MATLAB Central Newsreader - HARRIS Cornerness measure -- Confusion!?!

HARRIS Cornerness measure -- Confusion!?!

Nan W. — Wed, 11 Jun 2008 16:39:02 -0400

Dear all,

I am wondering that how come there are many solutions to
the Cornerness measurement on Harris corner dectection.

The novel one is:
C= det(M)-k*trace(M).^2

right?

but there are those in:
1) C= trace(M)./det(M)
2) C= det(M)./(trace(M)+eps)
3) C= (k+1/k)*det(M)-(trace(M).^2-2det(M))

Does anyone have any theories support these latter three
equations? It seems that some of them also concentrate on
edges plus corners but I really can't find any documents
supported these equations.

Please help me out. Thank you so much.

Re: HARRIS Cornerness measure -- Confusion!?!

Steven Lord — Wed, 11 Jun 2008 16:45:40 -0400

"Nan W." <iiuu_chan@yahoo.co.jp> wrote in message
news:g2ov36$fh4$1@fred.mathworks.com...
> Dear all,
>
> I am wondering that how come there are many solutions to
> the Cornerness measurement on Harris corner dectection.

*snip*

This doesn't seem to be a MATLAB question, but more of a general image
processing question, based on what Wikipedia says:

http://en.wikipedia.org/wiki/Corner_detection

You might have better luck asking about this in a group dedicated to image
processing, like sci.image.processing, which is available in Google Groups
if your news server doesn't carry it:

http://groups.google.com/group/sci.image.processing?lnk=srg

--
Steve Lord
slord@mathworks.com

Re: HARRIS Cornerness measure -- Confusion!?!

us — Wed, 11 Jun 2008 16:49:03 -0400

"Nan W.":
<SNIP theoretical question...

> I am wondering that how come there are many solutions to
> the Cornerness measurement on Harris corner dectection...

this NG is dedicated to MATLAB (ML) and its siblings...
now, where's your ML problem?

us

Re: HARRIS Cornerness measure -- Confusion!?!

roberson@ibd.nrc-cnrc.gc.ca (Walter Roberson) — Wed, 11 Jun 2008 17:02:48 -0400

In article <g2ov36$fh4$1@fred.mathworks.com>,
Nan W. <iiuu_chan@yahoo.co.jp> wrote:

>I am wondering that how come there are many solutions to
>the Cornerness measurement on Harris corner dectection.

Sorry, that is beyond my experience.

>but there are those in:
>1) C= trace(M)./det(M)
>2) C= det(M)./(trace(M)+eps)
>3) C= (k+1/k)*det(M)-(trace(M).^2-2det(M))

>Does anyone have any theories support these latter three
>equations?

Equation (2) is incorrect. The +eps in the denominator is attempting
to avoid division by 0 in case trace(M) is 0. However, trace(M)
can be positive or negative, and in particular trace(M) could be
a negative number greater than -1 * eps: in such a case, the addition
of eps in the denominator would change the sign of the result. The
addition of eps in the denominator could even -cause- a division by 0
where there was no division by 0 before.

--
  "I feel sorry for the person who can't get genuinely excited
  about his work. Not only will he never be satisfied, but he will
  never achieve anything worthwhile." -- Walter Chrysler

Re: HARRIS Cornerness measure -- Confusion!?!

Nan W. — Thu, 12 Jun 2008 14:57:01 -0400

Thank you to you all.
I will go and post my question again in google groups.

Thanks ^^

Re: HARRIS Cornerness measure -- Confusion!?!

Novina Soraya — Thu, 26 Jun 2008 04:03:04 -0400

Would you help me to learn all about Harris Corner? because
my final project is about Corner Detection with Harris
Method... Please help me... I really thankful for the
attention... Please..

Re: HARRIS Cornerness measure -- Confusion!?!

Novina Soraya — Thu, 26 Jun 2008 04:03:20 -0400

Would you help me to learn all about Harris Corner? because
my final project is about Corner Detection with Harris
Method... Please help me... I really thankful for the
attention... Please..