论文部分内容阅读
对视频序列离散余弦变换(DCT)系数统计分布进行了分析,提出了一种基于Weibull概率密度的视频DCT系数统计模型,该模型较Laplace概率密度与Cauchy概率密度更好地描述了视频序列的DCT特征.随后以该统计模型为基础,依据熵编码的理论分别给出了视频序列的速率-量化关系和失真-量化关系,并根据实际视频序列特性对其进行合理地简化近似,得到一种新的较为精确的视频序列率失真模型.经大量仿真实验证明,文中提出的基于Weibull分布的率失真模型对于Intra帧和Inter帧两种编码的真实视频序列率失真特性都具有较好的描述.
The statistical distribution of discrete cosine transform (DCT) coefficients in video sequence is analyzed, and a statistical DCT coefficient model based on Weibull probability density is proposed. This model describes DCT of video sequence better than Laplace probability density and Cauchy probability density Then based on this statistical model, the rate-quantization relationship and distortion-quantization relationship of the video sequence are given respectively based on the theory of entropy coding. According to the characteristics of the actual video sequence, a reasonable approximation is made to obtain a new The rate distortion model based on Weibull distribution proposed in this paper has a good description for the rate distortion characteristics of real video sequences in both Intra and Inter frames coding.