Floating small target detection based on the dual-polarization cross-time-frequency distribution in sea clutter

•A polarization fusion feature-based detector to detect small targets in sea clutter is given.•Dual-polarization cross-time-frequency distribution is applied.•Performance improvement is due to a SCR-based time-frequency distribution fusion method. In this paper, a polarization fusion detection method based on dual-polarization cross-time-frequency distribution is proposed to improve the detection performance compared with the original TF-feature-based detector in floating target detection in sea clutter background. Utilizing the received echoes information from the dual-polarization channels, time-frequency distributions (TFD) of radar returns from each polarization channel and the cross-time-frequency distribution (CTFD) of the two channels are computed. Weighted geometric average based on the signal-to-clutter ratio (SCR) is used to calculate a weighted fusion TFD based on the TFDs and CTFD from dual-polarization channels. To observe differences between sea clutter and target precisely, normalized time-frequency distributions (NTFD) of the cell under test (CUT) are calculated by the mean and variance of the fusion TFD computed from the reference range cells adjacent to the CUT. And the three existing features, the ridge integration (RI), the number of connected regions (NR) and the maximal size of connected regions (MS) are extracted from the NTFD and the convexhull learning algorithm is used to construct a polarization fusion detector in the feature space. Analyses of experimental results on the recognized IPIX database show that the proposed polarization fusion detector obtains superior detection performance compared to the original TF-feature-based detector and has competitive detection performance compared to other existing excellent feature-based detectors.