In many video encoding bit rate control models such as TMN8 and RHO domain, the distortion D has been described as a uniform weighted distortion. According to the invention an approximated but still easy-to-use Laplacian distribution model is calculated for the quantisation step size adapta- tion in the macroblock layer bit rate control. The inventive bit rate control uses a pre-analysis step (LPPREA, HPPREA) that calculates the standard deviation of the input signal amplitudes in each macroblock in a Group of Frames. Using the corresponding standard deviation value determined in the pre-analysis step, in a motion estimation/compensation step (CQPEMB) the best mode for the P and B pictures and the ac- tual standard deviation of the prediction residual for the macroblocks are obtained, whereby a rate distortion optimi- sation processing using the Laplace distribution is per- formed. For I frames an intra prediction is carried out by which the best prediction mode and the actual standard de- viation of the prediction residual for the macroblocks are obtained, and a rate distortion optimisation processing us- ing the Laplace distribution is performed. Thereby for I and P frames the quantisation parameter for each macroblock is obtained and used in a corresponding quantisation that fol- lows transform (TRF1) of the prediction or motion estima- tion/compensation residual. For the B frames the bit rate control is carried out once again in an actual coding step, i.e. the quantisation parameter for a current macroblock is computed with the inventive improved TMN8 bit rate control using the best mode and the actual standard deviation obtained from the ME/MC stage.

本发明公开一种视频编码性能评测方法，所述方法包括以下步骤：判定待评测编码器的实际输出比特率与目标比特率的相对误差在特定范围内；获取最大窗口比特率；获取图像质量指标PSNR，即PSNR-Y、PSNR-U、PSNR-V；根据最大窗口比特率与PSNR的值，构建条件率失真曲线图，获取编码器的视频编码性能。本发明可通过条件率失真曲线图，获取不同编码技术的视频压缩性能和对网络传输的适应性能，也可通过在相等PSNR值下，比较编码器最大窗口比特率，获取编码器的网络传输性能优劣，还可通过在相同最大窗口比特率下，比较PSNR的值，获取编码器在图像客观质量上的压缩性能。