Research On Image Quality Evaluation Of The Digital Video System

As the receiver of information, the vision of human acts as an important role which grasps lots of information rapidly and exactly. However, as the mainbody of information, the digital image processing technology has rapidly come to maturity for a couple of years, thus changing our lives dramatically. Infobahn which is under construction step by step has used clear image and video as an important means of information exchange. The digital video transporting technology in the computer field not only is the key to the technology such as infobahn, HDTV, videophone, video conference system, document facsimile, but also contributes much to the fields of aerial spy, remote sensing, exploitation of resource and biomedicine engineering. But with the rapid development of digital video technology, the measure of video system and the research into the methods of evaluation still hang behind. Therefore, it is urgent to probe into the methods of digital video image quality evaluation, which test the efficiency of the structure of system and arithmetic as well as acting as a guide of system optimizing.My paper covers the following points:1. The measure of image quality evaluation which is based on image optical characters and vision perception is set up. According to JND, we discard optical characters and details which is beyond human vision, so the objective evaluation is in accordance with human subjective perception.2. The experiments on luminance, hue and saturation are designed for threshold of contrast sensitivity. Besides, the general conditions of experiments and procedures are offered. As a result, the experiment can be repeated.3. In this paper a nev way of image sub-band division is proposed, namely, the employment of DCT and Zig-Zag is used to accelerate the division of image. Sowe can divide image sub-band at will. Finally, the process of studying human vision perception in agreement with different frequency and vision weight value proves to be more appropriate.4. By the definition of chroma in television principle, the mathematic model of signal based on basic chroma coding is established. The method concerning luminance, hue and saturation is also presented. The calculating method proves simple but solves the designing problem of signal which is limited between 0 and 1. It meets the demand of different luminance, hue and saturation. Moreover, it is conducive to the design and development of new digital video testing signal.5. By HVS, WSPSNR about the gray image is defined. Through the transformation of color space and the use of the up-to-date formula of color difference, the evaluating index such as SCAPSNR of the colorful images is defined. Clarity and continuity are used to evaluate the quality of motional images.6. We create VC systematic program in Windows in order to inspire the further studies and demonstration. In such a way the satisfying results can be achieved through plenty of experiments and simulations.