Target prescreening based on 2D gamma kernels

This work develops and tests a new target prescreening algorithm based on 2D gamma kernels. The key feature of the new kernel set is the existence of a free parameter that determines the size of its region of support. We show that the scale affects the false alarm rate of the two parameter CFAR test. We also show that a linear discriminant funtion composed from the linear and quadratic terms of the intensity in the test cell neighborhood improves the false alarm rate when compared with the two parameter CFAR.

[1]  José Carlos Príncipe,et al.  The gamma-filter-a new class of adaptive IIR filters with restricted feedback , 1993, IEEE Trans. Signal Process..

[2]  J.C. Principe,et al.  A gamma memory neural network for system identification , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[3]  José Carlos Príncipe,et al.  Parametric least squares approximation using gamma bases , 1995, IEEE Trans. Signal Process..

[4]  José Carlos Príncipe,et al.  Echo cancellation with the gamma filter , 1994, Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing.

[5]  Christine M. Netishen,et al.  Performance of a High-Resolution Polarimetric SAR Automatic Target Recognition System , 1993 .

[6]  José Carlos Príncipe,et al.  Generalized feedforward structures: a new class of adaptive filters , 1992, [Proceedings] ICASSP-92: 1992 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[7]  Norbert Wiener,et al.  Extrapolation, Interpolation, and Smoothing of Stationary Time Series, with Engineering Applications , 1949 .