2D Capon Method

Hi,

I am interested in implementing the 2D capon method, similar to what has
been done for the 1D case here:
http://www.dsprelated.com/showmessage/78531/1.php

Any suggestions as what is the best way to modify the above code to make it
work for 2d signals?

I am looking at the following paper, and am interested in applying capon
similarly:
http://www.emo.org.tr/ekler/23d0a5671efd29e_ek.pdf
However, their code seems different from what was posted here on
DSPRelated.

I appreciate your help and comments,
Nargess

On Apr 9, 6:34&#4294967295;am, "nargess" <nargessm@n_o_s_p_a_m.terpalum.umd.edu>
wrote:
> Hi,
>
> I am interested in implementing the 2D capon method, similar to what has

2D capon method. Does that involve a hit in a 2D restaurant somewhere?


Hardy

On 08-04-2010 at 22:15:04 HardySpicer <gyansorova@gmail.com> wrote:

(...)
>> I am interested in implementing the 2D capon method, similar to what has
>
> 2D capon method. Does that involve a hit in a 2D restaurant somewhere?
>
>
> Hardy

:)

-- 
Mikolaj

On 8 apr, 20:34, "nargess" <nargessm@n_o_s_p_a_m.terpalum.umd.edu>
wrote:
> Hi,
>
> I am interested in implementing the 2D capon method, similar to what has
> been done for the 1D case here:http://www.dsprelated.com/showmessage/78531/1.php
>
> Any suggestions as what is the best way to modify the above code to make it
> work for 2d signals?
>
> I am looking at the following paper, and am interested in applying capon
> similarly:http://www.emo.org.tr/ekler/23d0a5671efd29e_ek.pdf
> However, their code seems different from what was posted here on
> DSPRelated.

You shouldn't look at the code, but rather at the expressions.
The key is to realize how the 1D Capon works, and how to generalize
that methodology to 2D.

The 2nd equation 20 of the last paper seems to be a reasonable 2D
generalization of the first equation 20, which is the 1D version of
Capon's method. (But do work trough the arithmetics to see if the
result is correct. The inconsistent equation numbering is a hint
that the authors of the paper might not pay as acute attention to
details as one might have whished for.)

Apart from that, doce similarity means nothing. There are several
possible algorithms for each step, and several possible implemenations
of each algorithm, not to mention that coding style an attentiuon to
detail is a highly personal trait.

Don't expect to find very advanced, working code in usenet fora.

Rune

Mikolaj  <sterowanie_komputerowe@poczta.onet.pl> wrote:

>On 08-04-2010 at 22:15:04 HardySpicer <gyansorova@gmail.com> wrote:

>>> I am interested in implementing the 2D capon method, similar to what has

>> 2D capon method. Does that involve a hit in a 2D restaurant somewhere?

>:)

"D" is ancient for "shilling".  A "2D capon" would have been
quite an expensive capon in its day... maybe during a meat
shortage, or war rationing you could get that much.

A 2D capon, to justify its price, would have been LARGE
but not tender.... that size bird is 10 years old!


Cheers

Steve

>Mikolaj  <sterowanie_komputerowe@poczta.onet.pl> wrote:
>
>>On 08-04-2010 at 22:15:04 HardySpicer <gyansorova@gmail.com> wrote:
>
>>>> I am interested in implementing the 2D capon method, similar to what
has
>
>>> 2D capon method. Does that involve a hit in a 2D restaurant somewhere?
>
>>:)
>
>"D" is ancient for "shilling".  A "2D capon" would have been
>quite an expensive capon in its day... maybe during a meat
>shortage, or war rationing you could get that much.
>
>A 2D capon, to justify its price, would have been LARGE
>but not tender.... that size bird is 10 years old!

Actually, D was ancient for penny. However, since the Romans are said to
have originated the capon, 2 pence would have been darned expensive for a
capon at one time.

Steve

Thanks Rune for the comments.

On Apr 8, 2:34&#4294967295;pm, "nargess" <nargessm@n_o_s_p_a_m.terpalum.umd.edu>
wrote:
> Hi,
>
> I am interested in implementing the 2D capon method, similar to what has
> been done for the 1D case here:http://www.dsprelated.com/showmessage/78531/1.php
>
> Any suggestions as what is the best way to modify the above code to make it
> work for 2d signals?

The above link references

 http://www.sal.ufl.edu/eel6537/2006/ch4.pdf

which contains material on 2D.

Hope this helps.

Greg

On Apr 8, 3:15&#4294967295;pm, HardySpicer <gyansor...@gmail.com> wrote:
> On Apr 9, 6:34&#4294967295;am, "nargess" <nargessm@n_o_s_p_a_m.terpalum.umd.edu>
> wrote:
>
> > Hi,
>
> > I am interested in implementing the 2D capon method, similar to what has
>
> 2D capon method. Does that involve a hit in a 2D restaurant somewhere?
>
> Hardy

The 2D capons that I know are result from when they get run over on
the highway;-)

Clay

Thanks Greg, it is helpful.  Do I understand it correctly that each data
point in the two dimensional matrix 'x' is associated with two frequency
values w1,w2 as opposed to just a two dimensional frequency matrix that
would have a one to one mapping to each point in 'x'?


thanks,
n.
>On Apr 8, 2:34=A0pm, "nargess" <nargessm@n_o_s_p_a_m.terpalum.umd.edu>
>wrote:
>> Hi,
>>
>> I am interested in implementing the 2D capon method, similar to what
has
>> been done for the 1D case
here:http://www.dsprelated.com/showmessage/7853=
>1/1.php
>>
>> Any suggestions as what is the best way to modify the above code to make
=
>it
>> work for 2d signals?
>
>The above link references
>
> http://www.sal.ufl.edu/eel6537/2006/ch4.pdf
>
>which contains material on 2D.
>
>Hope this helps.
>
>Greg
>
>