GPR Data Denoising Using Combine Domain Of Wavelet And Curvelet Transform

Ground penetrating radar(GPR),as a nondestructive,fast and economical method of shallow geophysical exploration,is widely used in infrastructure construction,engineering investigation and other fields.In the stage of collecting data in complex environment,GPR observation is interfered by human,natural environment and instrument itself,and the collected data is adulterated by different levels and types of noise.Therefore,there are errors in the qualitative and quantitative analysis and interpretation of the ground penetrating radar noisy data,and the data denoising has always been concerned by people.High resolution,high SNR and high fidelity are the basic requirements of GPR data denoising.The traditional denoising method of GPR data can not meet the current needs of complex data denoising.Based on the advantages of wavelet transform and curvelet transform and the advantages of data de-noising processing,this paper studies the de-noising algorithm of GPR data in the combination domain of wavelet transform and curvelet transform,which provides theoretical basis for the de-noising of GPR data with high resolution,high signal-to-noise ratio and high fidelity,evaluates the effectiveness of the de-noising effect by using the synthetic data,and verifies its feasibility by the measured data.The specific work is as follows:(1)The research contents and methods of GPR data denoising are systematically investigated,and the application status and development trend of existing GPR data denoising methods are introduced in detail.In this paper,the theory of wavelet transform and curvelet transform are derived in detail.Aiming at the problem of wavelet threshold,the wavelet transform denoising algorithm of "window" high-order correlation statistics is proposed.Aiming at the problem of block complex threshold function and threshold control coefficient,the range of curvelet transform threshold control coefficient is discussed,which provides the theoretical basis for the algorithm of wavelet transform and curvelet transform combination domain.(2)This paper systematically analyzes and compares the denoising effect of traditional wavelet transform threshold algorithm,improved wavelet transform algorithm and curvelet transform algorithm,and clarifies the application scope and denoising effect of wavelet transform and curvelet transform in GPR data denoising.Based on the algorithm of wavelet transform and curvelet transform and its application limitations,the application of high-order correlation statistical calculation(wavelet analysis of curvelet coefficients)in the curved wave transform domaincan effectively count the random noise in the middle and high-frequency curved wave coefficient layers,effectively distinguish the related noise curved wave coefficients,further weaken the related noise energy,highlight the effective signal,and develop wavelet research on denoising Algorithm in combination domain of transform and curve transform.(3)The algorithm of wavelet transform and curvelet transform combination domain is applied to the comparative study and analysis of theoretical simulation data and measured data.Firstly,the model is established to simulate the synthetic data,and the effectiveness of the combined domain algorithm is verified by comparing with the denoising results of single domain wavelet transform and curvelet transform.Secondly,through the study of the de-noising of the ground penetrating radar measured data,it shows that the combination domain algorithm of wavelet and curvelet transform can effectively filter out the measured data noise,at the same time,it can reconstruct the effective signal more precisely and precisely,and solve the problem of the artifacts in the curvelet transform.In the curvelet transform domain,the vertical direction coefficients of the medium and high frequency scale are selected to zero,and then the high-order correlation statistics are carried out in the curved wave domain The combination domain algorithm of wavelet transform realizes the purpose of denoising and highlighting the weak effective signal.