Re:Free Lossless Image Format Publications
by jonsneyers on Tuesday October 04, 2016 @03:02AM
(#12633773)
Here are some results for a few black & white images from https://commons.wikimedia.org/wiki/Category:PD-ScottForesman :
(I only compare to JBIG2, not Group 4, since JBIG2 is supposed to be the best of those two)
JBIG2 FLIF PNG
11690 10903 27093 1971_San_Jose_Scale_Scales_on_Branch
10458 9761 23466 1971_San_Jose_Scale_Winged_Adult
44809 44301 105555 Accommodation_ladder
15966 14882 44073 Aconite
27031 25877 59353 Adder
13569 13302 40317 Adenoids
36409 35058 84826 African_elephant
6890 6899 12982 Aft
24465 23170 51297 Agate_A-40
18969 17896 24261 Aisle
8612 8590 23408 Allen_wrench_and_screw
9968 9104 11792 Alpaca
12258 11952 36657 Alternately
196083 192611 254895 Amazon_River
76572 73529 105637 Arabian_Sea
275124 256820 311448 Archery_colored
139482 137329 162432 Argonne_Forest
40681 39952 70999 Astrolabe
24950 23317 32007 Brace
43808 41113 116671 Deer_2
Re:Free Lossless Image Format Publications
by faragon on Tuesday October 04, 2016 @03:25AM
(#12633864)
Thank you! It looks amazing. So unless processing speed or memory requirements being a handicap, the FLIF could be a real generic graphics compression candidate :-)
Re:Free Lossless Image Format Publications
by jonsneyers on Tuesday October 04, 2016 @09:10AM
(#12635209)
Yes, that's the goal :)
Re:Free Lossless Image Format Publications
by jonsneyers on Monday October 03, 2016 @03:43PM
(#12630663)
Very good question. Do you or anyone else have a good test set of such images?
My gut feeling says that JBIG2 probably outperforms FLIF on repetitive images (e.g. text), since it specifically detects such patterns. On non-repetitive images, maybe FLIF is better. But it would be nice to have some benchmarks.
Re:Free Lossless Image Format Publications
by boramalper on Monday October 03, 2016 @10:30PM
(#12632936)
> Patents for JBIG2 are owned by IBM and Mitsubishi.
[1] https://en.wikipedia.org/wiki/Group_4_compression
[2] https://en.wikipedia.org/wiki/JBIG2