Study Of Shallow Underground Target Detection Imaging Algorithm

Ground Penetrating Radar (GPR) technology is a radar system to obtain the information of underground targets by utilizing the high frequency electromagnetic waves. Its working principle is to use the reflection and scattering produced by electromagnetic waves propagating onto the discontinuous of the medium. According to the information of received waveform, amplitude and time delay, we can extract the location, shape and nature of the targets to realize the image and position of the shallow underground targets. As one of the shallow underground non-destructive detective technology, GPR technology is widely used in military and civilian fields because of its quick detection speed and high space resolution. The imaging algorithm is one of the key technical indicators to realize the ground penetrating radars. Among the existing imaging algorithms, the time-domain synthetic aperture imaging algorithm is widely used due to the simple structure and be easy to implement. However, the time-domain synthetic aperture imaging algorithm suffers some disadvantages such as time-consuming computations, low computational efficiency and poor noise suppression. Therefore, some new imaging algorithms become eagerly desired for advanced radar systems.In this thesis, we first made a brief review of the theory of shallow underground targets detection, including the working principle of ground penetrating radar, the mode of electromagnetic waves transmission in the ground, the system implement and the imaging technology of the ground penetrating radar, etc, which provides a theoretical basis for the following study of this thesis.Secondly, we study the conventional back projection algorithm. Through theoretical studies and simulation experiments, we discuss and analyze the traditional back projection algorithm. Also we discuss the impact of different factors on the results of imaging. The results show that there are some inadequacies in traditional back projection algorithm.Then, we propose a new imaging algorithm called virtual echo projection algorithm. The proposed method can effectively suppress the environment noise, especially colored noise. Additionally, the method is not limited to the linear aperture radar for imaging. Through numerical simulations and experiment’s results analysis, we verify the effectiveness of the proposed method. The proposed method works well in the condition of non-linear aperture and in suppression the white noise and the colored noise.Finally, this thesis does a summary of the work, and analysis the advantages and disadvantages of the work, which will provide a reference for the subsequent studies.