Prev: textread: whitespaces help!
Next: question about bwperim
From: Lulu Ardiansyah on 20 Jul 2010 14:29
Hello all.

I'm trying to learn the super-resolution reconstruction, but I have a problem on the motion model. The equation that describes the model is:

f(k) = M(k, i) f(j) + e(k, j)

Where f(k) and f(j) is the vectors representing frames at the k-th frame and the j-th.
The matrix M(k,j), with size M x M which represents the motion transformation or warping.

My question is, how to form the matrix M? It says here that the matrix M is usually sparse.

Please advise. thanks.
From: us on 20 Jul 2010 14:51
"Lulu Ardiansyah" <halluvme(a)gmail.com> wrote in message <i24pu1$d6i$1(a)fred.mathworks.com>...
> Hello all.
>
> I'm trying to learn the super-resolution reconstruction, but I have a problem on the motion model. The equation that describes the model is:
>
> f(k) = M(k, i) f(j) + e(k, j)
>
> Where f(k) and f(j) is the vectors representing frames at the k-th frame and the j-th.
> The matrix M(k,j), with size M x M which represents the motion transformation or warping.
>
> My question is, how to form the matrix M? It says here that the matrix M is usually sparse.
>
> Please advise. thanks.

a hint:

help sparse; % <- and siblings...

us
 | 
Pages: 1
Prev: textread: whitespaces help!
Next: question about bwperim