Research And Development Of Undergroundstate Detection And Software Based Onground Penetrating Radar

With the rapid economic development of the country,the state council has also stepped up its efforts to build urban infrastructures.Road construction and underground construction are also being carried out in full swing,such as the laying of underground pipes and the construction of roads.This also brings more hidden dangers of underground disasters.In recent years,there have been frequent underground disasters,light tunnel collapse and pipeline leakage during construction.Which threatens people's property safety and life safety.This requires people to increase their research on the non-destructive testing of underground target objects.Based on the theory,this thesis introduces the basic theory of ground penetrating radar(GPR)detection,including the working principle of GPR,the propagation characteristics of GPR and the detection method of GPR.At the same time,the article introduces some knowledge about image processing,introduces the algorithm of image filtering,and uses Matlab simulation to filter the radar waveform,compares the advantages and disadvantages of the algorithm,finally selects the appropriate filtering algorithm to filter the radar image.At the same time,the edge detection technology of the image is also introduced.The radar waveform is detected by using Matlab,the advantages and disadvantages are compared,and the edge detection algorithm is selected to extract the contour of the radar image.When the GPR waveforms are preconditioned,the target recognition can be done.The main research of the support vector machine(SVM)algorithm here is that it has more advantages than the ordinary neural network learning algorithm in radar image recognition because it can solve the problem of neural network local minima and network structure selection and other issues,while not requiring too much sample size.At the end of the thesis a large number of samples are obtained for experiments,mainly different target objects.A large number of sample waveforms are obtained by changing the depth of the target object and the distribution of the target object,and the wave velocity of the electromagnetic wave in the soil is calibrated.Through the software described above to write the software to process the radar waveform,and then get a specific location through the calibration of the wavevelocity,and finally find the error by the test data.In order to obtain the radar waveform in the real environment,we choose to probe at the experimental site of China Institute of Radio wave Propagation,obtain a large number of samples through the field exploration of complex multi-target,and provide a reference for its future production and construction.