimum filter must be
constructed individually according to the statistics of the signal
(image) or, even better, from signal (image) regions.

Before you can construct a Wiener filter you need to know the
S/N-ratio, i.e. the power spectra of signal and noise. In general
however, you can only estimate the S/N-ratio because what you can
measure in your images is at best (S+N) and N.

A discussion of this problem and a (in certain cases sufficient)
pragmatic zero-order approximation based on the (S+N)/N-ratio is
given in a paper (1980) that can be downloaded from:
<http://www.gluender.de/Writings/WritingsTexts/HardText.html#Gl-1980-1>

Furthermore, be aware that Wiener filtering is limited to cases of
additive noise, i.e. it isn't advised to apply this technique to
signals disturbed by photon noise and the like.

>Hello Every one
>
>Do we have any plugin for Weiner Filter in ImageJ so far. I really
>appreciate your help.
>
>Regards
>
>Ahsan
>
>-----
>Ahsan
>--
>View this message in context:
>http://imagej.588099.n2.nabble.com/Weiner-Filter-tp6698937p6698937.html
>Sent from the ImageJ mailing list archive at Nabble.com.

HTH
--

                   Herbie

          ------------------------
          <http://www.gluender.de>

[ashamim]

Gluender-5 Thanks for the reply. I really appreciate that. I did realize this thing. What would you recommend me as a better option. Do you think LMS would be a good choice. In my case I have some small surface corrugations dominated by the noise in the scanner system which creates observable steps between subsequent scan lines. Also  temporary contamination of the probe sometimes causes the scan lines to be leveled differently. I actually want to level the individual scan lines to more probable levels.
I don't think Wiener performs any such functionality.  Someone suggested that The LMS Fit method subtracts a fitted polynomial function from each individual scan line with at polynomial degree defined by the Degree numerical text field. But I couldn't figure out how will I implement that.

Regards
Ahsan

[Herbie-6]

Ahsan,

as you state, the image defects in question appear quite different
from what can be effectively treated by use of a Wiener optimum
filter.

Your problem first requires a thorough signal theoretical analysis or
even better, improved image acquisition techniques. Post hoc
corrections never can cope with correctly acquired signals...

>Gluender-5 Thanks for the reply. I really appreciate that. I did realize this
>thing. What would you recommend me as a better option. Do you think LMS
>would be a good choice. In my case I have some small surface corrugations
>dominated by the noise in the scanner system which creates observable steps
>between subsequent scan lines. Also  temporary contamination of the probe
>sometimes causes the scan lines to be leveled differently. I actually want
>to level the individual scan lines to more probable levels.
>I don't think Wiener performs any such functionality.  Someone suggested
>that The LMS Fit method subtracts a fitted polynomial function from each
>individual scan line with at polynomial degree defined by the Degree
>numerical text field. But I couldn't figure out how will I implement that.
>
>Regards
>Ahsan

Good luck

                   Herbie

          ------------------------
          <http://www.gluender.de>