为研究探地雷达(GPR)对路面的缺陷检测识别,针对路用探地雷达信号的处理问题,通过在路面基层内各层界面处埋置金属板,分别对数字滤波、反褶积、波动方程偏移和小波多尺度去噪等方法的处理效果进行对比,提出了对提高异常的分辨率和奇异值分解具有明显效果的方法,并应用于道路现场检测中。结果表明,反褶积和波动方程偏移处理对提高异常的分辨率具有明显效果,小波变换技术分析具有很好的去噪和奇异值分解效果,可以减少处理数据的工作量和增加检测的准确性。 In order to study GPR detection and recognition on pavement defects, aiming at the problem of road GPR signal processing, the digital filtering, deconvolution and fluctuation Equation migration and wavelet multiscale denoising methods are compared. The methods to improve the resolution and singular value decomposition of anomalies are proposed and applied to the on-the-spot testing. The results show that the deconvolution and wave equation migration have significant effect on improving the resolution of the anomalies. The wavelet transform technique has good de-noising and singular value decomposition effects, which can reduce the workload of data processing and increase the accuracy of detection Sex.