分布式小波视频编码自适应嵌套量化模型

Adaptive Nested Scalar Quantization Model for Distributed Wavelet Video Coding

提出一种分布式小波视频编码自适应嵌套量化模型,对Wyner-Ziv帧做小波变换,利用Slepian-Wolf编码器对低频子带系数进行量化,采用SPIHT编码器对高频子带系数进行编码,通过分析Wyner-Ziv帧低频子带系数与参考帧低频子带系数的残差自适应度,判定Wyner-Ziv帧低频子带系数的量化步长,从而使运动剧烈的视频序列有较大的量化步长,而运动平缓序列的量化步长较小。实验结果表明,采用该模型能获得较好的编码效果。

An adaptive nested scalar quantization model for distributed wavelet video coding is proposed, which conducts wavelet transformation to Wyner-Ziv frame, and uses Slepian-Wolf coder to quantificate the low frequency subband coefficient while using SPIHT coder to program high frequency ones. By analyzing the error adaptive degree between low frequency subband coefficient and reference frame ones of Wyner-Ziv frame, the quantization step length of it is determined, which makes active video sequence have bigger step lenth, and the placid one have less step length. Experimental results show this model can obtain better coding effect.