Path: news.mathworks.com!not-for-mail
From: "Ahmed Zaki" <azaki07@gmail.com>
Newsgroups: comp.soft-sys.matlab
Subject: Re: Modulation Transfer function
Date: Fri, 16 May 2008 02:49:01 +0000 (UTC)
Organization: The MathWorks, Inc.
Lines: 41
Message-ID: <g0ismt$t80$1@fred.mathworks.com>
References: <g00eg4$3lp$1@fred.mathworks.com> <3b7b0902-e3f0-46ba-9e77-45190e6be339@b1g2000hsg.googlegroups.com>
Reply-To: "Ahmed Zaki" <azaki07@gmail.com>
NNTP-Posting-Host: webapp-03-blr.mathworks.com
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 8bit
X-Trace: fred.mathworks.com 1210906141 29952 172.30.248.38 (16 May 2008 02:49:01 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Fri, 16 May 2008 02:49:01 +0000 (UTC)
X-Newsreader: MATLAB Central Newsreader 1380892
Xref: news.mathworks.com comp.soft-sys.matlab:468752


ImageAnalyst <imageanalyst@mailinator.com> wrote in message 
<3b7b0902-e3f0-46ba-9e77-
45190e6be339@b1g2000hsg.googlegroups.com>...
> On May 8, 10:56=A0pm, "Ahmed Zaki" <azak...@gmail.com> 
wrote:
> > How can I calculate the Modulation Transfer function 
for an
> > image? Thanks.. AZ
> 
> ----------------------------------------------------------
--------------
> AZ:
> You either need the point spread function (and then you 
just take the
> magnitude of the Fourier tranform of the PSF), or you 
need an image
> with a bunch of cycles (e.g. sine waves or black and 
white line pairs
> or different frequencies) that you can measure the 
contrast of and
> build up the MTF frequency by frequency.  So which do you 
have?  If
> you have neither, you may be out of luck.  You might be 
able to
> estimate it if you have a known, perfectly sharp edge in 
the image
> (that gets blurred as it gets imaged).  I think you can 
estimate the
> PSF if you have a known point or known edge in your 
image.  What kind
> of image do you have?
> Regards,
> ImageAnalyst


Hi:

If I have an edge and it is blurred in the image, then how 
do I estimate the MTF from that image only?
Thanks,
AZ