Research And Implementation Of Video Coding Rate Control Algorithm Based On Fuzzy Logic

With the development of demand that higher resolution and clear video, In the case of storage space and network bandwidth is still limited, the rate control in video compression is particularly important.How to provide users with the best possible visual quality under the condition of network communication has become a research topic for a long time.This thesis focuses on the rate control principle, discussed the rate distortion theory and classical model, systematic analysis the basic theory and working principle of classical rate control algorithm in traditional video coding standard, and emphatically discussed the rate control algorithm in mainstream video coding standard HEVC and H.264/AVC. On the basis of summarizing and learning the classical rate control algorithm, a new rate control algorithm is proposed. The new algorithm includes a constant bit rate control algorithm, the initial QP automatic setting algorithm, variable bit rate control algorithm, adaptive GOP length adjustment algorithm. The main contents of this thesis are the following four aspects:(1) Established the fuzzy control model between buffer error, error of buffer error and video coding quantization parameters. On the basis of this, the relationship between the bit rate and the error is established, and the analysis of the error range of the buffer is given. Finally, a rate control algorithm based on fuzzy logic is proposed.(2) Modeling and analysis the linear relationship between the bit rate R, the quantization parameter Q and image gradient G. established the model of RQG and RQ by linear regression statistics, and proposed a new algorithm of automatic setting of initial QP. The new algorithm can better set the appropriate rate control of the initial QP, improve the rate control performance.(3) deeply studied the motion process and motion compensation technology and defined the image intensity and its calculation method. Proposed a variable bit rate control algorithm based on fuzzy logic rate control. The algorithm can satisfy the human perception of the desired video quality at the same time as high as possible to save the bit rate.(4) Analysis and discussion of the characteristics of the hierarchical coding structure,a statistical algorithm based on the motion characteristics of the adaptive GOP length adjustment algorithm is proposed. Experiments show that, compared with the fixed GOP length of the inter frame predictive coding, adaptive GOP length adjustment algorithm can better eliminate the temporal redundancy of the image and improve the coding performance.The above four aspects work meets the application requirements of the rate control in different scenarios and shows a good performance of the algorithm. In this thesis, the frame level, GOP rate control algorithm and the initial QP automatic setting algorithm have been adopted and integrated into the RD reference model by the AVS group as the only rate control algorithm.