Hi,
I am looking for a method to locate frequency maxima in the frequency domain image yielded by FFT. The procedure should be reasonable fast since large image stacks have to be processed. Regards Christian Kreutzfeldt |
Hi Christian,
do you only want to find the location and intensity of the "1st order" peak(s) (not counting the known zeroest order) or also want to get a (probably sorted by intensity) list of all peaks? In the former case, I have a plugin that uses this (it´s not a plugin on it´s own) but it could be extended to a more general plugin and also to the latter multi-peak case. I do not know if there is already another independent plugin that already does something similar. in my case a circular suppression area (of actually ~gaussian shape) is applied on the zeroest order an the highest remeining peak is located (in my case it makes no difference if it finds the +1 or -1 order because I have real input data). This is use for "fourier based" "spatial fringe demodulation" of interferograms This technique could be IMHO repeated to suppress the 1st order etc., the extent of the suppression zone clearly would need to be optimzed to the problem at hand. You can contact me by PM if you like. Mit freundlichen Grüßen / Best regards Joachim Wesner Phase GmbH <imaging@PHASE-HL .COM> An Gesendet von: [hidden email] ImageJ Interest Kopie Group <[hidden email]. Thema GOV> Fast Fourier Transform: Location of Maxima 30.10.2009 10:49 Bitte antworten an Phase GmbH <imaging@phase-hl .com> Hi, I am looking for a method to locate frequency maxima in the frequency domain image yielded by FFT. The procedure should be reasonable fast since large image stacks have to be processed. Regards Christian Kreutzfeldt ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________ |
On Monday 02 November 2009 12:50:32 Joachim Wesner wrote:
> In the former case, I have a plugin that uses this (it´s not a plugin on > it´s own) but it could be extended to a more general plugin and also to the > latter multi-peak case. [...] Alternatively one can consider the image as a landscape and use greyscale morphology to extract "domes" of a particular height. The morphology plugins in my site include a "Domes_" plugin, which supports only 8 bit greyscale. Not sure if that would be suitable for your problem, though. The paper to read about this approach is: Vincent L. Morphological greyscale reconstruction in Image Analysis: Applications and efficient algorithms. IEEE Trans Image Proc 2(2) 176-201, 1993. I hope it helps. G. |
Free forum by Nabble | Edit this page |