Mario,

   If you still have the use of the imaging device, record dark and light 
images with a uniform black target (or with all radiation blocked) and a 
uniform bright target. Then do a 2-point nonuniformity correction (NUC) using 
ImageJ 'Process/Image Calculator'. The targets can be out of focus; the 
measured black image needs to be fully on the gray scale, including its noise 
fluctuations; the white image, including its noise, should not saturate the 
digital gray scale.

   Otherwise, the 'unsharp masking' method described below seems good.

Jeff Stokes
Solid State Physicist
San Diego

-----Original Message-----
From: Michael Schmid <[email protected]> 
Sent: Friday, June 27, 2025 5:47 AM
To: [email protected]
Subject: Re: any suggestion for CIDRE plugin

Hi Mario,

of course, the gold standard is flat field (and dark frame) correction.
If you can't do that because your images were acquired long time ago or at a 
system where you have no access any more:

The very simple way to mitigate smooth variations of illumination in 
bright-field images with mostly background is 'Subtract background', best using 
the sliding paraboloid option.

If you have a linear detector and zero intensity roughly corresponds to a zero 
pixel value, duplicate the image, use 'subtract background' with 'create 
background' (and 'sliding paraboloid'), and divide the original by this 
background (32-bit float output).

If you have many images taken with the same setup, you could have them in a 
stack (import image sequence) and use Image>Stacks>z Project with 'max 
intensity'. If required, use the 'create background' option of 'subtract 
background' to remove remaining 'holes' (where all images of the stacks have 
had foreground objects). Some blurring or a polynomial fit [1] may help to 
smooth it.

Michael

[1] https://imagejdocu.list.lu/plugin/filter/fit_polynomial/start
________________________________________________________________
On 27.06.25 08:33, Faretta Mario Romolo wrote:
> Hi all,
> I would like testing two different plugins for correction of uneven 
> illumination, BASIC and CIDRE. However, while I have no problem on BASIC, I'm 
> not able to run CIDRE (win 11 imageJ v1.54 java 1.8). When I try ro build a 
> model after a couple of messages that announce successful reading of the 
> input, nothing happens in the destination directory where results have to be 
> written.
> Any experience on it?
> Do you any alternative for the same goal , i.e. retrospective flatfield 
> correction?
> Many thanks
> Mario
> 
> Mario Faretta
> Imaging Development Unit Director
> Department of Experimental Oncology
> European Institute of Oncology
> via Adamello 16
> 20139 Milan
> Italy
> Phone: +390294375027
> email: [email protected]
> [5x1000]<http://www.ieo.it/it/SCIENCE-IN-SOCIETY/Le-nostre-iniziative/
> 5-per-mille/>
> 
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

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

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

Reply via email to