Study On Intelligent Interpretation Method Of Radar Image Of Underground Pipe Network

In the process of modern urban construction and management of,the laying of underground pipeline is one of the important basic data,thus to seek effective method of detecting the underground pipe network,fast for underground pipeline information is particularly important,the external approach for detection of ground penetrating radar(GPR)technology applications and the most widely used its nondestructive,rapid,and can quickly detect the advantages of high resolution,but also needs to improve identification precision rate,GPR echo images automatic interpretation difficult problems,such as,according to the above problem,this paper puts forward the combined with digital image processing technique and artificial neural network model,Based on the forward numerical simulation of GPR detection of underground pipeline,the radar spectrum identification and feature extraction are carried out,and the inversion study of underground pipeline information is completed,so as to achieve the purpose of intelligent interpretation of radar images of underground pipeline.The main research results are as follows:(1)the forward numerical simulation study of underground pipeline shows that the greater the difference in dielectric constant between buried medium and pipeline material,the stronger the reflected signal in radar image and the clearer the hyperbolic shape;The shallower the pipe is buried and the larger its diameter,the higher the top of the hyperbola,the greater the curvature.(2)key feature parameters of GPR spectrum are extracted based on digital image processing technology through the identification research of spectrum.(3)through the inversion study of underground pipeline information,the ga-bp neural network pattern recognition algorithm with high inversion accuracy is selected to obtain the internal relationship between the characteristic parameters of spectral maps and the information of underground pipelines,which has a very important guiding significance for the spccific interpretation of radar echo images in practical engineering practice.