Research On Perceptual Quality Assessment For Emerging Multimedia Services

With the rapid development of multimedia technology,emerging multimedia services such as image sharing,image browsing,online video on demand,and 360-degree video on demand have been widely used and gradually become an indispensable part of people's daily life.The user's experience of a service determines the success or failure of a service.Therefore,the research on quality assessment techniques to evaluate the perceived quality of emerging multimedia services is becoming increasingly important.Nevertheless,the user's experience provided by the emerging multimedia services depends not only on the quality of the multimedia itself but also on various factors such as the media display device hosting the multimedia service.How to accurately analyze the relationship between the various factors and the perceptual quality is the difficulties in the research of evaluation of the multimedia perceived quality.For these emerging multimedia services,there is still lack of the relevant researches on the perceived quality assessment of emerging multimedia services.Therefore,this article focuses on the quality assessment methods of three mainstream multimedia services: terminal image browsing in smart phones,HTTP-based dynamic adaptive streaming service,and 360-degree video service in virtual reality scenario.The major contributions of this dissertation are summarized as follows:1.To deal with the issue of the assessment of user perceived image quality on a variety of smart phones,a perceived image quality assessment model is proposed by considering the characteristics of the screen of smart phone.First of all,according to the results of a large number of subjective experiments conducted on 9 smart phones,the characteristics of user perceived image quality are analyzed by considering the size and resolution of the screens of smart phones.Then,a new quality assessment parameter,named as effectively displayed pixels per inch,is proposed to integrate the impact of screen size,screen resolution and image resolution on user perceived quality.Based on this proposed quality assessment factor,an objective assessment model is proposed at last to evaluate the perceived image quality when a user views the image on a smart phone.The experimental results show that the proposed model achieves superior performance in prediction accuracy over other competing objective image quality assessment models.2.To predict the user perceptual quality of HTTP adaptive video streaming in real time,an event-based perceptual quality assessment model is proposed.Based on the type of events presents during the viewing process,the viewing timeline of subjects is divided into rebuffering,playback and quality switching.An event-based perceptual quality assessment framework is introduced which can assess the perceptual quality of the interrupted HTTP-based video streaming at any point in time.To analyze the characteristic of perceptual quality during each type of event,a subjective quality assessment experiment was designed to obtain the perceptual quality at a series of points in time.The between-subjects design was adopted where each subject can view the video content with no repetition.Based on the experimental results,a perceptual quality assessment model was proposed to predict the change of perceptual quality(i.e.,?PQ)after each rebuffering,playback and quality switching.The overall perceptual quality at any point in time is calculated as the summation of ?PQs of all the previous events.The experimental results reveal that the proposed model has a high performance than the international standard P.1203.3 when evaluating the user perceptual quality of HTTP dynamic adaptive streaming service.3.To evaluate the perceptual video quality of the 360-degree video on the Head-Mounted Display(HMD)more accurately,a perceptual video quality assessment model is proposed for 360-degree video by considering the characteristics of the head-mounted display device.The impact of the screen and field of view of the HMD is takes into account,and a new assessment parameter,named as the number of the effective video pixels per degree is proposed to integrate the impact of the screen resolution,field of view and video resolution.Simultaneously,the perceptual video quality on the HMD is found to be linearly related with the video coding quality(VCQ)on the PC monitor.Then,a perceptual video quality assessment model for 360-degree video is proposed based on the mapping formula and the new assessment parameter.Experimental results demonstrate that the proposed model can effectively and conveniently solve the challenge of PVQ assessment for 360-degree video on HMD.4.To evaluating the user experience of 360-degree video in the virtual reality environment,a presence assessment model for 360-degree video is proposed.According to the characteristics of the virtual reality system and the sense of presence,an assessment framework is firstly proposed to evaluate the presence of 360-degree video on the head-mounted display device.This framework has a hierarchical structure of three layers, namely the immersive factor layer,the intermediate quality layer and the presence layer.Based on this assessment framework,a subjective experiment is conducted on the HMD and the presence assessment is further established according to the results of subject experiment.The experimental results show that the proposed presence assessment model can accurately predict the user's presence when user views the 360-degree video on the HMD.