Research On The Efficient Coding Algorithm Of Game Video

With the increasing rise of the e-sports industry,the traffic of game video in the network has increased rapidly,which brings great challenges to its coding.However,the game video is different from the traditional video recorded by the camera,which shows the unnatural nature in the video content.The traditional video coding method can not meet its coding characteristics.In order to improve the coding performance,researchers began to consider using the commonness of unnatural video and the characteristics of game video in game video coding.However,the video coding has attracted the attention of researchers in recent years.The mining of its characteristics is not enough,and the coding performance still has room for improvement.In order to solve the above problems,this thesis studies the recurring characteristic of game video in depth,and proposes an efficient video coding optimization scheme based on scene adaptation.The specific research contents are as follows:(1)This thesis studies the recurring characteristic of game video,and proposes an efficient coding optimization strategy.First,from the change of game video content,the thesis studies the recurring characteristic of content and associates the recurring characteristic with video coding,analyzes the impact of the recurring characteristic on the coding rate.Then,based on the above analysis,the thesis proposes an efficient coding optimization strategy applicable to game video,gives the overall framework of the strategy,and details the implementation principle of the strategy.(2)This thesis analyzes the actual coding scene,and proposes and implements the scene adaptive coding optimization algorithm,further improves the coding performance of game video in the practical application.In order to ensure the performance of the algorithm,first,the thesis analyzes the influence of different scene characteristics on the strategy in the actual coding,and determines the main factors are video resolution and recurrence degree.Then,aiming at the scenes with different resolution and recurrence degree,the original similarity index is optimized,and structural similarity,hash value similarity and hash value matrix similarity are applied to different scenes respectively.Finally,the indexes suitable for different coding scenarios are integrated into the same algorithm,and the adaptive coding optimization algorithm is proposed,and the specific implementation flow of the algorithm is given.(3)Algorithm performance test and analysis.After the implementation of the proposed algorithm based on the Versatile Video Coding(VVC),in order to verify the performance of the proposed algorithm,first,the thesis tests the performance of the optimization algorithm using only one similarity index,and analyzes the advantages of the three similarity indexes in different coding scenarios.Then,based on the above test results,the thesis tests the performance of the scene adaptive optimization algorithm integrating three similarity indexes,and analyzes the advantages of the adaptive algorithm compared with the algorithm using one similarity index.Experimental results show that,compared with VVC standard encoder,the proposed adaptive scheme achieves 0.283 d B optimization on BDPSNR,1.0233%optimization on TS and 4.7106% optimization on BDBR.The proposed adaptive optimization scheme takes into account the recurring characteristic of game video,and uses the characteristic to improve the video coding performance,which has a certain reference value for related scientific research and industrial applications.