Research Of Complex Signal Analysis Applied In Data Processing Of GPR

The technique of ground penetrating radar (GPR) is a high-tech technique for examination, which has a lot of advantages such as high degree of accuracy, non-stop operation, non-destructive, economical, rapid survey, audio visual and so on. We can identify abnormalities in buried targets with GPR by transmitting electromagnetic waves of high frequency to buried targets and receiving the backward waves.It is very necessary to research on signal processing of GPR for improving performance of GPR with the applications of GPR increasing. Now there are more and more researches on signal processing technique.In recent years, complex signal analyzing method has been use to treat seismic prospecting data, and the outcome is preferably. Base on the characteristic of GPR wave similar with seismic wave, some scholars introduce complex signal analyzing method to GPR data processing, they draw a conclusion that complex signal analyzing method is an effective data processing method for GPR, which have many advantages such as several parameters, high degree of accuracy, high distinguishability, being able to distinguish the weak signal in deep place and so on. In this paper, some problems in traditional ground penetrating radar(GPR) complex signal analyzing extraction to use wavelet transform,generalized S transform,generalized Hilbert transform and complex signal transformation are investigated.First, this thesis demonstrates the history and the state-of-the-art of GPR, and then explains the feasibility and advantages of the complex signal analyzing technique applied to GPR data processing, introduces the basic idea and the mainly mathematics tool about the complex signal analyzing method, analyses the Hilbert transform method which is used to construct the complex signal. Due to the disadvantage of the Hilbert transform that the method is very sensitive to random noise, the author introduce some processing methods such as wavelet transform,generalized S transform,generalized Hilbert transform and complex signal transformation into the complex signal analyzing technique of GPR. The contents of this paper are as follows:1,One method for having instantaneous characteristics of GPR signals in wavelet domain is put forward, which obtains directly the instantaneous characteristics of GPR signals by taking the characteristics of both the real part(real signals, namely GPR signals) and the imaginary part(the Hilbert transform of real signals) of wavelet transform. The method has the functions of frequency division and noise removal. Choose the "mother wavelet" function is important when wavelet-based method is used to compute the instantaneous parameters. The GPR direct wave is analyzed; the wavelets called "GPR wavelets" are constructed to well simulate the GPR direct wave. At the same time, the method of how to construct a GPR wavelets in time domain is also discussed. Comparison with the conventional Hilbert method, the instantaneous parameter computed by wavelets called "the wavelets approaching to GPR wavelets" has higher antinoise function.2,The S transform can be regarded as a hybrid of the STFT and WT, it can keep the resolution relevant to frequency when S transform is used to detects the non-stable signals. A generalized S transform is achieved by adding an adjusted factor to the Gaussian window function of the normal S transform. The adjusted factor changes the velocity in which the width of the window function varies inversely with the frequency. The general S transform (GST) is achieved in the GPR data analysis. The general "three instantaneous" attribute(namely general instantaneous amplitude, frequency and phase)can be extracted from single frequency complex signal which is produced in S domain. The instantaneous amplitude of single frequency which is computed by general S transform is sensitive to reflected energy; this method has the potential to detect the location of reflection interface.3,Generalized Hilbert transform (GHT) which can be considered a windowed traditional HT extends HT. GHT can be implemented using different order, adjustable window shape and widths, whereas traditional HT is just one of many possible implementations for a box window with infinite length. Comparison with the conventional Hilbert method, the GHT is superior both at noise elimination and edge detection. The outcome is reliable and precise when using appropriate order and operator length.4,The complex signal transformation method this thesis introduced is an effective method which uses complex signal analyzing envelope and normalized phase information to improve the temporal resolution and to increase S/N. Using this method, the band-width of frequency broadens, so the GPR signal resolution can be increased. At the same time, the random-noise will be suppressed, and increase S/N. Complex signal transformation method is very simplest also does not require much mathematical computation, and the only parameter selection required is the time window length. We concluded that the time window length is an important parameter for complex signal transformation processes, both random-noise suppression and temporal resolution improvement is better when choosing the appropriate time window length.By comparing all these techniques, it summarizes both advantages and disadvantages of them and also points out the application conditions of them. According to the result of theory data and real data, the efficiency of the method has been demonstrated. Especially, in the case of signal to-noise ratio is very low. At last, the author points out the limitations of complex signal analyzing method of GPR, and also discuss the prospect and orientation of complex signal analyzing method of GPR.