Efficient Cloud-Based Virtual Reality (CloudVR):System Architecture,Optimization And Application

Virtual reality(Virtual Reality,VR)uses physical environments and devices to generate realistic images and sounds,which acts on the user's visual,auditory,and tactile sensations in an all-round manner,thereby creating a variety of immersive experiences for users to control richness and direct interaction,and widely used In the fields of education,medical,military and aviation.But since the development of virtual reality,VR has not opened a new era of user media consumption market as expected.At present,the VR experience is mainly a local VR experience mode(PC-VR experience mode,VR all-in-one machine).Both of these experience modes have certain defects in price universality and high quality portability,thereby reducing the user's desire for experience,Hindering the popularity of VR.The PC-VR system has extremely high demands for computing performance,and it is difficult to open the market of ordinary consumer users.In addition,it restricts the user's mobile space and affects the user experience.The VR all-in-one experience mode is limited by its mobile computing chip and cannot provide high-quality VR images.At the same time,the lack of VR applications is difficult to arouse user interest.Therefore,this paper proposes a cloud-based VR system that combines the advantages of both the PC-VR experience and the VR allin-one machine.The cloudized VR system moves VR rendering to a high-performance cloud server,and users only need to display the server-side rendered image on the client to complete the VR experience.This greatly reduces the user experience threshold,which is conducive to the popularity of VR.The main research contents of this article are as follows:1.Compared with the traditional local VR system architecture,this paper designs the architecture of a new cloud-based VR system and discusses the respective functionsand interaction methods of the components of the cloud-based VR system.The cloudbased VR system mainly calculates,renders and encodes the current VR scene into a video stream on the server side and transmits it to the client through the network.The client renders the received video stream to the user for viewing after decoding.2.On the time scale,analyze the entire workflow of the user experience cloud VR system from initiating an action until the action is truly reflected in front of the user.In addition,the evaluation indexes of cloud VR system are put forward from two aspects: subjective and objective.Respectively describe the measured parameters(resolution,refresh rate,motion perception delay,etc.)in the cloud-based VR system and the subjective experience(realism,interaction,pleasure)in the user experience process.3.In view of the two performances of the cloud VR system's picture quality and interactive experience,this article proposes corresponding optimization strategies from the three parts of the cloud VR system's codec,network transmission and experience optimization: the use of parallel codec based on VR images Relevant coding with left and right eye images to reduce codec delay,optimize codec fluctuations and reduce coding rate.For the same video sequence with a frame rate of 75 Hz,a resolution of2880X1440,and a GOP size of 75,compared with direct encoding,using this strategy can save 9% of the code rate and reduce the time of encoding and decoding by 30%.Customized private protocols of UDP,TCP and forward error correction codes optimize the transmission delay and transmission error tolerance rate of cloud-based VR systems,so that the system can run stably,and design different transmission strategies for different cloud-based VR application scenarios(based on visual Field transmission method and transmission method based on double-layer structure);using the end cloud asynchronous rendering scheme based on user posture prediction and asynchronous time warping greatly reduces the dizziness during the user experience of cloud VR system.4.According to the design of cloud-based VR system and related optimization schemes,this paper implements a cloud-based VR system,which respectively introduces the implementation details from the server,client and communication process of the system,including code rate control and parameter preset,The pixel format conversion and the ”sticky packet” problem in the network transmission process,compared with the local VR experience effect,there is no obvious experience difference in the cloudized VR system implemented in this article.The cloud-based VR system architecture and corresponding optimization strategies designed in this article can provide a user experience comparable to local VR systems.At the same time,it takes into account the extensive application content of the local VR system,reduces the user's VR experience threshold,and effectively meets the user's VR experience requirements.