Research On Acquisition Software Design And Denoising Method Of Ground Penetrating Radar Signal

With the increase of investment in infrastructure in China,the construction of road traffic is developing rapidly.The exploration of underground rock and soil prior to road construction,the quality inspection of new roads and the maintenance of roads during using period are becoming more and more serious.Traditional road detection methods are inefficient and can cause some damages to roads.Ground penetrating radar(GPR)is a nondestructive detection instrument with fast detection and real-time imaging,which can meet the increasing requirements of road detection tasks.GPR has various structures,among which,impulse-GPR is widely used because of its simple structure and good detection effect.In this paper,the software design of the data acquisition control module in impulse-GPR and the denoising method of GPR echo signal are mainly studied.The main research contents of this paper are as follows:(1)Due to the high frequency of GPR echo signal,it is difficult to realize the echo signal acquisition by real-time sampling method.In this paper,FPGA is used as the master device.Firstly,programmable delay chip is controlled by software design to produce step delay.And then,transmitting the pulse after the delay by radar is controlled.Finally,equivalent sampling of the radar echo is achieved.Through the experimental test,the acquisition software designed in this paper can achieve equivalent sampling of radar echo signal.(2)For solving the problems of discontinuous and fixed deviations of typical threshold functions in wavelet threshold denoising,an improved threshold function is proposed in this paper by combining the modulus squared threshold function containing parameters with the Polynomial interpolation threshold function,which overcomes the shortcomings of the typical threshold function.The simulation results show that the denoising effect of noisy GPR echo signal is improved by using the improved threshold function compared with typical threshold functions.(3)This paper focuses on the situation that without denoising of the low frequency wavelet coefficients for the GPR echo signal when using wavelet threshold denoising.A joint denoising method is designed in this paper based on wavelet threshold denoising and Wiener filtering,which overcomes the problem of incomplete denoising in wavelet threshold denoising method.The simulation results show that the joint denoising method has better denoising effect than wavelet threshold denoising method.