论文部分内容阅读
针对AVS-M帧间编码模式的复杂性,在对帧间预测编码最佳模式选择统计结果分析的基础上,提出了一种模式分类的帧间预测模式快速选择算法.该算法将全部编码模式进行分类,对SKIP模式采用提前终止判决,而对其它模式通过判别宏块的一致性和比较宏块的空时相关性逐类型比较.实验表明,与全模式选择算法相比,该算法的计算次数减小了52.5%~85.8%,在保持编码效率和重建图像质量的同时提高了编码速度.
Aiming at the complexity of AVS-M inter-frame coding mode, based on the analysis of the statistical results of the best mode selection for inter-frame predictive coding, a pattern classification fast interframe prediction mode selection algorithm is proposed, which combines all coding modes The classification of the SKIP model using early termination of the decision, and other models by identifying the consistency of the macroblock and compared macroscopic space-time correlation by the type of comparison.Experiments show that compared with the full-mode selection algorithm, the calculation of the algorithm The number of times is reduced by 52.5% -85.8%, which improves the coding speed while maintaining the coding efficiency and reconstructing the image quality.