论文部分内容阅读
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。
The direct coupling wave is one of the main clutter affecting the target echo of GPR. It has a small correlation between each channel, that is, it has strong correlation. The KL transform can decompose radar data into eigenvalue matrix and eigenvector Matrix product, in which the size of the eigenvalue reflects the strength of the data correlation, the less eigenvalues of the data can be reconstructed to irrelevant signals, thus inhibiting the direct-coupled wave. Combining the wavelet transform with KL transform with strong time-frequency analysis ability, we designed the KL transform of wavelet domain to make up for the shortcoming that the conventional KL transform considers the correlation only in the time domain. When the SNR is low, Can highlight the target echo, especially in the shallow. The method can remove the direct-coupled wave fast and has good effect and is especially suitable for the mass data processing field such as roadbed detection and railway subgrade detection.