Research On Key Technologies Of Low Latency Video Coding And Communication System

With the widespread use of video technology in various fields,high-definition real-time video communication has become the focus of the video industry development.Lower latency and better user experience are the core objectives of real-time video communication.Against this background,in order to further reduce the transmission pressure of video streaming media,a new generation of video coding standards such as HEVC has emerged.At the same time,with the full upgrade of the Internet,the demand for video communication based on the Internet is also growing.However,in the Internet,network bandwidth fluctuates greatly and communication quality is uncontrollable.Therefore,how to ensure QoE for real-time communication in the Internet is also an important research topic.Low-latency video coding is a key technology for real-time video applications,which can effectively reduce the delay in video communication.Aiming at the problem of low encoding quality in low-latency intra refresh coding mode,this paper proposes a motion-adaptive intra refresh coding scheme.In traditional intra refresh algorithm,it only relies on the fixed direction refresh strategy to encode the video.The proposed scheme introduces the motion vector information in the video coding,and the different regions of the video frame can freely choose the most suitable refresh strategy.Specifically,in each refresh cycle of the encoding,each band counts the motion vector inverse cost separately in the refresh period,and selects a less costly strategy as an option of the band refresh strategy.At the same time,it is necessary to take into account the correlation between the bands when deciding the best strategy.In this paper,the algorithm procedure and implementation of the scheme are described in detail.Experiments validate the effectiveness and analyze the coding complexity.This paper analyzes some of the key technologies of HD real-time video communication and the shortcomings of currently widely used systems in detail.Most of the existing open source video communication systems use the encoding standard of H.264/AVC,and the compression ratio cannot meet the transmission requirements of HD and Ultra HD video.Therefore,in order to reduce the bandwidth pressure effectively of real-time communication and improve the communication quality,this paper designs and implements a low-latency,high-quality and high-fluency real-time communication system.The system implemented in this paper uses the HEVC coding standard,and the system configures low-latency coding parameters to meet the requirements of low latency.Meanwhile,the existing open source systems perform poorly against network fluctuations and low network bandwidth,and have a large optimization space.This paper designs a simple and effective bitrate adaptive adjustment scheme,which enables the system to dynamically adjust the bit rate of video transmission according to the bandwidth fluctuation of the network.The test proves that the scheme can effectively improve the video communication quality of the system under weak network conditions.