A Video Quality Synthetic Evaluation Method Based On Motion Masking

Video not only includes images information itself, but also includes rich motion information, so the research of video quality evaluation methods possess more theory discussed and applied value. With the development of science, kinds of video become more and more, and the usage of video becomes more and more extensive, the traditional assessment methods cannot afford to various video quality assessment works. Therefore, now a large number of scholars at home and abroad set about research the movement information and the video quality assessment method based on motion, also made many good results. The basic idea of these assessment methods is to measure and estimate the motion of damage video, and get a motion model. Then to consider the human visual system (HVS) characteristics, and combined it with other methods. But if the motion in video foreground and background are complex, the analysis of movements often won’t be so precise, and the final results of model also won’t favorable. In addition, the HVS is a rather complex and nonlinear systems, so it is not an easy thing to simulation HVS. Therefore, the research of video quality needs to continue. As a whole, because subjective method in practice has certain difficulty, the current video quality assessment method has mainly focused on objective assessment, the ultimate goal is to build an objective quality assessment model which is coincide with subjective results.This paper mainly research the video quality assessment method based on the motion masking. First, we take example by the antecedent achievements. Then, we combine the motion masking and the space similarity model. The main research contents are as follows:1. In order to analyze the masking properties which movement parameters put on the visual perception, we design a kind of video test signal according to the signal amplitude, frequency and the movement parameters. This method can avoid the influence of various parameters on visual perception, and make for get single parameters changes to the inspection of the image quality assessment.2. Based on the measurement method of the motion information in video, according to the relationship between visual perception and movement parameters, analyzes the basic model of motion masking properties, and gives the corresponding block diagram and the experimental method. At last, get the motion masking evaluation results which is consistent with the visual perception.3. Combine motion masking with spatial similarity evaluation method, establish a comprehensive system model and platform. The experimental results show that the new evaluation algorithm is obviously superior to the traditional objective evaluation method and the evaluation method based on masking motion, and further coincide with the eye’s subjective feeling.