Making and applying custom masks to FTs

[bradscopegems-2]

I apologise if this message is a repeat: my previous postings were blocked for a while.
I am trying to select parts of the FT of an image and use only these parts in a resynthesised image. I have found examples of this being done with a circular mask, which allows the resynthesis of an image with , in effect, low- or high- pass filtration.
In biological electron microscopy it is often required to find the peaks in the FT, recognizing them as falling on a lattice, and select only these, which results in a low-noise image of ordered structure. The ideal software for doing this would allow the manual input of rectangular boxes of arbitrary size and position, and would allow the recording of the mask , visibly superimposed on the image. To be really convenient, the drawing of a mask anywhere in the FT would force an identical mask to appear symmetrically on the other side of the centre of symmetry of the FT.
   Does such a facility exist in Image J? This procedure is extremely useful: as well as reducing noise it allows two different superimposed lattices , arising from a multilayered specimen or the front and back of a helical particle like a virus or microtubule to be separated in the filtered image.

[Cammer, Michael]

On the power spectrum image, anywhere you draw pixels of zero will be masked and pixels of 255 will be passed.  Therefore, you can do thresholding to find peaks or manual drawing.  There are examples at
http://rsbweb.nih.gov/ij/docs/menus/process.html#fft
http://rsbweb.nih.gov/ij/docs/examples/FFT/
And  this picture was originally a window screen before filtering out the angled vertical lines:  http://www.flickr.com/photos/mcammer/5394773088/
-Michael C.



-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Brad Amos
Sent: Tuesday, February 01, 2011 7:56 AM
To: [hidden email]
Subject: Making and applying custom masks to FTs

I apologise if this message is a repeat: my previous postings were blocked for a while.
I am trying to select parts of the FT of an image and use only these parts in a resynthesised image. I have found examples of this being done with a circular mask, which allows the resynthesis of an image with , in effect, low- or high- pass filtration.
In biological electron microscopy it is often required to find the peaks in the FT, recognizing them as falling on a lattice, and select only these, which results in a low-noise image of ordered structure. The ideal software for doing this would allow the manual input of rectangular boxes of arbitrary size and position, and would allow the recording of the mask , visibly superimposed on the image. To be really convenient, the drawing of a mask anywhere in the FT would force an identical mask to appear symmetrically on the other side of the centre of symmetry of the FT.
   Does such a facility exist in Image J? This procedure is extremely useful: as well as reducing noise it allows two different superimposed lattices , arising from a multilayered specimen or the front and back of a helical particle like a virus or microtubule to be separated in the filtered image.

------------------------------------------------------------
This email message, including any attachments, is for the sole use of the intended recipient(s) and may contain information that is proprietary, confidential, and exempt from disclosure under applicable law. Any unauthorized review, use, disclosure, or distribution is prohibited. If you have received this email in error please notify the sender by return email and delete the original message. Please note, the recipient should check this email and any attachments for the presence of viruses. The organization accepts no liability for any damage caused by any virus transmitted by this email.
=================================

[Michael Schmid]

Hi,

you can easily filter images in the Fourier domain: in the FFT image,  
paint the areas that you want to keep in white (pixel value=255).  
With the standard white color and as long as you keep the original  
brightness/contrast, the 'delete' button does this for your selection  
(rectangular, circular, or whatever you like). When you transform it  
back, ImageJ makes it symmetrical automatically.
   http://rsb.info.nih.gov/ij/docs/menus/process.html#fft

If the ALT key is down during the inverse transform you get the  
symmetrized mask and you can use it later as a Fourier Custom Filter.

Michael
________________________________________________________________

On 1 Feb 2011, at 13:56, Brad Amos wrote:

> I apologise if this message is a repeat: my previous postings were  
> blocked for a while.
> I am trying to select parts of the FT of an image and use only  
> these parts in a resynthesised image. I have found examples of this  
> being done with a circular mask, which allows the resynthesis of an  
> image with , in effect, low- or high- pass filtration.
> In biological electron microscopy it is often required to find the  
> peaks in the FT, recognizing them as falling on a lattice, and  
> select only these, which results in a low-noise image of ordered  
> structure. The ideal software for doing this would allow the manual  
> input of rectangular boxes of arbitrary size and position, and  
> would allow the recording of the mask , visibly superimposed on the  
> image. To be really convenient, the drawing of a mask anywhere in  
> the FT would force an identical mask to appear symmetrically on the  
> other side of the centre of symmetry of the FT.
>    Does such a facility exist in Image J? This procedure is  
> extremely useful: as well as reducing noise it allows two different  
> superimposed lattices , arising from a multilayered specimen or the  
> front and back of a helical particle like a virus or microtubule to  
> be separated in the filtered image.