XSET - Emerging Trends in Image Processing, Computer Vision, and Pattern Recognition - page 81

2 Related Work: JPEG2000 Global Visual Frequency
Weighting
In JPEG2000, only one set of weights is chosen and applied to wavelet coefficients according
to a particular viewing condition (100, 200, or 400 dpi's) with fixed visual weighting [ 1 , Annex
J.8]. This viewing condition may be truncated depending on the stages of embedding, in other
words at low bit rates, the quality of the compressed image is poor and the detailed features of
the image are not available since at a relatively large distance the low frequencies are percep-
tually more important. Table 1 specifies a set of weights which was designed for the luminance
component based on the CSF value at the mid-frequency of each spatial frequency. The view-
ing distance is supposed to be 4000 pixels, corresponding to 10 in. for 400 dpi print or display.
The weight for LL is not included in the table, because it is always 1. Levels 1,2,…,5 denote
the spatial frequency levels in low-to-high-frequency order with three spatial orientations, ho-
rizontal , vertical , and diagonal .

Table 1
Recommended JPEG2000 Frequency ( s ) Weighting for 400 dpi'S ( s = 1 is the Lowest Fre-
quency Wavelet Plane)

s Horizontal Vertical Diagonal

1 1 1 1

2 1 1 0.731 668

3 0.564 344 0.564 344 0.285 968

4 0.179 609 0.179 609 0.043 903

5 0.014 774 0.014 774 0.000 573

3 Image entropy encoding: XSET algorithm
3.1 Perceptual Quantization
(1) Contrast band - pass filtering : The contrast band-pass filtering (CBPF) [ 2 ] is a low-level percep-
tual model of the HVS. It estimates the image perceived by an observer at a distance d just
by modeling the perceptual chromatic induction processes of the HVS. That is, given an im-
age I and an observation distance d , CBPF obtains an estimation of the perceptual image I ρ
that the observer perceives when observing I at distance d . CBPF is based on just three im-
portant stimulus properties: spatial frequency, spatial orientation, and surround contrast.
These three properties allow unifying the chromatic assimilation and contrast phenomena,
as well as some other perceptual processes such as saliency perceptual processes.
The perceptual image I ρ is recovered by weighting these ω s , o wavelet coefficients using the
extended contrast sensitivity function (e-CSF). The e-CSF is an extension of the psychophysical
CSF [ 3 ] considering spatial surround information (denoted by r ), visual frequency (denoted
by v , which is related to spatial frequency by observation distance), and observation dis-
tance ( d ). Perceptual image I ρ can be obtained by