基于数学形态学的海洋混响背景下的信号检测

Signal Detection Based on Mathematical Morphology in Ocean Reverberation

为了改善在海洋混响背景低信噪比环境下的信号检测性能,该文从传统的白化滤波器出发,结合图像处理的方法,提出了将白化滤波、波束形成和数学形态学图像处理多种手段结合起来的算法。给出了对海洋混响背景信号检测参数选择和海洋试验的测试结果。计算结果表明,该检测方法确实能够改善海底目标检测的性能。尤其反应在处理后的图像上,目标信号较为明显。与传统的单一抗混响降噪手段相比,该方法更加充分利用了目标反射信号与混响、直达信号、舰船反射噪声在时域和空间角度上的差别,将各自的特征进一步的放大,不仅提高了检测的效果,而且可以直观对目标的方向进行定位。

In the paper, to improve the signal detection performance in an ocean reverberation environment, the algorithm based on whiten filter, beamforming, and mathematical morphology from image processing was presented as well as the parameter selection method of the detector and ocean test result. The result of an ocean experiment indicates that the algorithm does improve the performance when detecting targets are on the bottom of ocean. Especially, in the processed beam image, the target is very distinct. Comparing with single traditional anti-reverberation method, the algorithm fully utilizes the temporal and spatial difference among detection signal with reverberation, directed signals, reflected noise from ship to largen their characters,improve detecting targets and optimize the localization.