2-D DCT/IDCT for JPEG Compression

Fast Implementation of 8X8 DCT/IDCT for JPEG Compression

Description

to understand the Algorithm go to matlab help in page dct2 and idct2 to get the mathematical expression

for M = N = 8, we can calculate the most of hard values and save it as LUTs to speed up the execution

now compare our special 8X8 functions with the internal general functions

use this code:

A = int32(100*rand(8,8));
tic;for i = 1 : 1000 IDCT_8X8(DCT_8X8(A));end;toc;
tic;for i = 1 : 1000 idct2(dct2(A));end;toc;

i had built the general functions too, but with low speed

• Required Products: Image Processing Toolbox
• MATLAB release: MATLAB 7.0.1 (R14SP1)
• Other Files: DCT_8X8.m, ; IDCT_8X8.m

[Ali OZGUL]

Here all sub-functions are similarly for program running procedure. Only, per functions are different to graphical-compression's functions;

DCT8X8 function:
 for n = 1 : 8
                s = s + (double(I(m,n)) * cosines(p,m) * cosines(q,n));
            end

IDCT function:
for q = 1 : 8
                s = s + (alpha(p,q) * double(I(p,q)) * cosines(p,m) * cosines(q,n));
            end

If you one-sub program's be running one module than I think very this function is very usefully.

For example: Dcomp.m(I,flag)
if flag==1 than running DCT_8x8.m
           else running IDCT_8x8.m
end

Good works, best regards