When the compound-Gaussian clutter can be clustered into groups of data with one and the same value of the texture

the Constant False Alarm Rate (CFAR) property and the detection performance of the adaptive normalized matched filter (ANMF) are analyzed in the mismatch situation

where the estimated clutter group size is unequal to the actual one.The theoretical analyses show that

the ANMF holds fully CFAR property when the actual clutter group size is multiples of the estimated one

while it is only CFAR to the clutter covariance matrix structure for other cases.It also shows that

the ANMF holds almost CFAR property for different subsets.Moreover

different mismatch cases cause different performance loss for the ANMF;and the worst performance results from the case when no a priori information about the texture correlation is available at all.