An Improved Ship Detection Algorithm for an Airborne Passive Interferometric Microwave Sensor (PIMS) Based on Ship Wakes
Hailiang Lu,Yinan Li,Liang Lang,Lulu Wu,Lin Xu,Songhua Yan,Qingbiao Fan,Wenchao Zheng,Rong Jin,Rongchuan Lv,Hao Li
DOI: https://doi.org/10.1109/tgrs.2023.3302029
IF: 8.2
2023-08-19
IEEE Transactions on Geoscience and Remote Sensing
Abstract:Airborne passive interferometric microwave sensors (PIMSs) were proposed as a powerful complementary tool for ship detection, especially in low visibility and high sea clutter environments. At present, there is only one algorithm for ship detection by airborne PIMSs, and this algorithm is only based on the "cold" brightness temperature (TB) characteristic of metal ships with wakes in passive interferometric microwave images (PIMIs). Moreover, it has been found that ship wakes always exhibit a "hot" TB characteristic in PIMIs. In this article, an improved algorithm is innovatively proposed for ship detection by airborne PIMSs to improve detection performance, and it is simultaneously based on the "cold" and "hot" TB characteristics of metal ships with wakes. The microwave TB model of metal ships with wakes is established to analyze its TB characteristics. Given the TB model and its characteristics, an improved algorithm is introduced, which comprises five steps: searching targets, improving TB maps, clustering targets, estimating ship headings, and estimating ship trajectories. The improved algorithm can detect ships with a low false-alarm probability (FAR) and accurately estimate ship headings using only a single snapshot and the ship trajectory. The practicability and detection performance of the improved algorithm is verified by airborne experiments using an X-band PIMS. The experiment results demonstrate that the improved algorithm can accurately detect ships, estimate ship headings using only a single snapshot, and estimate ship trajectories. The impact of the algorithm parameters on the detection performance is analyzed and discussed. Moreover, the detection performance of the improved algorithm is analyzed and compared with the performance of the only existing algorithm, and the experiment results indicate tha- the improved algorithm has a better detection performance with a lower FAR value than the only existing algorithm, as expected.
imaging science & photographic technology,remote sensing,engineering, electrical & electronic,geochemistry & geophysics