最优对称性的快速自适应环路滤波算法  

Fast adaptive loop filter algorithm with optimal symmetry

作　　者：李宏伟[1] 吴成柯[1] 张捷[1] 宋锐[1] 

机构地区：[1]西安电子科技大学综合业务网理论及关键技术国家重点实验室,西安710071

出　　处：《控制与决策》2012年第7期1087-1091,共5页Control and Decision

基　　金：国家自然科学基金项目(60802076);高等学校创新引智计划项目(B08038);中央高校基本科研业务费专项基金项目(72115046)

摘　　要：为了进一步提高基于块与四叉树的自适应环路滤波算法(BQ ALF)的编码性能,提出了最优对称性的快速自适应环路滤波算法(FALF OS).首先根据区域对称性和平均绝对差值门限确定最优的对称性滤波器;然后在I帧中按照基于块与四叉树的方法得到需要滤波的区域,在P或B帧中利用运动矢量、运动估计和率失真模型A(RDOA)得到需要滤波的区域;最后采用得到的最优对称性滤波器对这些区域进行滤波.仿真结果表明,与BQ ALF算法相比,所提出的算法在保证重建图像质量的同时,编码时间大幅度下降.In order to further improve the coding performance of block,based and quadtree-based adaptive loop filter（BQ_ALF） algorithm, fast adaptive loop filter algorithm with the optimal symmetry（FALF_OS） is proposed. Firstly, this algorithm determines the optimal symmetry filter according to area symmetry and average sum of absolute difference. Then the filter areas are obtained by using the block-based and quadtree-based method in I frame, and the filter areas are obtained through motion vector, motion estimation and rate distortion optimization model A（RDOA） in P or B frame. Finally, the filter areas are filtered by the optimal symmetry filter. Simulation results show that compared with BQ_ALF algorithm, the proposed algorithm reduces the coding time greatly, mean while, the reconstructed picture quality it kept.

关 键 词：对称性 环路滤波 运动矢量 率失真 

分 类 号：TN919.81[电子电信—通信与信息系统]