2026· IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing· Vol 19, pp. 32990-33008· 0 citations· 40 references
Abstract
Weak radar targets in maritime radar remote sensing are often masked by strong, time-varying sea clutter, while deep-network-enhanced echoes may no longer satisfy the statistical assumptions of conventional detectors. To address these issues, this study proposes a joint progressive enhancement and adaptive detection framework. A progressive multiscale enhancement network (PMENet) is first developed to suppress sea clutter and preserve compact target responses. It uses an iso-dimensional multiscale backbone, adaptive cross-scale feature fusion, high-frequency dual-residual reconstruction, and multiexit deep supervision to balance target preservation, clutter suppression, and computational efficiency. Then, a two-stage maximum between-class variance (TS-MBCV) detector is designed for the enhanced-output domain. By combining two successive threshold searches, peak-neighborhood censoring, a noise-floor constraint, and nonmaximum suppression, TS-MBCV makes detection decisions without requiring an explicit clutter distribution model. Experiments on measured $X$-band sea clutter data show that, for input signal-to-clutter ratios from $-$20 to 0 dB, the deepest exit achieves an average improvement factor of integrated signal-to-clutter ratio of 31.97 dB and an average F1-score of 0.9920. These results demonstrate that the proposed framework achieves effective weak-target enhancement and false-alarm suppression while retaining low computational cost in complex maritime environments.
Reliable weak target detection in nonstationary sea clutter is essential for maritime surveillance and navigation safety. Fluctuating sea-surface returns and evolving Doppler spectra can mask weak targets, while conditioning solely on the cell under test may introduce target components into the background representatio...
Yu-Jie Chen, Gui-Geng Li, Man-Li Zhou et al.· Journal of Marine Science an...· 0 citations
Synthetic aperture radar (SAR) ship detection is essential for maritime surveillance, yet coherent speckle can obscure tiny offshore vessels, while high-response coastal structures can produce target-like responses. Repeated downsampling further weakens sparse vessel signatures, whereas a separate high-resolution predi...
Space-Time Adaptive Processing (STAP) demonstrates excellent clutter suppression performance in airborne radar systems when sufficient independent and identically distributed (i.i.d.) samples are available. However, heterogeneous and non-stationary clutter environments often lead to a shortage of i.i.d. samples, degrad...
Compound noise, clutter, and multipath reduce target–background separability in frequency-modulated continuous-wave (FMCW) radar Range–Doppler (RD) maps. Rawonly processing can leave interference-obscured weak targets insufficiently salient, whereas replacing Raw with an enhanced representation or directly mixing the t...
Ground Penetrating Radar (GPR) is an important non-destructive sensing technology for subsurface detection. However, strong clutter from surface reflection and heterogeneous media often masks weak target responses in B-scan images. Existing low-rank and sparse decomposition (LRSD) methods, such as Robust Principal Comp...
Cheng Chen, Le-Tian Wang, Xiao-Wen Liu et al.· Remote Sensing· 0 citations
Interrupted sampling and repeater jamming (ISRJ) severely affects the detection of real targets by generating highly realistic coherent false targets. To address this issue, this article proposes a false target discrimination and rejection method that combines joint fast–slow time dimension processing and deep learning...
Jia-Wei Lu, Yi-Duo Guo, Xiao-Wei Hu et al.· IEEE Transactions on Aerospa...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.