**5. Objective quality assessment**

In this section, the most important metrics are described to make to offer an overview of the techniques that develop this type of quality assessment.

There are three types of objective quality assessment depending on the presence and availability of a reference image or any of its features to develop the study: Full-Reference (FR), Reduced-Reference (RR) and No-Reference).

Old metrics designed for digital imaging systems, such as MSE (Mean Sqaured Error) and PSNR (Peak Signal-to-Noise Ratio), which are still very used to develop quality assessment, are defined in this section. They are still adequate for evaluating error measures

The analysis of measurement of artifacts such as tiling or blurring, especially introduced by video compression algorithms, will be interesting to offer the reader a perspective of evolution of metrics.

### **5.1 Objective quality metrics**

Depending on the presence of a video reference, three kinds of analysis are defined:

 Full Reference (FR) metrics, when the original image is present and can be used to compare it to the degraded image in order to obtain the reduction of quality because of the process of encoding and decoding.


Video Quality Assessment 143

As SNR used the same signal to compare with, it is more difficult to export the conclusion from one study to another, that is why the original signal was changed by the peak value (255 in a RGB channel, or 240 in luminance, for example) to obtain more general results,

 

*X Y*

*Max*

*i i*

2

[dB]

1, 1

*M N*

*i j M N*

 

0, 0 <sup>10</sup> 1, 1 <sup>2</sup>

0, 0

*i j*

A collection of metrics attempt to assess the effects of artifacts described on section 0, instead of offering a global idea of quality, as with pixel-based metrics. Some examples of

The metric defined by MSU Graphics & Media Lab measures subjective blocking effect in video sequence, based on energy calculation and gradients. In contrast areas of the frame

Other metrics are based on the structure of the pixelized image. The model included in Lee et al. research, first extracts edge pixels and computes horizontal ( H(t,i, j) ) and vertical (V (t,i, j) ) gradient component of the edge pixels. The gradient is calculated using the Sobel operators. From the horizontal and vertical gradient images, the magnitude (R) and angle

2 2 *Rti* (,,) (,,) (,,) *j Hti j Vti j*

<sup>1</sup> (,,) ( , , ) tan (,,) *Vti <sup>j</sup> tij Hti <sup>j</sup>*

Analyzing the angles of the gradient, the pixels with gradient parallel to the picture frame are considered as belonging to blocking region, if they have a determined magnitude.

Other interesting metric on this field are the ones by Winkler et al., 2001 and Wang et al., 2002.

Blurring metrics are based on the analysis of energy in high frequencies and analysis of edges and their spread. As it proposes Marziliano et al. in 2004. The reduction of edge energy between the original and the impaired image shows the loss of quality due to blurring artifact. Other metrics, such as proposed in Lee et al. Research, utilizes the gradient calculated in every pixel of the image to detect the blur artifact by analyzing the diminution of this

10log

blocking is not appreciable, but in smooth areas these edges are conspicuous.

Comparing to the original image, errors are avoided due to real edge pixels.

*PSNR*

most representative metrics are introduced next.

**PSNR (Peak Single-To-Noise Ratio)** 

with a more efficient method.

Based on artifacts

**Blockiness or tiling** 

() are extracted:

**Blurring** 

are of vital importance in environments which are difficult to provide any reference, such as mobile or internet multimedia services.

Fig. 13. No Reference (NR) metric diagram
