一种被动声呐信号检测的低虚警率方法

针对被动声呐多目标信号检测中的噪声背景归一化问题,提出了一种基于数学形态学滤波的噪声背景归一化新方法。该方法利用数学形态学处理中的膨胀和腐蚀算子,以及基于多项式拟合的数据均值估计方法,构造出了一种能够较为准确的估计噪声门限的方法,并以之进行噪声背景归一化,在较好保留信号信息的前提下较大程度的抑制了噪声,有效降低了多目标信号检测的虚警概率。通过计算机仿真对比了该算法与S3PM算法、OTA算法的性能,结果表明该噪声背景归一化算法能够在检测概率损失较小的情况下较大幅度地降低检测的虚警率。实际被动声呐数据处理的对比结果同样验证了该算法的有效性。 

Passive sonar signal detection algorithm with lower false alarm rate

A novel noise background normalization algorithm is proposed to deal with the multi-target signal detection problem of passive sonar.The key of the proposed algorithm is composed of three main parts,Dilation operator, Erosion operator and polynomial fitting method.The Dilation operator and Erosion operator are commonly used in Mathematical Morphology signal processing,and the polynomial fitting method is used to evaluate the mean value of the data.By using these three basic methods,the threshold of the noise is estimated accurately,and the proposed algorithm performs well in noise suppressing while preserving as much signal information as possible.Performance comparisons are given through computer simulations with the Split Three-Pass Mean(S3PM) algorithm and the Order Truncate Average (OTA) algorithm.Results showed that the proposed noise background normalization algorithm could depress the false alarm rate substantially when the cost of the detection probability is slightly.Real Bearing-Time Record(BTR) data of passive sonar is also used to check the capability of three algorithms mentioned above.The novel proposed algorithm presents a better performance than the other two methods in practical applications too.