Floating Small Target Detection in Sea Clutter Based on the Singular Value Decomposition of Low Rank Perturbed Random Matrices

Detection of sea-surface small floating target is always a challenging problem under the influence of sea clutter. In this paper, a feature-based detector is proposed for effective and robust detection of floating small target in sea clutter. Three features have been extracted from the received signal based on the singular value decomposition (SVD) of large dimensional rectangular random matrix to distinguish target echo and sea clutter. The maximum singular value, the mean spectral radius (MSR) and the l
1
norm of the principle right singular vector are exploited to form the 3D feature space. Based on the three features extracted, the support vector machine (SVM) algorithm is used to detect floating small target within sea clutter background. Experiments on all public datasets of IPIX radar show that our proposed method achieves higher detection probability compared with the existing methods.