为了降低下一代通用视频编码(VVC)帧内预测编码单元(CU)划分的计算复杂度,提出一种基于梯度幅值相似度的CU快速划分方法.首先,计算当前编码单元下层的四个子编码单元的平均梯度幅值相似度偏差(M GM SD),根据该信息来确定当前编码单元是...为了降低下一代通用视频编码(VVC)帧内预测编码单元(CU)划分的计算复杂度,提出一种基于梯度幅值相似度的CU快速划分方法.首先,计算当前编码单元下层的四个子编码单元的平均梯度幅值相似度偏差(M GM SD),根据该信息来确定当前编码单元是否进行四叉树划分或不划分.其次,当不满足四叉树划分和不划分的条件时,通过遍历得到三叉树划分和二叉树划分的子块像素方差的方差,根据该信息来选择二叉树和三叉树中最佳的划分方式.在全I帧条件下,本文方法与VTM7.0(VVC Test Model 7.0)标准模型相比,编码时长平均降低了50.69%,在大幅降低编码复杂度的同时码率仅增加1.36%.展开更多
The Gutenberg-Richter law (G-R law) of the magnitude-frequency distribution of earthquakes has been an important base in seismology for a long time. However, the actual magnitude-frequency distribution usually deviate...The Gutenberg-Richter law (G-R law) of the magnitude-frequency distribution of earthquakes has been an important base in seismology for a long time. However, the actual magnitude-frequency distribution usually deviates from the G-R law. Based on the experimental results of three different rock samples, which contain macro-asperity, compressional en-echelon fault and model-Ⅲ shear fault, the correlative coefficient (r) was calculated by fitting the sequences of the acoustic emissions with the G-R law in every scanning time window. We investigated the changes of the correlative coefficient, which describes the actual deviation of magnitude-frequency distribution of earthquakes from the G-R law within the specific scanning time window. According to the results of the rock samples containing macro-asperity and compressional en-echelon fault respectively, the value decreases prior to the rock fracture, meaning that the deviation of magnitude-frequency distribution from the G-R law tends to be larger. The result of the model-Ⅲ shear fault didn’t show obvious decrease before the final rock fracture. Actually, the studies of some earthquakes also show deviation before the occurrence of moderate earthquakes. The results obtained in this paper will provide us with some useful clues for studying precursors before the occurrence of moderate earthquakes with the data of regional earthquake activities.展开更多
文摘为了降低下一代通用视频编码(VVC)帧内预测编码单元(CU)划分的计算复杂度,提出一种基于梯度幅值相似度的CU快速划分方法.首先,计算当前编码单元下层的四个子编码单元的平均梯度幅值相似度偏差(M GM SD),根据该信息来确定当前编码单元是否进行四叉树划分或不划分.其次,当不满足四叉树划分和不划分的条件时,通过遍历得到三叉树划分和二叉树划分的子块像素方差的方差,根据该信息来选择二叉树和三叉树中最佳的划分方式.在全I帧条件下,本文方法与VTM7.0(VVC Test Model 7.0)标准模型相比,编码时长平均降低了50.69%,在大幅降低编码复杂度的同时码率仅增加1.36%.
基金This research was sponsored by the Joint EarthquakeScience Foundation,China (A07007)
文摘The Gutenberg-Richter law (G-R law) of the magnitude-frequency distribution of earthquakes has been an important base in seismology for a long time. However, the actual magnitude-frequency distribution usually deviates from the G-R law. Based on the experimental results of three different rock samples, which contain macro-asperity, compressional en-echelon fault and model-Ⅲ shear fault, the correlative coefficient (r) was calculated by fitting the sequences of the acoustic emissions with the G-R law in every scanning time window. We investigated the changes of the correlative coefficient, which describes the actual deviation of magnitude-frequency distribution of earthquakes from the G-R law within the specific scanning time window. According to the results of the rock samples containing macro-asperity and compressional en-echelon fault respectively, the value decreases prior to the rock fracture, meaning that the deviation of magnitude-frequency distribution from the G-R law tends to be larger. The result of the model-Ⅲ shear fault didn’t show obvious decrease before the final rock fracture. Actually, the studies of some earthquakes also show deviation before the occurrence of moderate earthquakes. The results obtained in this paper will provide us with some useful clues for studying precursors before the occurrence of moderate earthquakes with the data of regional earthquake activities.
