点评：一种用于视频图像编码过程中的量化方法，残差块信号Dn经整数变换的核心变换后，其输出信号W再经系数调整与量化处理|Zij|＝(|Wij|·MFij[Qm]+f)＞＞qbits，sign(Zij)＝sign(Wij)，得到量化输出信号Z，特点是在量化处理的过程中，首先变换量化器乘法因子，使MFij[Qm]＝MF′ij[Qm]·2n，n为1到13之间的整数，根据四舍五入取整数的原则确定MF′ij[Qm]的值，然后用加法器对变换输出信号W进行系数调整与量化处理，与原H.264量化方法相比，优点在于以率失真性能的微小下降为代价，换取量化过程复杂度和功耗的大幅降低，软件算法实现时可节省92%以上运算时间；硬件电路实现时，在相同CMOS工艺下，各量化单元资源面积和功耗平均节省75.2%和76.3%；绝对平均率失真误差小于0.0315dB。

Comment: A quantization method used in video image coding course obtains quantization output signal Z by carrying out coefficient regulation and quantization treatment of |Zij|＝(|Wij|•MFij[Qm]+f)＞＞qbits，sign(Zij)＝sign(Wij) on output signal W of residual error block signal Dn after said signal Dn is core coverted with integer transform . In the method , quantizer multiplication factor is converted first to let MFij[Qm]＝MF′ij[Qm]·2n be n of integer between 1 to 13 , then adder is used to carry out coefficient regulation and quantization treatment on conversion output signal W .