论文部分内容阅读
针对现有的多种视频标准,在总结分解后矩阵的共同特点的基础上,定制实现了一种资源共享结构。该结构通过定制实现处理单元、蝶形变换网络和转置,实现了附带有“阶流水”结构的高效率计算结构。实验结果显示,本文所提出的结构,除了能够在低开销的前提下支持现有的视频标准H.264、视窗媒体视频(VC-1)、音视频编码标准(AVS)和高效视频编码(HEVC)之外,还能够对格式为4:2:0的Full-HD和宽四叉扩展图形阵列(WQXGA)的视频序列进行实时处理。通过在0.13μm工艺下的综合实现,相比于现有的设计,本文所提出的设计逻辑资源开销下降近44%,功耗减小近20%。
Aiming at the existing multi-video standards, a resource sharing structure is customized based on the common characteristics of the decomposed matrix. This structure realizes the high-efficiency computing structure with “order flowing ” structure by customizing the processing unit, butterfly transform network and transpose. The experimental results show that the proposed architecture can support existing video standard H.264, VC-1, AVS and HEVC with low overhead. ), It is also capable of real-time processing of the video sequences of Full-HD and W quad-racket extended graphics arrays (WQXGA) in 4: 2: 0 format. Through the integrated implementation in the 0.13μm process, compared with the existing design, the proposed design logic resource overhead reduced by nearly 44%, power consumption reduced by nearly 20%.