Research On Temporal And Spatial Quality Consistency-Aware 360-degree Video Rate Adaptive Transmission Method

With the great achievement of information technology,the requirement of video applications is changing from traditional two-dimensional videos to immersive videos.As a typical kind of immersive video,360-degree video is becoming popular because it allows users to watch the scene from multi-view by switching their viewports.However,the huge data volume of the 360-degree video limits its applications in network transmission.To provide high quality of experience(QoE)for remote web users,this paper presents a temporal and spatial consistency-aware bitrate adaptation method for 360-degree videos based on MPEG dynamic adaptive streaming over HTTP(MPEG-DASH).The method consists of four parts:bitrate request,field of view(FoV)prediction,tile priority determination,and bitrate allocation.First,we propose a buffer-quality-based bitrate request algorithm(BQA)to balance the buffer length and video quality for the current pre-fetched video segment.Second,the Gaussian model is adopted to predict the user's FoV at the beginning of each pre-fetched video segment to estimate the track of the user's viewport during video playback.Afterwards,for possible FoV switching during video playback,we propose to use Zipf model to calculate priorities of different tiles in the 360-degree video.To obtain the bitrates of all tiles in each video segment,we propose a two-stage coarse to fine optimization algorithm.In the first stage,coarse bitrates for all tiles are allocated to optimize the overall video quality.In the second stage,we fine-tune the bitrate of each tile so as to improve the temporal and spatial quality consistency while guaranteeing the quality of the tiles in FoV.Finally,a video transmission platform based on Libdash is built and the performance of the proposed algorithm is verified by comparing with other state-of-the-art bitrate request algorithms and allocation algorithms.Experimental results demonstrate that the proposed method can reduce the video playback interruption,preserve the temporal-spatial quality consistency of tiles in FoV,and thus improve the user QoE effectively.