一种优化的极化SAR图像海面目标检测方法

提出一种优化的极化SAR图像海面目标检测方法,结合改进的极化SAR四分量分解中的螺旋散射分量与Wishart分类器,充分利用极化散射特性、结构特征、统计特性来进行目标的自动检测。同时通过纹理特征相似性克服了Wishart分类器在无目标海域检测时容易将强度值较高的海杂波误认为目标的缺陷。采用美国无人机UAVSAR在Mexico海域和巴拿马Barro Colorado Island海域获取的两组L波段全极化数据进行实验验证。实验结果表明:文中的优化方法能够较准确检测海面目标,很好地降低虚警率;同时解决了Wishart分类器在无目标海域发生错检的问题。

An optimized PolSAR sea-surface target detection method

An optimized sea-surface target detection method based on Poarimetric SAR (PolSAR ) data is proposed which combines the enhanced Yamaguchi four-component decomposition with the Wishart classifier ,therefore being able to detect sea-surface targets automatically by using the advantages of scattering property ,structural feature and statistic characteristics .Meanwhile ,this method can overcome the deficiency that the Wishart classifier will falsely recognize the sea clutter with high intensity as targets in the no‐target sea area ,which selects two L‐band quad-polarization data covered USA UAVSAR in the sea area of the Gulf of Mexico and Panama Barro Colorado Island for experiments .Experimental results indicate this optimized method can detect the sea-surface targets effectively ,reduce the false alarm rate and overcome the w rong detection by Wishart classifier in no‐target sea areas .