SAR图像伪加性相干斑噪声统计特性

Statistics of Pseudo Additive Speckle Noise in SAR Image

研究了SAR图像加性模型中伪加性相干斑噪声分量的统计特性。分析了SAR图像中真实信号分量和乘性相干斑噪声分量的统计特性;在此基础上结合Edgeworth展开式和以Mellin变换为基础的第二类型特征函数及其对数累积量详细推导了伪加性相干斑噪声概率密度函数的近似表达式;各信号分量的概率密度函数中的参数采用对数累积量方法进行估计。最后利用SAR实测数据仿真分析了伪加性相干斑噪声的统计特性。实验结果表明,乘性的相干斑噪声转换为加性噪声后,其统计特性非常接近于高斯分布。

The statistical properties of pseudo additive speckle noise in synthetic aperture radar （SAR） image are investigated in this paper. First, the statistical properties of the noise-free signal and the muhiplicative speckle noise in SAR image are analyzed, then the approximate probability density function （PDF） of the pseudo additive speckle noise is derived by using the Edgeworth expansion. The parameters involved in PDF of various signals are estimated by using the method of log-cumulants based on Mellin transform. The real SAR images are utilized to analyze the derived PDF, and the results show that the statistical properties of the pseudo additive speckle noise closely approach to Gaussian distribution.