Détection CFAR de cibles réparties dans un clutter K-distribué de paramètres inconnus

Abstract

In high resolution radars (HRR), unlike point-target model, targets are described as a reflection from isolated points, referred to as MDS (Multiple Dominant Scattering). In this work, we consider the detection of spatially distributed targets (MDS model) embedded in K-distributed clutter with a partially correlated texture, and without any a priori knowledge about the clutter parameters. To this end, we propose a new detection approach, based on the Cell-Averaging (CA) scheme, using Look up Tables (LT), which contain the pre-computed thresholding factors that maintain a constant false alarm rate. The parameters of the K-distribution are estimated, in order to assign to the threshold its appropriate value according to the actual environment. The performances of the proposed detector were analyzed and compared to those of the Logarithmic-based Cell Averaging detector (CAL), for different shape and scale parameters and correlation degrees. The simulation results showed that the proposed detector (CA-LT) acts robustly and outperforms the CAL in the different situations under investigation.