论文部分内容阅读
在可分级视频编码(SVC,scalable video coding)的框架下,利用分布式视频编码(DVC,distributed video coding)技术,设计了一种低编码复杂度的SVC方案。该系统具有空间可分级的特性,各分层中仅用到了传统的帧内编码技术和DVC技术,最大限度的减小了SVC系统的编码复杂度。在该系统中,充分利用分级系统的特点,在增强层(EL)的解码中提出了一种基于二次搜索和残差补偿(DSRCB)的边信息(SI)生成算法和一种基于时空域虚拟噪声模型的估计算法,并针对各分层图像的频域特性优化了量化模型。实验表明,与基于传统视频编码技术的SVC系统相比,该系统具有极低的复杂度,性能超过了非可分级的DVC系统,而且在较小的GOP(group of pictures)尺寸下获得了接近传统SVC系统的性能。
In the framework of scalable video coding (SVC), a low coding complexity SVC scheme is designed by using distributed video coding (DVC) technology. The system has the characteristics of spatial scalability, and only uses the traditional intra-coding technology and DVC technology in each layer to minimize the coding complexity of the SVC system. In this system, taking full advantage of the characteristics of hierarchical system, an edge information (SI) generation algorithm based on quadratic search and residual error compensation (DSRCB) and an algorithm based on spatio-temporal domain Virtual noise model estimation algorithm, and optimize the quantization model for the frequency-domain characteristics of each hierarchical image. Experiments show that compared with SVC based on traditional video coding technology, this system has very low complexity and outperforms non-scalable DVC system and gets closer to the smaller GOP size Traditional SVC system performance.