From: Manal on
Hi,
I hope some can help me
I am struggling on this for 4 days

I am trying to implement the star skeleton algorithm
I extracted the border, smoothed the border by using (DFT + low pass filter + IDFT)
the next step is to find the five extreme points which are detected by finding zero-crossings of the different function

s(i)=d(i)-d(i-1)

could some one explain this to me

I'll really appreciate that

thanks in advance
From: Manal on
to be clear
this is the algorithm that I am trying to implement

http://www.vision.cs.chubu.ac.jp/04/pdf/VSAM08.pdf

Waiting for some one to help me :(
From: Manal on
please guys help me
I just need some one to explain it for me
From: TideMan on
On Jun 17, 8:36 am, "Manal " <be....(a)live.com> wrote:
> please guys help me
> I just need some one to explain it for me

I know nothing about star skeleton and I'm not interested in reading
the reference you posted.
But I can advise on zero crossing analysis.
Here is how you can locate them:
t1=y(1:n-1);
t2=y(2:n);
tt=t1.*t2;
indx=find(tt<0);

What it does is locate the places where adjacent points have different
signs (i.e., they are either side of the axis).
From: Manal on
TideMan <mulgor(a)gmail.com> wrote in message <9e9eebf5-6e64-468f-9dd7-b43f8be4d44e(a)r27g2000yqb.googlegroups.com>...
> On Jun 17, 8:36 am, "Manal " <be....(a)live.com> wrote:
> > please guys help me
> > I just need some one to explain it for me
>
> I know nothing about star skeleton and I'm not interested in reading
> the reference you posted.
> But I can advise on zero crossing analysis.
> Here is how you can locate them:
> t1=y(1:n-1);
> t2=y(2:n);
> tt=t1.*t2;
> indx=find(tt<0);
>
> What it does is locate the places where adjacent points have different
> signs (i.e., they are either side of the axis).

thanks a lot for your replay
thats what I want
some one to give me the sample code for the zero-crossings

but meaning of the zero-crossings for the difference function??????

I've tried your code and it display all the vector content as "indx" value