Marine Radar Image Sequence Target Detection Based on Space-Time Adaptive Filtering and Hough Transform

The performance of marine radar target detection is largely affected by the intricate and dynamic space-time variations of sea clutter signals, which cause substantial numbers of false and missed alarms. To improve the target detection performance of rotating scanning marine radar, this study proposes a marine radar image sequence target detection algorithm based on space-time adaptive filtering and the Hough transform algorithm. The algorithm adopts a two-stage approach of coarse detection followed by precise detection. During the coarse detection stage, the sea clutter energy in the 3-D frequency-wavenumber spectrum of the marine radar image sequence is suppressed by a sea clutter suppression algorithm in the space-time domain, space-time clutter suppression (STCS). Subsequently, moving targets are extracted from the image sequence using a target energy extraction method based on the Hough transform algorithm in the 3-D frequency-wavenumber domain. The result is a processed image sequence with sea clutter signal reduction and target signal extraction. The precise detection stage detects the target point in this processed image sequence using a constant false alarm rate method based on a real clutter background distribution model. During verification tests on real X-band marine radar data, the detection probability of the proposed method reaches 99.89% under low sea state, 95.34% under medium sea state, and 94.44% under high sea state. Compared with the WHOS-CFAR and GMOS-CFAR, the average improvement is 10.1% and 16.6%, respectively. Furthermore, compared to the STCS, there is a maximum improvement of 3.7%. The enhancement in detection performance is significant.