论文部分内容阅读
研究了在复合高斯杂波中利用先验知识自适应检测目标的问题,证明了基于双参数逆高斯分布纹理的复合高斯模型比传统的K分布、复t分布和单参数逆高斯分布纹理的复合高斯分布模型能够更好地拟合实际杂波。文中选择双参数逆高斯分布作为纹理分量的先验分布、基于广义似然比准则和贝叶斯方法设计得到了一种复合高斯杂波中的自适应检测器。理论分析和数值仿真表明,与自适应匹配滤波器和正则化自适应匹配滤波器相比,该检测器具有更好的检测性能。
The problem of adaptively detecting a target by using prior knowledge in complex Gaussian clutter is studied. It is proved that the complex Gaussian model based on the two-parameter inverse Gaussian distribution texture is more complex than the traditional K distribution, complex t distribution and single parameter inverse Gaussian distribution texture Gaussian distribution model to better fit the actual clutter. In this paper, two-parameter inverse Gaussian distribution is chosen as the prior distribution of texture components. Based on the generalized likelihood ratio criterion and Bayesian method, an adaptive detector in complex Gaussian clutter is proposed. Theoretical analysis and numerical simulation show that the detector has better detection performance than adaptive matched filter and regularized adaptive matched filter.