基于组合滤波设计的背景噪声归一化方法

Background Noise Normalization Method Based on Combinational Filter Design

针对阵列信号处理输出结果中的背景噪声归一化问题,提出了一种基于组合滤波器设计的噪声背景归一化方法。该方法利用形态学运算子对输出结果进行组合处理,降低背景噪声波动对目标检测影响;根据目标稳定特性设计频域波器对输出结果进行卷积处理,降低背景高频噪声对目标检测影响;通过多项式拟合方法实现背景噪声门限的自动推荐,对背景噪声进行归一化处理,在保留目标信号的条件下,有效抑制了背景噪声,降低了目标检测中的虚警概率。

For the background noise normalization problem of array signal processing output results,a background noise normalization method based on combinational filter design is proposed.Firstly,the output results are combinational processed by morphology operators to reduce the impact of background noise fluctuation on target detection.Then,the output results are convolution processed by frequency domain filter which is designed according to the stability characteristic of target.Lastly,the noise limit value is automatically recommended by polynomial fitting method to normalize the background noise.This method performs well in noise suppressing under the condition of retaining the target signal,and depress the false alarm rate substantially.