| Recently, with the development of digital communication technology and the maturity of image compression technology, and the formulation of relevant international standards and the rise of network, video communication has penetrated into people's daily life and work. Therefore, frequently large number of image needs storage and exchange of video information. This storage and transmission capacity of the existing mechanisms brings new demands and new problems. Under the premise of the limited network bandwidth, digital video compression is particularly important.In March 2003, the Joint Video Team (JVT) officially launched the new generation of video compression coding standard H.264/AVC.H.264/AVC standard support many applications from low bandwidth, high-error-bit rate of mobile wireless communications to high-bit-rate, low delay video broadcasting and streaming media and other online applications, hence is increasingly concerned by the industry. The main objective of video compression coding is to make coding at the minimum distortion under the bit rate restricted conditions. In order to achieve optimal compression efficiency, to ensure the success transmission of encoding data in the channel, to get the best video quality at receiving end, rate-distortion optimization and rate control technology in video coding plays an important role. Video coding technology consists of two parts: the foundation of coding model and the choice of coding model.In this paper, the rate-distortion optimization technology for H.264, coding model and rate control algorithm are deeply studied.Rate control as a key component of video coding algorithm directly affects coding performance. Aiming at specific application choosing a reasonable rate control scheme, is an important research direction in video coding. Many real-time coding systems output constant bit-rate (CBR) video stream. But CBR rate control algorithm has inherent defects, not be able to provide a stable video quality. It is a sacrifice of image quality to obtain constant output rate. Variable bit rate(VBR)rate control algorithm allows bit rate changes in a certain range according to complexity of the scene, so as to access to approximately constant video quality. It can better balance the relationship between video quality and output rate fluctuations.A major advantage of variable bit rate video compression coding technology is significantly reducing the "Filled Bit", greatly increasing the utilization of transmission (or storage media) spectrum (or storage capacity), which have huge economic benefits. But the cost is the technical difficulty of encoder, and the high cost (the amount will not be much, such as radio studio). But adding the receiving equipment (the number of home to millions of units) will not increase the cost too much. Therefore, the application of this technology is especially suitable for real-time broadcasting, which has high performance and low cost.Multiple VBR encoding is frequently used in many VBR coding methods, from the angle of coding quality, which has very high coding efficiency. However, due to its high complex operations, it was impossible to use in real-time coding system. In this paper, the complexity of single VBR rate control algorithm is low. It uses statistical information of previous coding image and characteristics of current coding image to control coding parameters. As the uncertainty of non-coding image signal exists, it will not have comparable effects of coding rate control with multiple VBR encoding. But compared with CBR rate control, constraint condition of the highest rate and the average rate of single VBR rate control is loose. At the same average coding rate, much better results can be achieved than CBR. Therefore, in real-time coding system it has great practical value.This paper is based on an analysis of a single variable bit rate coding rate control advantages, characteristic, elements (such as decoder buffer bound, bit allocation etc.),in view of the H.264 stream characteristics, a low complexity of a single variable bit rate encoding algorithm is proposed, and the experimental result is finally present. Based on R-D model the algorithm predict the image complexity of coding frame you own in DCT using the information of coded frame, and according to the complexity of predicted image and the deviation of the actual bit rate and the target bit rate, DCT coefficients sub-coded bits factor are calculated on four different case, then DCT coefficient target bits are distributed in the GOP layer. Complexity predicted values of the current image are then used to distribute the target bits in frame layer and the buffer underflow is protected. Finally considering the characteristics of the standard H.264 the actual quantitative step is sought. Quantitative parameters are selected to realize H.264 VBR video coding. The algorithm is also on the scene change detection and treatment, in order to get relatively stable image quality at the time of scene change.Finally this paper use H.264 encoder on the proposed algorithm for the test, using experimental test sequence. From an objective evaluation criteria and subjective evaluation criteria of this two aspects, comparing JM10.1 algorithm with the proposed algorithm verify the feasibility of the proposed algorithm.The results showed: compared with the JM10.1 algorithm, this paper presents an algorithm to meet the long-range average bit error rate within the constraints set, to provide a more stable video image quality, better able to adapt to scene changes, and can effectively prevent buffer underflow.The low VBR complexity of the algorithm can be better integrated into various real-time video compression systems, having practical application of certain reference value.This paper only realize a low complexity bit-rate video compression coding, in order to achieve application in a real-time packet exchange system, and in the scope of allowed bit rate variance and under the condition of scene changes to have a relatively smooth video quality, not using more advanced tools of H.264 encoder to achieve variable bit rate video compression coding algorithm. But after further studies, as well as with the development of DSP technology, I believe H.264 encoder can be used to achieve good control of variable bit-rate video compression coding rate. H.264 coding technology will be widely used in various fields. |
PDF Full Text Request