基于时空域的探地雷达杂波的去除

直耦波和随机噪声是影响雷达信号的杂波,根据各种杂波在各频率段分布不同,在各道间相关性也不同的特性,采用时空域相结合的去噪方法,在时间域上采用小波变换将其分解到各频域上,在空间上采用KL变换提取出有效信号。去噪后的雷达图像更加清晰地反映地下介质情况,且在信噪比较低的情况下,仍能突出目标回波。该方法去除杂波速度快、效果好,特别适用于公路、铁路路基检测等海量数据处理领域。

Removing Clutter of GPR Based on Temporal and Spatial Domain

Direct coupled wave and random noise are clutters affecting the shallow and deep signal of the ground penetrating radar(GPR).According to the differences in distribution of frequency and relevance in traces,these clutters are removed based on temporal and spatial domain.In temporal domain wavelet transform is used to break down GPR signal into different frequency-domain.In spatial domain KL transform is applied to extract effective signal.After removing the noise,GPR image shows underground medium clearer,gives prominence to the target echo in low SNR(signal-to-noise ratio),especially in the shallow and deep layers.The method wipes off clutters quickly with good effects,and is applicable to roadbed and railway trackbed inspection with large quantity of data.