首页> 中文期刊>计算机工程 >高速AVS去块滤波器的VLSI实现

高速AVS去块滤波器的VLSI实现

     

摘要

为提高AVS视频编解码系统去块滤波的速度,提出一种基于乒乓并行存储结构及乒乓转置矩阵的环路滤波器.使用2个存储器组存储相邻的2个宏块,在当前宏块滤波过程中,读出另一组存储器中已滤波的宏块并存入新的宏块.通过2个转置矩阵对水平边界进行不间断的行列转换,以保证对水平边界滤波过程中数据的连续性.实验结果证明,该结构处理一个宏块只需122个时钟,在100 MHz的工作频率下能够满足1080p高清编解码的需求.%This paper proposes an Audio and Video coding Standard(AVS) loop filter based on ping-pong parallel storage and ping-pong transposed matrix structure to improve AVS video system deblocking filtering speed. Two sets of memory storage are used for storing two adjacent macroblock. In current macroblock deblocking process, filtered macroblock is read and new macroblock is deposited at the same time in another storage. With two transpose matrix, horizontal boundaries are transposed continuously, which ensures the process of continuous horizontal boundaries filtering. Experimental result shows that the proposed design processing a macroblock only needs 122 clocks, and in 100 MHz working frequency, the design can satisfy the demand for 1 080 p high-definition decoding.

著录项

相似文献

  • 中文文献
  • 外文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号