舰船尾迹光学信号异常特征贝叶斯识别方法

Bayesian recognition method for abnormal features of ship wake optical signals

为了提升舰船尾迹光学信号异常特征识别效果,提出舰船尾迹光学信号异常特征贝叶斯识别方法。针对合成孔径雷达系统采集的舰船尾迹SAR图像中舰船尾迹与海杂波边界区分不清晰的情况,使用图像分割和归一化的Hough变换检测方法实现舰船尾迹图像增强;依据气泡运输方程提取舰船尾迹直方图,根据直方图内峰值点密集程度,提取舰船尾迹光学信号特征,将该特征作为输入,使用贝叶斯分类模型输出舰船尾迹光学信号异常特征识别结果。实验结果表明:该方法可有效增强舰船尾迹SAR图像,也可有效提取舰船尾迹直方图,并准确提取舰船尾迹光学信号特征和识别其中的异常特征。

In order to improve the recognition effect of abnormal features in ship wake optical signals,a Bayesian recognition method for abnormal features in ship wake optical signals is proposed.In response to the unclear distinction between ship wakes and sea clutter boundaries in SAR images of ship wakes collected by synthetic aperture radar systems,image segmentation and normalized Hough transform detection methods are used to enhance the ship wakes image;After extracting the ship wake histogram based on the bubble transport equation,the optical signal features of the ship wake are extracted based on the density of peak points in the histogram;Using this feature as input,use a Bayesian classification model to output the abnormal feature recognition results of the ship wake optical signal.The experimental results show that this method can effectively enhance SAR images of ship wakes,extract ship wakes histograms,and accurately extract optical signal features of ship wakes and identify abnormal features within them.