Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
JPEG uses DCT for compression whereas JPEG2000 relies on Wavelet/Subband Coding. Techincally, JPEG2000 has reversible compression with precision arithmetic, something not achievable with JPEG. JPEG2000 uses bit plane coding for more efficiency.
The novel aspect of JPEG2000 is that you encode once and decode multiple times on whatever device, resolution, you want. It is highly scalable. It will decode only the part of the picture that you want to see unlike JPEG. Also, there are many security related features supported by JPEG2000.
Hi ,
i would like to add few more points
1) in jpeg we lose high frequency information , but not in jpeg-2000
2) At lower bitrate jpeg produces blocking artifacts ,
3) in the same jpeg2000 bit stream can have lossy & losless compression .
The main different is,
1. Transform: JPEG is DCT and JPEG2000 is DWT
2. Entropy Coding: JPEG ia Huffman and JPEG2000 id QM (Arithmetic Coding)
3. Pre-Entropy Coding: JPEG is RLE and JPEG2000 is EBCOT
You can get more papres form IEEE with search on EBCOT and JPEG2000 as keyword.
Added after 5 minutes:
Ok, I will post some IEEE papers on JPEG200 to EDA, check with the IEEE paper section.
jpeg uses discrete cosine transform
jpeg 2000 uses wavelet transform
in general first split the image into blocks
then apply the transform
then quantization using quantization table
zig zag conversion and then apply RLE
refer any image compression book for further ref.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
