Research On Macroblock Codec Algorithm Based On H.264

In recent years,more and more video contents have been encoded by the H.264/AVC standard.In practical communication applications,it is required to perform local content re-editing processing on a compressed coded video stream at a transmitting node,such as inserting a logo to protect copyright or labeling station information.The traditional transcoding way is to decode the video bit stream completely and re-encode the reconstructed signal.Although easily to achieved,the Full Decode and Recode process is time consuming and filled with huge complexity.Otherwise,it may cause the loss of the overall video quality due to re-quantization.The local content re-editing method achieves the purpose of local content modification by only encoding the macroblocks relevant to logo insertion.While ensuring the quality of the video,it reduces the amount of calculation and system complexity.Therefore,the local content re-editing method is a research focus in the field of the station logo subtitle system.This thesis proposed an effective logo insertion scheme based on macroblock-level local processing,which is for the video stream of the H.264/AVC standard.This scheme not only overcomes various dependencies that exist in video-encoded data,reduces the amount of computation,but also ensures the overall quality of the image by maximizing the use of the original video stream data.At the same time,in order to ensure that the re-edited stream can be effectively transmitted on the original bandwidth,a corresponding rate control method is proposed in this thesis.The main works of this thesis can be presented as follows:First,this thesis introduces the basic frame structure,codec processing flow and key technologies of the H.264 standard in detail.In order to make the recombined bit stream decode correctly,this thesis analyzes the dependencies of video content and coding parameters as well as the performance of drift-error propagation during the transcoding architecture.In addition,this thesis introduces several commonly used transcoding system structures and classical rate control algorithms.Then,this thesis designs the overall system scheme of the station logo superposition technology based on the video local macroblock editing,which mainly includes the key technologies of macroblock type division,station logo superposition processing,the isolation of pixel and encoding parameter drift-error.In order to track target coded bits accurately,this thesis introduces the ? domain source model to the H.264 standard.Based on information of the original stream,this thesis analyzes the establishment of the ? domain source model relationship in detail,and formulates a rate control scheme in this system.Finally,on the MFC platform of visual studio2017,the system simulation is designed to verify the effectiveness of the parameter setting algorithm of interface MB,test the performance of the rate control scheme and evaluate the output image quality as well as the algorithm efficiency.The experimental results show that compared with the traditional "full decode and re-encode" logo insertion scheme,the proposed scheme improves the algorithm efficiency by about 120%,and improves the signal-to-noise ratio of the output image quality by 1.73 dB.Due to the deviation of the bit rate between the edited bit stream and the source video stream within 3%,the editing code stream can well adapt to the adaptive requirements of network transmission bandwidth.It has very important practical value in the application of real-time communication video.