Digital Signal Processing Reference
In-Depth Information
User experience indicators can be considered as QoE (Quality of Experience, QoE).
Sometimes, it can be quantified to reflect the network's quality about the gap between
user's experience and hops. As we all know, for television programs, image eventual-
ly gives users the perception of influence eventually with audio-visual experience.
From the channels between station and end user, source videos are often noised by
different disturbed factors before signal reached terminator [6]. Therefore, image
quality assessment is an important factor in the real practice and the measurement can
be using in communication in order to improve a network's quality [7].
For an image quality assessment (IQA) problem, there are two different categories:
subjective assessment by humans and objective assessment by algorithm automatical-
ly [8]. For human subjective assessment, an algorithm is defined by how well it corre-
lates with human perception of quality. Among the algorithms, automatic “reduced-
reference” (RRED Indices) image quality assessment (QA) algorithms from the point
of view of image information change. Such changes are measured between the refer-
ence- and natural-image approximations of the distorted image. Algorithms that
measure differences between the entropies of wavelet coefficients of reference and
distorted images, as perceived by humans, are designed [9]. Feature similarity (FSIM)
index for full reference IQA is proposed, which is a dimensionless measure of the
significance of a local structure, is used as the primary feature in FSIM [10]. However,
these models are algebra method and limited for video sequence or required more
precise time. Some application is only make a measure between received image and
original one. Meanwhile, real image is often chaotic sequence received or image qual-
ity is low. These methods are invalid for this kind of real application.
The paper is organized as follow. Section 2 gives basic definitions and description
image calculation in high-dimension space based on Gabor filters; section 3 presents
an algorithm of image measurement for coving features in high-dimension space.
Results of different type images from video and measurement with real-noised image
are provided in Section 4, with conclusions in Section 5.
2
Gabor Feature and Model
For wireless application, the signals are transmitted for source to user's destinations.
Among this way, signals are often noised by out-side noise. The way is shown in
figure 1.
Fig. 1.
Wireless video and noise
