JPEG image compression is Fourier based DCT while modern image compression technique like JPEG2000 is based on more multi-scale technique like Wavelets.I want to know how Fourier and Wavelets are useful in Image compression.

So can anybody explain advantages and disadvantages of JPEG and JPEG2000 with the help of (characteristics of) transforms they use?


JPEG is far simpler. It divides the image into 8x8 pixel blocks, and processes each using a Discrete Cosine Transform. The results are quantised and then encoded. The quality is fixed by the encoder.

JPEG2000 uses a 2D wavelet function, the output of which is four "images", each a quarter the size of the original. One of those is actually an image, while the others are high-frequency components that can be added to it to re-construct the full-resolution image. The wavelet process may be repeated multiple times. The result is a tiny image, and a series of high frequency components that may be combined with it. Each resulting component is quantised and encoded.

JPEG is fine for high quality and modest compression, which is why it is still very widely used.

JPEG2000 offers several advantages:

  • To achieve very high compression, it is possible to throw away or heavily quantise the high frequency components. This gives a poor quality - but usable - image where JPEG would fail completely.
  • Images can be re-constructed progressively at ever improving quality. This can be either in terms of increasing resolution or bit depth as required.
  • It supports the JPIP protocol for progressively transmitting images to a client. The client may retrieve low-resolution thumbnails and then just the parts of the image they want at better resolutions.

JPEG uses, in its core, a block-based transform called Discrete Cosine Transform. The inherent result of this choice is known as the blocking artifact. To overcome this artifact and to accomodate many new advanced features required by the complex communication environment and demanding applications of the global communicaitons era it was required to be replaced by a wavelet type of multiresolution transform within the JPEG2000 codec.

JPEG2000 provides more efficient compression, with many added modern features. The most important artifact of JPEG2000 is called as the washed-out effect observed at very low bitrates, which results in muddy outputs. Note however that JPEG at such low bitrates totally fails.

As a result, JPEG 2000 is a better but more complicated codec.

  • $\begingroup$ I think JPEG 2000 is also processed by blocks yet instead of DCT uses Waveltes, are you sure about "Non Block Based"? I think JPEG 2000 allows any size of blocks, bust still works in blocks (Though you really can take the whole image as one block, yet the compression will be low). $\endgroup$
    – Royi
    May 27 '15 at 10:49
  • $\begingroup$ you are right that is called tiling in JPEG2000, actually "nonblock based" seems to be a wrong description that I have to edit. $\endgroup$
    – Fat32
    May 27 '15 at 10:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.