Low-Complexity Encoder Framework for Window-Level Rate Control Optimization

摘要

There exists a tradeoff between visual quality smoothness and buffer constraint in rate control. Thus, a window model and a window-level rate control algorithm were proposed to handle such a tradeoff recently. However, the extra computational complexity and encoding delay were introduced due to a pre-analysis process in the window-level rate control. In this paper, the window model is first improved for a more convenient usage in practical encodings. Second, a new low-complexity encoder framework is proposed based on an open-loop motion estimation (ME). Third, a new memory-efficient algorithm is proposed to eliminate the drift error of open-loop ME. Fourth, the window-level rate control algorithm is improved to adapt to the proposed encoder framework. The experimental results show that the visual quality smoothness is much improved by the proposed algorithm against the traditional one of H.264/AVC. In addition, the improvement of the window-level rate control is better than the previous one in terms of both bit control accuracy and visual quality smoothness. Besides, the proposed encoder could be used to any other rate control algorithms with pre-analysis to reduce computational complexity.