| The rapid development of communication and multimedia technology creates astrong need for better video services. With the high definition video becoming more andmore popular, the ultra high definition video would be the next generation videoapplication in the next few years. On the other hand, the traditional video coding standardsare not efficient enough to handle these ultra high definition videos. So Video CodingExperts Group and Moving Pictures Experts Group established a joint working team JointCollaborative Team on Video Coding to develop the next generation video codingstandard——High Efficiency Video Coding(HEVC). Compared with H.264, the codingeffiency is doubled.HEVC standard is still making process, but in February2012, JCT-VC has completedthe first milestone. The framework of the entire coding standard has been basicallyestablished. On this basis, some of the key technologies for the HEVC—quantizationmatrix and rate control are studied. Quantization matrix is used in the the quantizationprocess, playing a significant role in improving the objective quality of the video. On theother hand, sometimes people require the encoded video stream having a constant bit ratein the actual video applications, so that rate control algorithm is needed.This paper presents a new quantization matrix compression method, usingprediction-based method, such as bi-directional prediction, inter prediction,intra prediction,interpolation to build reference matrix. Experimental results show that using thisalgorithm, the proportion of the quantization matrix part in the final stream can bereduced.This paper also researched on the latest HEVC rate control algorithm, which isnamed as Unified Rate-quantization-based rate control algorithm. The quantization matrixis applied to this algorithm to optimize the video quality. The experiments show that, whenthe quantization matrix is used, the video quality has been improved, the video bit rate isalso maintained in a very stable state. |
PDF Full Text Request