Research On Video And Image Objective Quality Assessment Algorithm Based On Structural Information

With the progress of the era and the development of science and technology, the image as an important carrier of information, has an important role in our daily life. Also, people gradually improve to the requirement of image quality. Because the distortion exists in the process of image, it leads to a decline of image quality which brings inconvenience to the subsequent image processing and analysis and finally will influence visual experience. So, it has become one of the key technologies in the field of image processing.Image quality assessment methods include two kinds. One is the subjective evaluation method and the other is objective evaluation method. Because the person is the ultimate viewers of the image, the subjective image quality assessment method, which is depended on subjective perception, is the most accurate method. But it has many shortcomings such as high cost, long cycle, and real-time. These methods are not convenient to be used in real life application. Because objective image quality evaluation method has many advantages such as timeliness strong, less cost, application field widely, not influenced by the subjective condition, it attracts attention to by the experts and scholars. So, it is a research hotspot in the field of image processing. In this project, objective image and video quality assessment methods are researched. And the research content and innovation are as follows:First, this article has systematically illustrated the methods of image quality assessment: subjective assessment methods and objective assessment methods. We analyzed all the factors which affect the human perception system from the physiology and psychology of human visual system, such as brightness, contrast, mask, visual interested areas, movement and other important information.Second, to the hot research subject of stereo image, this article has proposed an image quality assessment method which based on the structure information and in-depth information. In the method, image segmentation has been adopted, then layered the depth image based on the statistical property of the histogram, and then summarized the gradient information of the image, extracted the structure characteristics and established the quality evaluation algorithm model of the stereo image based on the human visual system perception sensitivity. The experimental results show that the proposed metric is highly consistent with the subjective test scores compared with the existing related metrics.Finally, this article has also studied the 2D video quality assessment algorithm, the biggest difference between video and image is that movement in the video scenarios. This article has analyzed the movement in different scenarios, introduced the typical human visual attention model and extracted the motion vector in the video, and then proposed the video quality assessment method based on the structure information and motion information, tested and analyzed by the experiment, the method mentioned in this article is better than any other ones. By studying of the 2D video, we have mastered the technology of video quality assessment. It has also laid the foundation for the further study of stereo video quality assessment algorithm in the future.