Need help on my DCT and Quantization code for Image Compression.

Question

Hiroshi Nakamura 2015-5-16

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/217004-need-help-on-my-dct-and-quantization-code-for-image-compression

编辑： Prathyusha K 2017-11-20

Hello guys, I would really appreciate it if anyone could point out the mistakes in my code. I am trying to encode and decode an image by reading it in, performing DCT, Quantization then dequantizing it and performing inverse DCT. After running this code, the output Image, I2 is kind of pixellated. I have no idea how to fix it. The output should be somewhat similar to the original image but slightly blurred as it has undergone compression. Please help! My code is as follows :-

I = imread('cameraman.tif');
 I = im2double(I);
T = dctmtx(8); % dct matrix
%Performing DCT on blocks of 8 by 8
dct = @(block_struct) T * block_struct.data * T';
B = blockproc(I,[8 8],dct);
B = ceil(B); 
% A Standard Quantization Matrix
q_mtx =     [16 11 10 16 24 40 51 61; 
            12 12 14 19 26 58 60 55;
            14 13 16 24 40 57 69 56; 
            14 17 22 29 51 87 80 62;
            18 22 37 56 68 109 103 77;
            24 35 55 64 81 104 113 92;
            49 64 78 87 103 121 120 101;
            72 92 95 98 112 100 103 99];
   %PErforming Quantization by Dividing with q_mtx on blocks of 8 by 8
   c = @(block_struct) (block_struct.data) ./ q_mtx;        
   B2 = blockproc(B,[8 8],c);
  %    B2 = ceil(B2)
%Performing Inverse Quantization By Multiplying with q_mtx on Blocks of 8
%by 8
B3 = blockproc(B2,[8 8],@(block_struct) q_mtx .* block_struct.data);
%Performing Inverse DCT on Blocks of 8 by 8
invdct = @(block_struct) T' * block_struct.data * T;
% B3 = ceil(B3);
I2 = blockproc(B3,[8 8],invdct);
imshow(I), figure, imshow(I2)

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

matlabgeek 2016-1-11

Have you fixed this problem?

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Hiroshi Nakamura 2015-5-16

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/217004-need-help-on-my-dct-and-quantization-code-for-image-compression#answer_179308

Just in case you're wondering after running the code, I2 looks like this :-

Please help!

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

B.k Sumedha 2015-5-16

I think this link might help. http://in.mathworks.com/help/wavelet/ref/wcompress.html

请先登录，再进行评论。

Answer 2

B.k Sumedha 2015-5-16

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/217004-need-help-on-my-dct-and-quantization-code-for-image-compression#answer_179311

编辑：B.k Sumedha 2016-3-11

在 MATLAB Online 中打开

I = imread('cameraman.tif');
I = im2double(I);
T = dctmtx(8);
B = blkproc(I,[8 8],'P1*x*P2',T,T');
mask = [1   1   1   1   0   0   0   0
          1   1   1   0   0   0   0   0
          1   1   0   0   0   0   0   0
          1   0   0   0   0   0   0   0
          0   0   0   0   0   0   0   0
          0   0   0   0   0   0   0   0
          0   0   0   0   0   0   0   0
          0   0   0   0   0   0   0   0];
B2 = blkproc(B,[8 8],'P1.*x',mask);
I2 = blkproc(B2,[8 8],'P1*x*P2',T',T);
imshow(I), figure, imshow(I2)

2 个评论
显示无隐藏无

Hiroshi Nakamura 2015-5-16

Thanks but after looking up Quantization Matrix on Wiki, I should divide it by some Quantization Matrix other than what you have shown me. Thanks anyway!

B.k Sumedha 2015-5-16

If the answer is accepted then pleae click on accepted answer:)

请先登录，再进行评论。

Answer 3

Prathyusha K 2017-11-20

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/217004-need-help-on-my-dct-and-quantization-code-for-image-compression#answer_292147

编辑：Prathyusha K 2017-11-20

I = imread('cameraman.tif');

I = double(I);

T = dctmtx(8); % dct matrix

dct = @(block_struct) T * block_struct.data * T';

B = blockproc(I,[8 8],dct);

q_mtx = [16 11 10 16 24 40 51 61; 12 12 14 19 26 58 60 55; 14 13 16 24 40 57 69 56; 14 17 22 29 51 87 80 62; 18 22 37 56 68 109 103 77; 24 35 55 64 81 104 113 92; 49 64 78 87 103 121 120 101; 72 92 95 98 112 100 103 99];

c = @(block_struct)(block_struct.data) ./ q_mtx;

B2 = blockproc(B,[8 8],c);

B2 = round(B2);

B3 = blockproc(B2,[8 8],@(block_struct) q_mtx .* block_struct.data);

invdct = @(block_struct) round(T' * block_struct.data * T);

I2 = blockproc(B3,[8 8],invdct);

imagesc(I); colormap gray; figure, imagesc(I2), colormap gray;