Currently we have implemented and verified gray jpeg-ls encoder with pixel width 8-, 10-, 12-, 16 bits. Also we have written RGB encoder, gray and RGB decoder, so if you are interested in this project send us your pictures and we are testing core on them.

Result of the hardware implementation of the gray jpeg-ls encoder for HD images (1920x1080) on the Altera StratixII FPGA (EP2SGX30F780C3) looks like:

  8  bits: 11% ALUT, 2% memory, 91,51 MHz
 10 bits: 12% ALUT, 3% memory, 83,48 MHz
 12 bits: 13% ALUT, 3% memory, 82,62 MHz
 16 bits: 15% ALUT, 4% memory, 80,76 MHz                                                                

We have implemented video Encoder IP Core according to ITU-T Recommendation T.87 (1998) | ISO/IEC 14495-1:1999 – Lossless and near-lossless compression of continuous-tone still images: Baseline.   These compression method are widely referred to as "JPEG-LS".

Lossless compression is the most optimal to work in the spheres where preserving a 100% image quality is of vital importance.

The primary purpose of the hardware implementation JPEG-LS is the real-time packing raw HighDefinition videos  for saving storage space without any impact in quality and usability of the pictures. The achieved compression ratio for  real images is about 2:1 that is the significantly better compared to the use of a common arithmetic coding. See our  comprasion table and expert's articles concerning into selection of the best lossless image compression method.