Is it possible to combine, lossy and lossless image compression???

2 views (last 30 days)
Sir, is it possible to combine lossy and lossless compression together,, for eg, I've a brain MRI image (tumor) , i segmented as ROI and non ROI , so now, i've to compress ROI with lossless, and non ROI with lossy compression and to combine both.... so far i segmented the roi and non-roi separately,.. with salience i separated compressed roi, but how to combine both??? thanks in advance..
  1 Comment
Radhika UD
Radhika UD on 11 Nov 2016
Edited: Radhika UD on 11 Nov 2016
Hi..I am in need of same type of compression now.. Please share for any solutions.. Thanks..

Sign in to comment.

Accepted Answer

Image Analyst
Image Analyst on 2 Apr 2013
I wasn't sure that was legal, but maybe times are changing. See this 2006 paper that proposes a method exactly as you described: http://www.vast.uccs.edu/~tboult/PAPERS/IJBRA-06-Zukowski-Boult-Medical-Image-Compression.pdf

More Answers (1)

Walter Roberson
Walter Roberson on 11 Nov 2016
For example you can take
data_in_ROI = YourData(ROI_mask);
data_outside_ROI = YourData(~ROI_mask);
compressed_mask = run_length_encode_bitwise(ROI_mask);
compressed_inside = lossless_compression(data_in_ROI);
compressed_outside = lossy_compression(data_outside_ROI);
compressed_result = [compressed_mask, compressed_inside, compressed_outside];
Restoration would be decoding the compressed result to separate the parts, then uncompressing the mask, uncompressing the losslessly-compressed inside data, storing it in the positions where the mask is true, uncompressing the lossy outside data, storing that in the positions where the mask is false.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!