vis
|
The CSIQ image database
The CSIQ
database is a new
database released by our
Lab. It consists of 30
original images, each is
distorted using six
different types of
distortions at four to
five different levels of
distortion. CSIQ images are subjectively rated base on a
linear displacement of the images across four calibrated LCD
monitors placed side by
side with equal viewing
distance to the observer. The database contains 5000 subjective ratings
from 35 different observers, and ratings are reported in the form
of DMOS. If you use this database, please cite the following paper:
E. C. Larson
and D. M.
Chandler,
"Most
apparent
distortion:
full-reference
image
quality
assessment
and the role
of strategy,"
Journal
of
Electronic
Imaging,
19 (1),
March
2010.
.PDF .bibtex
This page is for the CSIQ image database, the video database can be found here.
|
|
|
Each
original image in the
database is distorted
using six different
types of distortions at
four to five different
levels of distortion.
The distortions used in CSIQ are: JPEG compression, JPEG-2000 compression,
global contrast decrements, additive pink Gaussian noise, and
Gaussian blurring. This result in 866 distorted versions of
original images. Examples of these distortion types are shown
below:
|
JPEG compression |
|
|
JPEG-2000 compression |
|
 |
 |
 |
 |
 |
 |
|
Global contrast
decrements |
|
Gaussian
blurring |
|
|
 |
 |
 |
 |
 |
 |
|
Additive
Gaussian white
noise |
|
Additive
Gaussian pink
noise |
|
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
Subjective Rating of Perceived Distortion:
|
|
CSIQ images are subjectively rated base on a
linear displacement of the images. Four Sceptre X24WG LCD monitors at
resolution of 1920 x
1200 were calibrated
following sRGB color
space and placed side by side with equal viewing distance to the
observer. Subjects were instructed to keep the viewing distance
stable of 70cm.
All of the distorted
versions of an original
image were viewed
simultaneously across the monitor array.
Each subject placed
these images in the
monitor array in which
the horizontal distance
between two images
reflexes the
perceived quality of
them. One image placed
close to another means
that the subject thinks
their quality is not
much different. On the
other hand, one image
place further away to
the left of another
means the subject thinks
the former image is much
worse in quality
compared to the latter.
As the final step, across image
ratings are realigned according to a separate, but identical,
experiment in which observers place subsets of all the images
linearly in space.
There are 35 observers
participated in this
experiment. Subjects
include both male and
female with normal or
correct-to-normal visual
acuity. Subjects' ages
range from 21 to 35.
Overall the database contains 5000 subjective ratings
and are reported in the form
of DMOS. |
|
|
|
|
|
|
|
|
|
|
|
This project was
supported by the
National Science
Foundation,
"Content-Based
Strategies of Image and
Video Quality
Assessment,"
PI: Damon Chandler,
Oklahoma State
University; Award
#0917014. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
 Copyright © 2004-2011 Computational Perception and Image Quality Lab. All
Rights Reserved.
 |
|
|
|
|
|
|
|
|
|
|
|
|
