| The ground-penetrating radar was widely used in engineering geology exploration at home and abroad for used detection for its high resolution, short detection cycle, and undamaged to detect object characteristics etc. But due to the variety of factors influence in the actual detection process, made the ground penetrating radar signal with noises, affected the radar signal’s quality, caused harmful effect in accurate judgment for the underground geological structure. In order to maximize reduction of the underground geologic configuration, obtain good quality radar signal is the key. So it is necessary to deal with the noise for radar signal.Currently, there are many removing noises methods were applied to radar signals, however the most commonly method is based on wavelet transform. In the wavelet transform de-noising approach, the wavelet threshold de-noising method was used widely and has de-noising effect. The de-noising effect of the wavelet threshold de-noising method is related to the wavelet threshold function and the threshold value selection. The different threshold function and the combination of the threshold choice have different effect for de-noising effectiveness. From the wavelet theory, several kinds of common threshold de-noising method were studied in this paper, such as soft threshold de-noising method, hard threshold de-noising methods, Garrot threshold de-noising method. Comparison analysis the de-noising effectiveness for these de-noising methods in different threshold condition. And their advantages and disadvantages were compared by the simulated data. The disadvantages of soft threshold de-noising method were improved. In addition, singular spectrum analysis (SSA) is mainly used in the seismic signal processing. In addition, singular spectrum analysis (SSA) is mainly used in the seismic signal processing, and made the time series period decomposition strongly associated with the time scales. It is based on singular value decomposition for the time series of matrix. The basic principle is reduced rank, represent the original signal singular by refactoring the low-order modes. The SSA was used to the radar signal de-noising processing in this paper, and the simulation data showed that the method has good de-noising effect.Finally, the ground penetrating radar echo signal imaging software was compiled based on MATLAB GUI programming, The software has the advantage of friendly interface, simple operation and strong practicability. in the software implementation several common wavelet threshold de-noising method and singular spectrum analysis de-noising method were performed. Different threshold was chose to de-noising processing when using threshold de-noising method, and different order number was chose for signal reconstruction when using singular spectrum analysis method. The wavelet threshold de-noising method and singular spectrum analysis de-noising method all had de-noising effect by processed the outdoor radar data. However, the de-noising method of singular spectral analysis had better de-noising effect than the wavelet threshold de-noising method. |
PDF Full Text Request