Re: spectrogram...but for images?
- From: aruzinsky <aruzinsky@xxxxxxxxxxxxxxxxxxxx>
- Date: Sun, 13 Jan 2008 09:10:04 -0800 (PST)
On Jan 13, 5:36 am, -G- <-...@xxxxxxxxxxxxxxxxxx> wrote:
ImageAnalyst wrote:
On Jan 8, 8:38 pm, "MZ" <m...@xxxxxxxxxxx> wrote:
I have an image. I want to evaluate the spatial frequency components asSounds confusing. Are you saying that you'd like a Fourier Transform
a function of position in the image.
(which is the spatial frequency spectrum) in an N by N window centered
on each and every pixel in the image? So that if you had a mega pixel
image, you'd have a million Fourier Transforms? Seems like a strange
need. What are you going to do with them? Please explain further.
Perhaps not confusing, but I agree that it would be time consuming.
Nobody else but you said it would be time consuming.
The problem with DFT based 1D spectrograms such as here,
http://en.wikipedia.org/wiki/Spectrogram
is that they only provide useful information when the frequency
changes slowly compared to window length. In typical images, the
frequency changes rapidly over a length of a few pixels. However, the
definition of "time varying frequency" is pretty much determined by
the test method or application. As I recall, in the paper I
previously cited, various spectrogram methods were tested on a
sinusoid with a frequency that was a step function of time.
.
- References:
- spectrogram...but for images?
- From: MZ
- Re: spectrogram...but for images?
- From: ImageAnalyst
- Re: spectrogram...but for images?
- From: -G-
- spectrogram...but for images?
- Prev by Date: De-ringing and de-blocking source code
- Next by Date: Software to make PNG transparent
- Previous by thread: Re: spectrogram...but for images?
- Next by thread: Re: spectrogram...but for images?
- Index(es):
Relevant Pages
|