I am trying "nldif" function, and find that there are two options for diffusion: one is "AOS" by aosiso.m and the other is "iso" by isodifstep.
For "AOS", the result is good when I choose stepSize = 2; But for "iso", I cannot get good result for any stepSize (from 0.01 to 1...).
I want to know what value of stepSize for "iso" case can get the same result as "aos". I suppose "aos" and "iso" should get the same result when choose proper stepSize. The only different should be the speed. Is it correct?
And when I try to feed back from "aos" result to "iso" stepSize ( (aosiso(y, g, 2) - y)/dy ), but I can not get a uniform stepSize value.
Can you help me to explain above case?
code for feedback:
% Calculate dy/dt
% if plotflux
% yo = y;
% y = aosiso(y,g,stepsize(i)); % updating
% dy = isodifstep(y, g);
% y = y + stepsize(i) * dy; % updating
tmp = y;
tmp1 = aosiso(y, g, 2) - tmp;
dy = isodifstep(y, g);
ios_step = tmp1 / dy;
There is no productive 2D example provided that show the superiority of this method over regular gaussian smoothing. A simple combination of Gaussian and Median filtering provides better denoising than what is proposed:
s = phantom(512) + randn(512);
num_iter = 55;
delta_t = 1/7;
kappa = 30;
option = 2;
ad = anisodiff2D(s,num_iter,delta_t,kappa,option);
It is very likely that the authors generalization of using 8 point neighbourhoods is valid, but if so it changes the valid range of the parameters, and thus one cant easily recreate the results of the original Perona-Malik paper. If a simple example was provided that showed how the 2D method actually outperforms gaussian blurring, then I might be inclined to read the secondary paper referenced.
the interpretation of the kappa parameter is nicely given in weickerts book: http://www.lpi.tel.uva.es/muitic/pim/docus/anisotropic_diffusion.pdf, chapter 1, figure 1.1, b... (its called lambda there)
I cant get that interpretation to work for me here. Whats wrong? How can i strengthen the edges in the image using this, while blurring along isolines!?
Again, if only a positive example was provided for the 2D case, this might all be cleared up.