|
Lossless compression results on the New Test Images.
Implementations Used
- For Jpeg-ls, UBC's open-source implenemtation
v2.2 was used. (No parameters)
- For Jpeg 2000, Kakadu v6.0 demo executables were
used (Parameters: Creversible=yes -rate -)
- For HD Photo, MS DPK v1.0 was used (-c X,
depending upon image bit-depth, color)
Tests were run on a AMD 3200+ (1.28GHz) with 1.25 GB RAM
running MS Windows XP.
Let me know if you want to see some other
algorithm/implementation included in the benchmark. Goal is
not to be exhaustive and cover all
algorithms/implementations, the results must be interesting
enough.
Compression ratio and speed results are given below, in
both cases smaller bar means better
results. Click on the images to enlarge, you can enlarge multiple
images at same time.
RGB 8 Bit
.png "Click to enlarge")
Compression Ratio
|
.png "Click to enlarge")
Time Taken
|
RGB 16 Bit
.png "Click to enlarge")
Compression Ratio
|
.png "Click to enlarge")
Time Taken
|
RGB 16 Bit Linear
.png "Click to enlarge")
Compression Ratio
|
.png "Click to enlarge")
Time Taken
|
Gray 8 Bit
.png "Click to enlarge")
Compression Ratio
|
.png "Click to enlarge")
Time Taken
|
Gray 16 Bit
.png "Click to enlarge")
Compression Ratio
|
.png "Click to enlarge")
Time Taken
|
Gray 16 Bit Linear
.png "Click to enlarge")
Compression Ratio
|
.png "Click to enlarge")
Time Taken
|
|
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
.png "Click to enlarge")
