Getting 32-bit floating point FFT

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Getting 32-bit floating point FFT

Aaron Hendrickson
Hello,

I am trying to analyze and filter noise in images using Fourier Transforms.
 In ImageJ when you create a FFT from an image you get an 8-bit "power
spectrum"....not the full 32-bit floating point FFT.  Is there a way to get
this uncompressed FFT?

Cheers,
Aaron Hendrickson.

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: Getting 32-bit floating point FFT

Michael Schmid
Hi Aaron,

the 8-bit power spectrum has logarithmic scaling; the real FFT data are invisible in the background (for doing 'inverse FFT'). If you want the true power spectrum for analysis, use Process>FFT>FFT Options and select 'raw power spectrum'. You can't do a back transform of the power spectrum, however, it lacks the phase.

Michael
________________________________________________________________
On 2013-Mar-18, at 13:34, Aaron Hendrickson wrote:

> Hello,
>
> I am trying to analyze and filter noise in images using Fourier Transforms.
> In ImageJ when you create a FFT from an image you get an 8-bit "power
> spectrum"....not the full 32-bit floating point FFT.  Is there a way to get
> this uncompressed FFT?
>
> Cheers,
> Aaron Hendrickson.

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