高效视频编码的快速编码单元深度遍历选择和早期编码单元裁剪  被引量:10

Fast CU depth range selection and early CU pruning for HEVC

在线阅读下载全文

作  者:蒋刚毅[1,2] 杨小祥[1] 彭宗举[1] 郁梅[1,2] 邵枫[1] 陈芬[1] 

机构地区:[1]宁波大学信息科学与工程学院,浙江宁波315211 [2]南京大学计算机软件新技术国家重点实验室,江苏南京210093

出  处:《光学精密工程》2014年第5期1322-1330,共9页Optics and Precision Engineering

基  金:国家自然科学基金资助项目(No.61171163;No.61271270;No.61271021;No.61311140262;No.U1301257)

摘  要:为了降低新一代高效视频编码(High Efficiency Video Coding,HEVC)标准的编码复杂度,提出了一种基于四叉树结构类型分析和早期编码单元(Coding Unit,CU)裁剪的HEVC快速编码新算法。首先,通过分析已编码的最大编码单元(Largest CU,LCU)四叉树结构类型,确定其深度遍历区间(Depth Range,DR)类型。然后,利用相邻已编码的LCU以及前向参考帧和后向参考帧中坐标位置相同的LCU的DR类型来预测当前LCU的DR类型,并根据预测得到的DR类型对当前LCU设定CU深度遍历区间。最后,采用贝叶斯决策原理获取阈值,并利用该阈值在CU分割过程进行早期CU裁剪。实验结果表明:相对于原始HEVC编码结构,本文算法在随机访问模式下编码时间平均减少41.55%,BDBR(Bjontegaard Delta Bit Rate)只增加约1.94%,BDPSNR(Bjontegaard Delta Peak Signal-to-noise Rate)只降低了0.06dB;与Shen方案相比,该算法可以降低12%左右的计算复杂度,BDBR只增加约1.09%,BDPSNR只降低了0.03dB。To reduce the computational complexity of the emerging High Efficiency Video Coding(HEVC),a novel fast coding algorithm based on quad-tree type analysis and Coding Unit(CU)pruning was proposed.First,the CU Depth Range(DR)of an encoded Largest CU(LCU)was determined by analyzing its quad-tree type.Then,the DR of the current LCU was predicted based on those of neighboring LCUs and the co-located LCUs in the previously coded reference frame which was before or after current frame.The CU depth range of the current LCU could be determined based on the predicted DR.The proposed method also took advantage of Bayesian decision to obtain thresholds,bywhich early CU pruning could be determined.Experimental results show that the proposed method reduces encoding time by 41.55%on average in a random access case.Compared with the original HEVC encoding scheme,Bjontegaard delta bitrate(BDBR)increment is 1.94% and BD-PSNR loss is 0.06 dB.Compared with Shen's algorithm,it saves 12% encoding time with a BDBR increment of 1.09% and a BD-PSNR lossof 0.03dB.

关 键 词:高效视频编码 编码单元(CU) CU深度遍历选择 CU裁剪 四叉树结构类型分析 

分 类 号:TN919.81[电子电信—通信与信息系统]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象