论文部分内容阅读
为了进一步提高H.264中码率控制的性能,提出了自适应预测残差和量化矩阵的码率控制.首先根据宏块梯度方向权值和简化率失真模型,确定是否对宏块进行离散余弦变换(DCT);然后通过新的比特分配模型计算宏块的量化参数,并通过坡度加权确定宏块的坡度系数,最终根据得到的量化参数(QP)和坡度系数计算宏块的量化矩阵,并使用该量化矩阵对宏块进行编码.仿真结果表明,与H.264基本单元层码率控制算法相比,本算法在严格控制编码时间和输出码率的情况下,使重建图像的峰值信噪比(PSNR)值提高0.21dB.
In order to further improve the performance of rate control in H.264, an adaptive prediction residual and the rate control of the quantization matrix are proposed.Firstly, according to the weight of the gradient direction of the macro block and the simplified rate distortion model, it is determined whether to perform discrete cosine (DCT). Secondly, the quantization parameter of the macroblock is calculated by the new bit allocation model, and the gradient coefficient of the macroblock is determined by the slope weighting. Finally, the quantization matrix of the macroblock is calculated according to the obtained quantization parameter (QP) and the gradient coefficient The quantization matrix is used to encode the macroblocks. The simulation results show that compared with the H.264 basic unit layer rate control algorithm, the proposed algorithm can control the peak signal noise of the reconstructed image under the strict control of the coding time and output bit rate (PSNR) value increased by 0.21dB.