现有的VLSI(verylarge scale integration)视频编码芯片多使用全搜索运动估计(ME)方法,且没有搜索中心偏移(CB)的并行实现方法。本文提出一种适合VLSI的H.264、AVS CB并行搜索方案,减少搜索点数量,降低逻辑资源的消耗,并且使用预测高概率区域的方法,保证ME精度。实验表明,本方法具备较好的率失真性能。在现场可编程门阵列(FPGA)平台上实现了本算法,逻辑综合的数据表明,硬件资源消耗降低了64%。本算法可应用于标清和高清电视(HDTV,hign-definition television)视频编码器。
Most VLSI implementations of video coding adopt full search motion estimation approach,and there is no high efficient center-biased method.A candidates selected search algorithm is proposed which can decrease the resource cost of search.A Center-biased method which is used to predict the high probability area is also proposed.The experiments show that the algorithm owns a good rate-distortion performance.And the algorithm has been implemented on FPGA platform.The synthesis result shows that the cost is decreased by 64%.The algorithm can be used in SDTV and HDTV encoder.