Differentiation with convolution operators

  • Thread starter anja.ende
  • Start date
  • #1
anja.ende
5
0
Hello,

I have really been banging my head the whole day and trying to figure this derivative out. I have a function of the following form:

F = W * (I.J(t)) - (W * I).(W*J(t))

where I and J are two images. J depends on some transformation parameters t and W is a gaussian kernel with some fixed standard deviation and zero mean. * represents the convolution operator. Now, I want to compute the derivative of F wrt to the transformation parameters 't'.

So, I try the following:

[itex]\frac{dF}{dt} = \frac{d}{dt} [(I . W*J(t)) - (W*I)(W*J(t))][/itex]

I can talk 'I' out as it can be treated as a constant. This gives (I think):

[itex]\frac{dF}{dt} = (I. W*J'(t)) - (W*J'(t)) . (W*I)[/itex]

Can I treat the convolution operators this way or is this wrong? The convolution kernels are fixed width Gaussians and do not depend on the parameters 't'.

Thanks for any help you can give me.

Anja
 
Last edited:

Answers and Replies

  • #2
CompuChip
Science Advisor
Homework Helper
4,309
49
The definition is
$$(f \ast g)(t) = \int_\mathbb{R} f(\tau) g(t - \tau)$$
so with the appropriate smoothness conditions and all that, you can easily verify
$$(f \ast g)'(t) = \int_\mathbb{R} f(\tau) g'(t - \tau) = (f \ast g')(t)$$
 

Suggested for: Differentiation with convolution operators

  • Last Post
Replies
2
Views
728
Replies
46
Views
2K
Replies
17
Views
892
Replies
5
Views
448
  • Last Post
Replies
1
Views
476
Replies
14
Views
650
  • Last Post
Replies
5
Views
609
  • Last Post
Replies
2
Views
777
  • Last Post
Replies
16
Views
1K
Top