基于离散小波分析的非接触探地雷达数据去噪研究

Research on Denoising of Non-Contact GPR Data Based on Discrete Wavelet Analysis

探地雷达非接触检测工作方式已在铁路路基检测领域广泛采用,相较于传统的接触式探地雷达检测方式具备更安全的检测环境和更高的采集效率。但由于雷达天线离开道砟面且保持一定的高度空间,信号更容易受噪声影响,因此信号分析过程常因噪声干扰而变得困难。为了改善信号质量,研究离散小波变换技术对非接触式探地雷达回波信号的去噪方法意义重大。试验使用不同子波基进行处理,深入探索小波去噪方法在铁路路基检测数据的应用效果,筛选出最适宜朔黄重载铁路路基的小波波形和变换参数。在实际数据测试过程中,通过选取适当的小波基函数和重构尺度系数,实现了更好的噪声抑制和高信噪比。通过量化参数评估,在滤波处理效果和信息保留方面取得了较好的平衡。研究结果可为非接触式探地雷达铁路路基检测数据处理与图像识别领域提供新思路。

The non-contact detection mode of GPR has been widely used in the field of railway subgrade detection.Compared with the traditional contact detection mode,the GPR has a safer detection environment and higher acquisition efficiency.However,since the radar antenna leaves the track surface and maintains a certain height space,the signal is more susceptible to noise,so the signal analysis process is often difficult due to noise interference.In order to improve signal quality,this study used discrete wavelet transform technology to denoise the non-contact GPR signal,which is of great significance.This study uses different wavelet bases for processing and explores in depth the application effects of wavelet denoising methods,and then selects the most suitable wavelet waveforms and transformation parameters for the Shuozhou-Huanghua Heavy-Duty Railway subgrade.In the actual data testing process,better noise suppression and high signal-to-noise ratio are achieved by selecting appropriate wavelet basis functions and reconstructing scale coefficients.Through quantization parameter evaluation,a good balance is achieved between filtering effect and information retention.The research results can provide new ideas for non-contact GPR railway subgrade detection data processing and image recognition.