Outlier-Robust Truncated Maximum Likelihood Parameter Estimation of Compound-Gaussian Clutter with Inverse Gaussian Texture

Compound-Gaussian distributions with inverse Gaussian textures, referred to as the IGCG distributions, are often used model moderate/high-resolution sea clutter in amplitude. In maritime radars, parameter estimation of from radar returns data plays an important role adaptive target detection. Due inevitable existence outliers high amplitude targets and reefs, must be outlier robust. this paper, outlier-robust truncated maximum likelihood (TML) method is proposed mitigate effect data. The first transferred into by removing a given percentage largest samples From data, function constructed, its corresponds TML estimates scale shape parameters. Further, iterative algorithm presented obtain outliers, which extension ML case that contain outliers. comparison outlier-sensitive methods bipercentile methods, performance close best without outlier, it better