Multiscale modeling of sea clutter to facilitate detection of low observable targets within sea clutter

Abstract

Sea clutter, the radar backscatter from the ocean surface, is highly complicated and non-stationary, due to multipath propagation of the radar returns and multiscale interactions at the air-sea interface. To facilitate robust detection of low observable targets within sea clutter, which is an important issue in coastal security, navigation safety and environmental monitoring, we propose a systematic multiscale approach to the modeling of sea clutter. Specifically, we (i) develop new methods to better fit non-stationary and non-Gaussian sea clutter, (ii) fully characterize the correlation structure of sea clutter on multiple time scales, (iii) develop a highly accurate cascade model for sea clutter, and (iv) develop accurate and readily implementable methods to detect low observable targets within sea clutter.