Research On Quality Evaluation Method Of High Dynamic Video

High dynamic range video contains much wider color gamut,higher contrast,and more details,which can accurately restore the light and shadow effects close to the real scene.However,because of the technical level of acquisition,display and transmission,the high dynamic range brightness value often exceeds the capability of capturing,displaying and transmitting in low dynamic range devices,resulting in the loss of visual information and thus affecting the quality of user experience.Therefore,it has become a hot research topic to use effective methods to evaluate the quality of high dynamic range video.Due to the fundamental difference between high dynamic range video signals and low dynamic video signals,the existing low dynamic range video quality evaluation methods can't be directly used to the high dynamic range video quality evaluation.Currently,low dynamic range video quality evaluation algorithms are relatively developed maturely.It is of great significance to study how to apply low dynamic range method to high dynamic range video evaluation.This paper focuses on the subjective and objective quality evaluations of high dynamic video,the main work is as follows:1)Considering the high attention of the human visual system to the information of the edge structure,the movement object and the distortion information,the motion information of the video sequence is extracted by using improved frame difference method,and then the spatio-temporal domain video block containing the motion information and the distortion information is constructed.Combining with the advantage of Haar wavelet transform in extracting edge information,a video quality Assessment model in wavelet domain based on background subtraction(BSWQ)is proposed in this paper,which has better performance and lower complexity,compared with the current international full-reference video quality assessment algorithms.2)In the subjective test,first of all,30 testers scored 49 high dynamic range videos that had been compressed and decompressed.Second,the obtained 1470 subjective data are selected and processed in order to remove outliers.Then,a high dynamic range video quality subjective evaluation test data set with high accuracy is created.3)Using the perceptually uniform(PU)encoding,the BSWQ method is extended to the quality assessment of high dynamic range video.Firstly,the luminance value of a high dynamic range video sequence is processed.Then,the BSWQ method is applied to the coded sequence to evaluate objective quality of high dynamic range video.Compared with the other quality evaluation methods,the proposed algorithm has better consistency with human visual perception quality through testing in high dynamic range video quality subjective evaluation test data set.