...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Sensing and imaging >SAR Image Despeckling Based on a Mixture of Gaussian Distributions with Local Parameters and Multiscale Edge Detection in Lapped Transform Domain
【24h】

SAR Image Despeckling Based on a Mixture of Gaussian Distributions with Local Parameters and Multiscale Edge Detection in Lapped Transform Domain

机译：重叠变换域中基于局部参数高斯分布和多尺度边缘检测混合的SAR图像去斑。

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

A new Lapped transform domain SAR image despeckling algorithmrnusing a two-state Gaussian mixture probability density function that uses localrnparameters for the mixture model, is proposed. The use of lapped orthogonalrntransform (LOT) is motivated by its low computational complexity and robustnessrnto oversmoothing. It is shown that the dyadic rearranged LOT coefficients of logarithmicallyrntransformed SAR images can be well approximated using two-staternGaussian mixture distribution compared to Laplacian, Gamma, generalized Gaussianrnand Cauchy distributions, based on the Kolmogorov–Smirnov (KS) goodness ofrnfit test. The LOT coefficients of speckle noise are modeled using zero meanrnGaussian distributions. A maximum a posteriori (MAP) estimator within Bayesianrnframework is developed using this proposed prior distribution and is used to restorernthe noisy LOT coefficients. The parameters of mixture distribution are estimatedrnusing the expectation-maximization algorithm. This paper presents a new techniquernto identify LOT modulus maxima which allows us to classify LOT coefficients intornthe edge and non edge coefficients. The classified edge coefficients are keptrnunmodified by the proposed algorithm whereas the noise-free estimates of non-edgerncoefficients are obtained using Bayesian MAP estimator developed using two staternGaussian mixture distribution with local parameters. Finally the proposed techniquernis combined with the cycle spinning scheme to further improve the despecklingrnperformance. Experimental results show that the proposed method very efficiently reduces speckle in homogeneous regions while preserving more edge structuresrncompared to some recent well known methods.
机译：提出了一种新的重叠变换域SAR图像去斑算法，该算法利用两态高斯混合概率密度函数对混合模型使用局部参数。重叠正交变换（LOT）的使用是由于其较低的计算复杂度和过平滑的鲁棒性。结果表明，基于Kolmogorov-Smirnov（KS）拟合优度检验，与Laplacian，Gamma，广义Gaussianrn和Cauchy分布相比，使用二态高斯混合分布可以很好地近似对数变换后的SAR图像的二重排列LOT系数。使用零均值高斯分布对斑点噪声的LOT系数进行建模。贝叶斯框架中的最大后验（MAP）估计器是使用此提议的先验分布开发的，并用于恢复嘈杂的LOT系数。使用期望最大化算法估计混合物分布的参数。本文提出了一种识别LOT模数最大值的新技术，该技术使我们能够将LOT系数分为边缘系数和非边缘系数。所分类的边缘系数通过所提出的算法保持不变，而非边缘系数的无噪声估计使用贝叶斯MAP估计器获得，贝叶斯MAP估计器使用具有局部参数的两种状态高斯混合分布开发。最后，提出的技术与循环纺纱方案相结合，进一步提高了去斑点性能。实验结果表明，与最近的一些众所周知的方法相比，该方法可以有效地减少均匀区域中的斑点，同时保留更多的边缘结构。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号