AVS视频解码器中运动矢量预测的硬件设计和实现  

Hardware Design and Implementation of Motion Vector Prediction in AVS Decoder

作　　者：温海波[1] 潘良华[1] 

机构地区：[1]同济大学通信软件及专用集成电路设计中心,上海200092

出　　处：《自动化信息》2008年第5期52-53,48,共3页Automation Information

摘　　要：AVS是我国具有自主知识产权的音视频编码国家标准。本文介绍了基于该标准的视频解码器中运动矢量预测模块的硬件设计与实现。根据宏块类型将运动矢量预测的计算方式分成三大类，采用流水线加并行的处理方式，在满足速度指标的要求下有效地减小了逻辑电路结构的面积，且完全满足高清实时解码要求。AVS, with independent intellectual property rights, is the national audio and video coding standards. This paper introduces the design and implementation of the hardware in motion vector prediction module in video decoder based on AVS. The calculation method of motion vector prediction is divided into three categories according to the macro-block types, pipeline and parallel structures are adopted, in order to meet the speed required on condition that the area of logic circuit architecture is reduced and realize the high definition real-time decoding.

关 键 词：视频解码 AVS 运动矢量预测 FPGA 

分 类 号：TN919.81[电子电信—通信与信息系统]