ID 原文 译文
12154 该算法以非局部欧氏中值(nonlocal Euclidean medians,NLEM)滤波算法为基础,根据含噪图像梯度幅值在一定噪声范围内服从Rayleigh分布这一特性,求得以梯度幅值和噪声标准差为自变量的二元自适应滤波参数,并将它引入到邻域的权值计算中。 The algorithm to nonlocal Euclidean median (nonlocal Euclidean medians, NLEM) filter algorithm as the foundation, according to the image gradient amplitude in a certain noise signals with noise obey Rayleigh distribution within the scope of the properties, obtained by gradient amplitude and the noise standard deviation for binary adaptive filter parameters of the independent variables, and to introduce it to the neighborhood of the weight calculation.
12155 其次,噪声的变化影响着p范数回归的选择,在一定范围内以噪声标准差为自变量对参数p进行多项式拟合,得到自适应p范数回归。 Secondly, the change of noise affects the selection of p norm regression, and the parameter p is polynomically fitted with the noise standard deviation as the independent variable within a certain range to obtain adaptive p norm regression.
12156 在自适应滤波参数基础上,用自适应p范数回归进一步改进NLEM滤波算法的1-范数回归。 Based on the adaptive filtering parameters, the 1-norm regression of the NLEM filter algorithm is further improved by adaptive p norm regression.
12157 所选图像的实验结果表明,本文算法在一定噪声范围内不但获得满意的去噪效果,而且有效地减少人机交互程度。 The experimental results for the selected images show that the algorithm is within a certain range of noise not only obtain satisfactory denoising effect, and effectively reduce the degree of human-computer interaction.
12158 针对自相似业务流量下的高突发性及重尾性所引起的空间数据系统调度性能下降问题,分析了高级在轨系统(advanced orbiting system,AOS)虚拟信道存取(virtual channel access,VCA)子层调度策略以及现有基于短相关模型调度算法的不足。 For self-similar business flow under high spatial data caused by sudden and heavy tail system scheduling performance degradation problems, analyzes the advanced orbit system (advanced orbiting system, AOS) virtual channel access (virtual channel access, VCA) sub-layer scheduling policy and the shortage of the existing scheduling algorithm based on short correlation model.
12159 引入Hurst参数、紧迫度、流量离差、成帧时间因子等权值参量。 Introducing the Hurst parameter, pressing into the frame, flow deviation, time factor weight parameters.
12160 提出一种基于延时累积的自适应轮询调度(scheduling of delay accumulated adaptive polling,SDAAP)算法。 Is proposed based on adaptive polling delay accumulated scheduling (scheduling of delay accumulated the adaptive polling, SDAAP) algorithm.
12161 通过自适应改变延时阀值因子实现多业务的差异化调度,从而优化AOS虚拟信道服务质量及调度性能。 Through the adaptive change of time-delay valve value factor to realize the business differentiation scheduling, thus optimizing the AOS service quality and virtual channel scheduling performance.
12162 采用多信源重尾分布的ON/OFF流量分布模型进行仿真验证。 ‭Multi source of heavy-tailed ON/OFF traffic distribution model simulation.
12163 实验结果表明,针对自相似业务流,SDAAP算法在溢出率、平均延迟等方面优于AOS固定阀值和等时调度算法。 The experimental results show that for the self-similar traffic stream SDAAP algorithm in such aspects as overflow ratio, average delay is superior to the AOS fixed threshold and scheduling algorithm etc.