Help using idwt3 - high and low pass filters

I am trying to use idwt3 on my data after applying soft thresholding but I am not sure how to do this exactly; I used dwt3(x, sym2) on my data but as I understand it in order to get back to the original displacement domain I can just use idwt3(x, sym2) I must decide low and high pass filters and input characters such as aaa etc. Is anyone well versed with this toolbox and can guide me through it

thanks bran

Answer 1

[Wayne King]

dwt3 returns a structure array. The 3-D wavelet coefficients are in the dec field and are ordered as explained in the documentation for dwt3.m

For example, wt.dec{1,1,1} contains the LLL subband so to construct an image approximation solely based on the the LLL image, I can set all the others to zero and invert.

X = reshape(1:64,4,4,4);
wt = dwt3(X,'sym2','mode','per');
decnew = cell(2,2,2);
decnew(:) = {zeros(2,2,2)};
decnew{1,1,1} = wt.dec{1,1,1};
wt.dec = decnew;
X1 = idwt3(wt);

[Bran]

Hi there thanks so much for your help so I am trying to use this function on a signal that is 17x17x17 which has gone through the fast fourier transform. Which would be the best parameters to use in my case? This is my code;

tmp = fftn(MAT); Adash = fftshift(tmp);

 Awave=dwt3(Adash,'sym2'); Y = wthresh(Awave,'s',0.1);PLEASE NOTE! HERE IS WHERE I MADE MY ORIGINAL ERROR!!! Awaveup = idwt3(Y, 'sym2');

 Adashup=ifftshift(Awaveup);
 Adashup2 = ifftn(Adashup);Many thanks in advance

Answer 2

[Bran]

Hi there thanks so much for your help so I am trying to use this function on a signal that is 17x17x17 which has gone through the fast fourier transform. Which would be the best parameters to use in my case? This is my code;

tmp = fftn(MAT); Adash = fftshift(tmp);

 Awave=dwt3(Adash,'sym2');

 Y = wthresh(Awave,'s',0.1);

PLEASE NOTE! HERE IS WHERE I MADE MY ORIGINAL ERROR!!! Awaveup = idwt3(Y, 'sym2');

 Adashup=ifftshift(Awaveup);
 Adashup2 = ifftn(Adashup);

Many thanks in advance