| Japanese | English |
14th LSI Design Contests・in Okinawa Design Specification - 7
7. IMAGE COMPRESSION BY MATLAB®
Here we provide image compression code using MATLAB®:compress.m
In order to fully understand the computation, reading the MATLAB® code is one of good method. Please pay attention that in MATLAB index starts from 1 not 0.
As shown in Figure 8, we can see that the 8x8 pixels data (a) is compressed into quantized coefficients (c) which have a lot of zero.
(a)
(b)
(c)
Figure8 The example of input and computation result of the compression module.
(a) 8x8 pixles data input (b) The result of DCT processing (c) The result of Quantization.
At the decompression module, the quantized coefficient from previous processing can be used to reconstruct the original image. The computation result example can be seen in Figure 9. We can see that the pixels value in Figure 8 (a) is similar to Figure 9(b).
(a)
(b)
Figure9 The example of input and computation result of the decompression module.
(a) The result of Quantization processing (b) The result of IDCT processing.
After processing is done for all pixels, we can see the reconstructed image in Figure 10.
|
|
Figure10 The comparison between original and reconstructed image