视频编码中自适应的初始量化参数预测算法

Adaptive initial QP prediction algorithmin vedio coding

通过分析在H.264/AVC中预测初始量化参数的方法过于简单,导致视频序列整体编码量化参数和重建图像PSNR的波动,甚至PSNR急剧下降的问题,利用基于Cauchy分布的R-Q模型,结合I帧图像复杂度建立R-C-Q模型,提出了自适应的初始量化参数预测算法。该算法能够根据不同视频序列的特性,准确地预测I帧的初始量化参数,有效地抑制视频序列的整体量化参数及PSNR的波动,从而提高码率控制精度和重建图像质量。

Based on the analysis of initialization scheme in H.264/AVC, which is so simple that makes fluctuations of coding QPsand reconstructed image PSNR, even video sequence PSNR reduced sharply,an adaptive initial QP prediction algorithm is proposed.The Cauchy distribution-based R-Q model combined with the complexity of I-frameis applied to modeling R-C-Q relations.The proposed algorithm can accurately predict the I-frame initial QP according to the characteristics of different video sequences,which effectively inhibits the fluctuations of overall QPs and PSNR, and improves the accuracy of rate control and the quality of reconstruction image.