面向硬件的AVS帧内预测算法改进
Improvement of Hardware-Oriented AVS Intra Prediction Algorithm
蔡江震 1吴林煌 1黄贺焜1
作者信息
- 1. 福州大学物理与信息工程学院,福建 福州 350100
- 折叠
摘要
为了面向低延时的浅压缩场景提供更加适配的编码方案,并降低硬件实现成本,提出一种基于数字音视频编解码技术标准(Audio Video coding Standard,AVS)浅压缩算法的帧内预测模式优化以及快速率失真优化算法.该算法通过减少原有算法帧内预测所需的预测循环次数,以及打破各块之间的数据依赖关系等措施,克服了原始方案不适合硬件流水并行处理的限制,提高了编码的效率和稳定性,从而既保障了算法的视频质量,又使新的硬件实现方案更符合实际应用需求.实验结果表明,该算法优化方案能够有效改善实际面向低延时浅压缩场景下的编码效果.
Abstract
In order to provide a more suitable coding scheme for shallow compression scenarios with low delay and reduce the cost of hardware implementation,an algorithm based on the shallow compression algorithm of Audio Video coding Standard(AVS)is proposed,which is based on intra-frame prediction mode optimization and fast rate distortion optimization.By reducing the number of prediction cycles required by the original algorithm and breaking the data dependence relationship between blocks,the algorithm overcomes the limitation that the original scheme is not suitable for hardware pipeline-based parallel processing,and improves the efficiency and stability of the coding,so as to ensure the video quality of the algorithm and make the new hardware implementation scheme more in line with the practical application requirements.The experimental results show that the algorithm optimization scheme can effectively improve the actual coding effect in the low-delay shallow compression scenario.
关键词
数字音视频编解码技术标准(AVS)/浅压缩/帧内预测/快速算法Key words
Audio Video coding Standard(AVS)/shallow compression/intra-frame prediction/fast algorithm引用本文复制引用
出版年
2024
NETL
NSTL
万方数据