H.264低码率视频应用的率失真模型及码率控制

Rate distortion model and rate control for H.264 low bit-rate applications

提出了一种通用的率失真(RD,rate distortion)模型,进而提出纹理复杂度(TC,texture-complexity)和图像块头用字节自适应线性预测模型及宏块(组)比特分配模型。利用以上模型,构建了一种面向低码率视频应用的码率控制方案。大量仿真实验结果表明,利用新提出的码率控制方法,测试序列的峰值信噪比(PSNR)均有不同程度提高,其码流曲线和PSNR曲线更平滑平稳,目标码率控制更准确。

A general rate-distortion model is proposed.On the basis,the optimum bit allocation model and two adaptive linear predicative models:texture complexity and overhead bits count predicative models are presented.Based on above models,a new rate control scheme is proposed for low bit-rate and real-time video coding applications.The simulation results show that the new rate control algorithm provides smoother bit stream,higher video quality and more accurate rate control than the previous algorithms.