3-D TARGET FEATURE-EXTRACTION VIA INTERFEROMETRIC SAR

The authors consider 3-D (three-dimensional) target feature extraction via an interferometric synthetic aperture radar (IFSAR). The targets of interest are relatively small and consist of a small number of dist inct point scatterers. Since using IFSAR to extract the features of su ch targets has not been addressed before, a self-contained detailed de rivation of the data model is presented. A set of sufficient parameter identifiability conditions on the data model and the Cramer-Rao bound s (CRBs) on the parameter estimates are also derived. Four existing tw o-dimensional feature extraction methods (FFT, windowed FFT, Capon, an d MUSIC) are extended to the 3-D parameters of the target scatterers. A new nonlinear least squares parameter estimation method, referred to as IFRELAX, is also derived to extract the target features. Finally, numerical examples are presented to compare the performances of the pr esented methods with each other and with the corresponding CRBs. The a uthors show by means of numerical examples that, among the three nonpa rametric methods (FFT, windowed FFT, and Capon), Capon has the best re solution. The parametric methods MUSIC and IFRELAX can have much bette r resolution and provide much more accurate parameter estimates the no nparametric methods. It is shown IFRELAX can be faster and provide muc h better parameter estimates than MUSIC.