Low Observable Targets Detection by Joint Fractal Properties of Sea Clutter: An Experimental Study of IPIX OHGR Datasets

Abstract

We exploit the joint fractal properties of sea clutter extracted from detrended fluctuation analysis (DFA) for targets detection. We find that two specific fractal statistics, i.e., the intercept at the crucial scale and the Hurst exponent of optimal scales provide valuable information for targets detection. The first statistic measures the discrepancy between sea clutter and low observable targets at the crucial fractal scale, and the second one evaluates the average fractal difference within the optimal multi-scales. A target detection method integrating these two statistics is proposed, which is validated by real-life IPIX radar datasets. We find that this joint fractal detection approach achieves more accurate results for low observable targets detection.