I'm recently learning image splicing detection. I've already known image splicing can bring some abrupt changes to an image,which generate the sharp edges.And sharp edges correspond to high frequency. And I want to extract features from images for splicing detection. In my opinion,I should use a high-pass filter to enhance the edges,but I've read some papers which mentioned using a low-pass filter for gradient detection techniques applied in discontinuity detection. Now I am very bothered with this.Which kind of filters should I use?I would really appreciate if someone can help me!
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
3
Since you're after a filter which emphasizes abrupt changes you should use High Pass Filter.
The issue is you'd be also sensitive to noise.
Hence one way to do it is to apply High Pass Filter on slightly blurred image (Which actually results in a Band Pass Filter).
One easy choice would be using the Gradient of The Gaussian Filter.
Why is it?
Because it is easy to calculate (Analytically) and efficient to implement because of its separability property of the filter.
-
$\begingroup$ Thank you for answering!But I have another problem.Should I care about the issue of a high pass filter sensitive to noise when it comes to splicing detection?If it is necessary to apply a pre-smoothing of Gaussian filter before a high pass filter? $\endgroup$ Dec 14, 2015 at 2:15
-
$\begingroup$ It depends how smooth is the transition between the two images. $\endgroup$– RoyiDec 14, 2015 at 16:26
-