GLRT detector based on knowledge aided covariance estimation in compound Gaussian environment

In order to alleviate the effect of the limited secondary data in the non-Gaussian clutter, a knowledge aided adaptive detector is proposed. The covariance matrix estimation is modeled as a general linear combination of prior covariance matrix and sample covariance matrix. Within this consideration, we obtain an adaptive detector based on the generalized likelihood ratio test. Experimental results on simulation and real data demonstrate that the proposed detector achieves better performance than the existing one-step GLRT (1S-GLRT) detectors when the secondary data are insufficient.