Given a regular compact set E in , a unit measure μ supported by , a triangular point set , and a function f , holomorphic on E , let πβ,fn,m be the associated multipoint β-Padé approximant of order (n,m) . W...Given a regular compact set E in , a unit measure μ supported by , a triangular point set , and a function f , holomorphic on E , let πβ,fn,m be the associated multipoint β-Padé approximant of order (n,m) . We show that if the sequence πβ,fn,m , n∈Λ , ∧∈n,k are uniformly distributed on with respect to u as n∈Λ . Furthermore, a result about the behavior of the zeros of the exact maximally convergent sequence Λ is provided, under the condition that Λ is “dense enough”.展开更多
基于四叉树的分层马尔可夫随机场(Markov random field,MRF)模型在层间存在因果性,不需要像非因果马尔可夫随机场模型那样的迭代算法,但是传统的分层MRF模型常常导致分割结果具有块状现象和非连续边缘．本文提出一种新的基于区域确定的...基于四叉树的分层马尔可夫随机场(Markov random field,MRF)模型在层间存在因果性,不需要像非因果马尔可夫随机场模型那样的迭代算法,但是传统的分层MRF模型常常导致分割结果具有块状现象和非连续边缘．本文提出一种新的基于区域确定的半树分层MRF算法,并推导出它的最大后验边缘概率(Maximizer of the posteriori marginal,MPM)算法．在流域算法过分割结果的基础上,该模型将层间的点概率转换为区域概率,采用区域概率实现各层图像分割．从SAR图像的监督分割实验结果来看,本文提出的模型较好地克服了基于像素分层模型和单分辨率MRF模型带米的块现象和非连续边界,因而具有更好的分割结果．展开更多
文摘Given a regular compact set E in , a unit measure μ supported by , a triangular point set , and a function f , holomorphic on E , let πβ,fn,m be the associated multipoint β-Padé approximant of order (n,m) . We show that if the sequence πβ,fn,m , n∈Λ , ∧∈n,k are uniformly distributed on with respect to u as n∈Λ . Furthermore, a result about the behavior of the zeros of the exact maximally convergent sequence Λ is provided, under the condition that Λ is “dense enough”.
文摘基于四叉树的分层马尔可夫随机场(Markov random field,MRF)模型在层间存在因果性,不需要像非因果马尔可夫随机场模型那样的迭代算法,但是传统的分层MRF模型常常导致分割结果具有块状现象和非连续边缘．本文提出一种新的基于区域确定的半树分层MRF算法,并推导出它的最大后验边缘概率(Maximizer of the posteriori marginal,MPM)算法．在流域算法过分割结果的基础上,该模型将层间的点概率转换为区域概率,采用区域概率实现各层图像分割．从SAR图像的监督分割实验结果来看,本文提出的模型较好地克服了基于像素分层模型和单分辨率MRF模型带米的块现象和非连续边界,因而具有更好的分割结果．
