【摘要】 直耦波是影响探地雷达目标回波的主要杂波之一,其在各道间变化很小,即具有很强的相关性,采用KL变换能够将雷达数据分解为特征值矩阵和特征向量矩阵的乘积,其中特征值的大小反映各道数据相关性的强弱,选取特征值较小的数据便能重建出不相关信号,从而抑制了直耦波。将具有很强时频分析能力的小波变换与KL变换结合起来,设计出小波域KL变换能够弥补常规KL变换仅在时间域上考虑相关性的不足,在信噪比较低的情况下,仍能突出目标回波,尤其是在浅层。该方法去除直耦波速度快、效果好,特别适用于公路路基检测、铁路路基检测等海量数据处理领域。更多还原

【Abstract】 The direct coupled wave is one of the main radar clutter affecting target echo of the ground penetrating radar(GPR).It changes little with various trace. In other words,it is with great relevance. By using KL transform,the radar data can be decomposed into the product of eigenvalue matrix and eigenvector matrix,and the size of the eigenvalue can reflect how much they are relevant among them. Selecting data with small evignvalue can rebuild irrelevant signals and inhibits the direct coupled wave. Combining wavelet transform with great frequency analyzing abilities and KL transform,and designing KL transform in wavelet domain can make up the shortcomings of the general KL transform’s considering only the relevance in the time dimension,give prominence to the target echo in the condition of low SNR(signal-to-noise),especially in the shallow layer. The method wipes off direct coupled wave quickly with good effects,and is applicable to road and railway trackbed inspection with large quantities of data.更多还原