基于SVC空域/质量域宏块级码率控制
首发时间:2011-11-22
摘要:可伸缩视频编码 (SVC) 的发展满足了互联网视频、移动无线视频、视频监控、视频电话以及实时转播等不同视频服务的多样化需求。为了适应实际信道带宽的限制同时又能重建高质量的视频图像,码率控制(RC)发挥了重要的作用。本文在SVC基础层码率控制的基础上,提出一种基于层级和宏块级的目标比特分配方案,每一层的编码取决于新定义的缓冲器和RC控制参数,同时提出了新的空域/质量域MAD预测模型。实验结果表明,文中算法能够有效地控制各层码率,使得目标码率跟实际产生码率之间的偏差最大不超过±2%;同时文中算法提高了解码图像的质量。与JSVM自带工具FixQpEncoder相比,不需要多次迭代,降低复杂度的同时能够满足实时性需求。
For information in English, please click here
Macroblock Level Rate Control of SVC Spatial-SNR Scalability
Abstract:Scalable video coding (SVC) addresses the application need of a more fexible format of coded video in heterogeneous and time-varying environments. Rate control (RC) plays an important role in allocating H.264/SVC bits to adapt to the actual channel bandwidth and meanwhile to achieve high-quality reconstructed images. A macroblock level rate control algorithm based on layer level and MB level bit-rate allocation is proposed for SVC spatial and MGS scalability in this paper. At the same time, layer level buffer control is used. The coding process of each layer is controlled by a respective buffer controller and a train of RC parameters. The accurate bit-rate can be achieved and the maximum mismatch between actual bit-rate and target bit-rate can be maintained, using the proposed algorithm. Comparing with FixQpEncoder adopted in the joint scalable video model (JSVM), high peak signal to noise ratio (PSNR) can be obtained without multiple iterative and the complexity is reduced. The proposed algorithm can meet the real-time demand.
Keywords: Scalable video coding (SVC) Rate control (RC) Bit-rate allocation MAD prediction
基金:
论文图表:
引用
No.****
同行评议
勘误表
基于SVC空域/质量域宏块级码率控制
评论
全部评论0/1000