How to extract the highest 300 coefficient from each level of decomposition in shearlet transform( from level 1 to 5).

1 次查看（过去 30 天）
Adnan Ibrahim
Adnan Ibrahim 2015-8-17
评论： Walter Roberson 2018-1-5
I have extracted some statistical features (mean, standard deviation, constranst.. etc), but Now I want to test using highest 300 coefficients from each level of decompostion... please I need your suggestion. Thank you
if nargin < 1
img = double(imread('Img205.png','png'));
end
[row col] = size(img);
ndir = [2 2 1 1 0];
for j = 1:5
for k = -2^(ndir(j)):2^(ndir(j))
coeff_h{j}(:,:,k+(2^(ndir(j))+1)) = obtain_shear_coeff(img,[3 3 4 4 5],ndir(j),'hor',j,k);
coeff_v{j}(:,:,k+(2^(ndir(j))+1)) = obtain_shear_coeff(img,[3 3 4 4 5],ndir(j),'ver',j,k);
end
end
% Display original image
colormap gray;
subplot(1,1,1), imagesc( img, [0 255] ) ;
title( sprintf('Original Image' )) ;
% Display shearlet coeff for horizontal cone
figure; clf;
colormap gray;
tmp1 = []; tmp2 = []; tmp3 = []; tmp4 = []; tmp5 = [];
for k = 1:9
if k < 9
tmp1 = [ tmp1; coeff_h{1}(:,:,k); 255*ones(5,col/2) ];
else
tmp1 = [ tmp1; coeff_h{1}(:,:,k)];
end
end
for k = 1:9
if k < 9
tmp2 = [ tmp2; coeff_h{2}(:,:,k); 255*ones(5,col/4) ];
else
tmp2 = [ tmp2; coeff_h{2}(:,:,k)];
end
end
for k = 1:5
if k < 5
tmp3 = [ tmp3; coeff_h{3}(:,:,k); 255*ones(5,col/8) ];
else
tmp3 = [ tmp3; coeff_h{3}(:,:,k)];
end
end
for k = 1:5
if k < 5
tmp4 = [ tmp4; coeff_h{4}(:,:,k); 255*ones(5,col/16) ];
else
tmp4 = [ tmp4; coeff_h{4}(:,:,k)];
end
end
for k = 1:3
if k < 3
tmp5 = [ tmp5; coeff_h{5}(:,:,k); 255*ones(5,col/32) ];
else
tmp5 = [ tmp5; coeff_h{5}(:,:,k)];
end
end
subplot(1,5,1), imagesc( tmp1 ) ;
subplot(1,5,2), imagesc( tmp2 ) ;
subplot(1,5,3), imagesc( tmp3 ) ;
title( sprintf('Shearlet Coefficients for horizontal cone' )) ;
subplot(1,5,4), imagesc( tmp4 ) ;
subplot(1,5,5), imagesc( tmp5 ) ;
% Display shearlet coeff for vertical cone
figure; clf;
colormap gray;
tmp1 = []; tmp2 = []; tmp3 = []; tmp4 = []; tmp5 = [];
for k = 1:9
if k < 9
tmp1 = [ tmp1 coeff_v{1}(:,:,k) 255*ones(row/2,5) ];
else
tmp1 = [ tmp1 coeff_v{1}(:,:,k)];
end
end
for k = 1:9
if k < 9
tmp2 = [ tmp2 coeff_v{2}(:,:,k) 255*ones(row/4,5) ];
else
tmp2 = [ tmp2 coeff_v{2}(:,:,k)];
end
end
for k = 1:5
if k < 5
tmp3 = [ tmp3 coeff_v{3}(:,:,k) 255*ones(row/8,5) ];
else
tmp3 = [ tmp3 coeff_v{3}(:,:,k)];
end
end
for k = 1:5
if k < 5
tmp4 = [ tmp4 coeff_v{4}(:,:,k) 255*ones(row/16,5) ];
else
tmp4 = [ tmp4 coeff_v{4}(:,:,k)];
end
end
for k = 1:3
if k < 3
tmp5 = [ tmp5 coeff_v{5}(:,:,k) 255*ones(row/32,5) ];
else
tmp5 = [ tmp5 coeff_v{5}(:,:,k)];
end
end
subplot(5,1,1), imagesc( tmp1 ) ;
title( sprintf('Shearlet Coefficients for vertical cone' )) ;
subplot(5,1,2), imagesc( tmp2 ) ;
subplot(5,1,3), imagesc( tmp3 ) ;
subplot(5,1,4), imagesc( tmp4 ) ;
subplot(5,1,5), imagesc( tmp5 ) ;
%
% Copyright (c) 2010. Wang-Q Lim
  2 个评论
shakin banu
shakin banu 2018-1-4
Sir, While executing this code , I am getting error as Undefined function 'obtain_shear_coeff' for input arguments of type 'double'.
Error in sheacoef (line 8) coeff_h{j}(:,:,k+(2^(ndir(j))+1)) = obtain_shear_coeff(img,[3 3 4 4 5],ndir(j),'hor',j,k); Kindly help me to run this code. Also I would like to know how to extract sparse coefficients from shearlet transform.
by Shakin, Ph.D Research Scholar

请先登录，再进行评论。

请先登录，再回答此问题。

采纳的回答

Walter Roberson
Walter Roberson 2015-8-17
[sorted_data, sort_index] = sort(DataYouWantToGetTop300ValuesOf);
top300values = sorted_data(end-299:end);
index_of_top300values = sort_index(end-299:end);

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Subplots 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by