1 #+TITLE: ImageSqueeze - lossy image codec
4 - [[http://www2.svjatoslav.eu/gitweb/?p=imagesqueeze.git;a=snapshot;h=HEAD;sf=tgz][download latest snapshot]]
6 - This program is free software; you can redistribute it and/or modify it under
7 the terms of version 3 of the [[https://www.gnu.org/licenses/lgpl.html][GNU Lesser General Public License]] or later as
8 published by the Free Software Foundation.
12 - Homepage: http://svjatoslav.eu
13 - Email: mailto://svjatoslav@svjatoslav.eu
15 - [[http://www.svjatoslav.eu/programs.jsp][other applications hosted at svjatoslav.eu]]
17 * (document settings) :noexport:
18 ** use dark style for TWBS-HTML exporter
19 #+HTML_HEAD: <link href="https://bootswatch.com/darkly/bootstrap.min.css" rel="stylesheet">
20 #+HTML_HEAD: <script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/1.11.2/jquery.min.js"></script>
21 #+HTML_HEAD: <script src="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/3.3.1/js/bootstrap.min.js"></script>"
22 #+HTML_HEAD: <style type="text/css">
23 #+HTML_HEAD: footer {background-color: #111 !important;}
24 #+HTML_HEAD: pre {background-color: #111; color: #ccc;}
27 Lossy image codec. Optimized for photos. I developed it to test out an
28 image compression ideas.
30 I believe my [[id:1d917f74-e763-4a71-976e-4aa60732efa6][algorighm has interesting advantages.]]
32 Below are original photo and the same image being compressed down to
33 ~93 Kb and then decompressed.
35 [[file:originalAndCompressed.png][file:originalAndCompressed.png]]
37 When looking very closely, slight grainyness, loss of color precision
38 and blurriness (loss of detail) could be noticed as a compression
39 artifacts. Still sharp edges are always preserved. Also no blocks
40 typical to JPEG are ever seen. I think that is awesome result for
41 just ~ 2.5 bits per pixel on that color photo.
42 * Algorithm description
43 + Color image is split into three separate channels: Y, U and V.
44 + Each channel is independently compressed.
46 + Single channel compression method:
48 + Algorithm is inspired by [[https://en.wikipedia.org/wiki/Diamond-square_algorithm][diamond-square algorithm]] that is meant to
49 produce random heightmap/plasma effect.
51 + During image compression: 2D image surface is iterated in the
52 similar manner to diamond-square algorithm. Average color from
53 neighbors is calculated and difference between neighbors average
54 and actual pixel color is saved. As the pixel grid becomes
55 gradually more dense, difference between neighbors tends to get
56 smaller, thereby requiring less bits per pixel for storing the
58 ** Algorighm advantages
60 :ID: 1d917f74-e763-4a71-976e-4aa60732efa6
62 + It can be applied to any amount of dimensions, even for sound and
65 + Algorithm can operate in lossy and lossless mode.
67 + Algorithm naturally handles progressive loading. That is: low
68 resolution thumbnail of entire thing is immediately available and
69 gets gradually more dense during entire loading process.
71 + Algorithm naturally supports variable resolution. That is: different
72 areas can be encoded with different resolutions / pixel densities.
74 + Fast: Very little computations per pixel.
76 * TODO Things to improve
77 - Code documentation is weak.
79 - Better sample applications needed:
80 - Commandline image conversion utility.
83 - Add lossless support.
85 - Extract algorithm key parts into reusable library and make it
86 variable dimensional. So that the same code can be used for sound,
87 image and volumetric data compression/decompression.