Abstract: When using ground penetrating radar(GPR)to detect the underground target distribution,valid signal in receiving data tend to be susceptible to noise and the interference of air-ground waves,which will affect the accuracy of target recognition and increase the difficulty of target recognitionThe successful application of Shearlet transform in image and seismic data shows its superiority for denoising processWhen using the Shearlet transform to reduce the noise of the data,the choice of threshold has a great influence on the denoising effectIn order to improve the denoising effect for radar data,a novel denosing method combining singular value decomposition(SVD)method for suppression of random noise to enhance the reflection signal caused by underground targetsThe method is applied to process was proposed the simulation and field datasets,and the experimental results show that the method is effective