Study On 3D GPR Road Disease Change Automatic Recognition Technology Based On Face Recognition

Urban road is an important infrastructure for urban development.It not only plays an important role in urban safety,but also affects the economic development of cities.However,in recent years,road subsidence has become more frequent in major cities throughout the country and causes great security risks and economic losses to the community and residents.The collapse of urban roads is often caused by the gradual expansion of vacancy,which is concealed,sudden and dangerous.Now the road cavities have become the key and difficult point of urban governance.Among ways of road detection,ground penetrating radar(GPR)has been widely used in urban road disease detection for its advantages of convenience,visualization,high resolution and so on.However,the road detection work is mainly based on two-dimensional detection,although it can reflect some information of the underground structure,but the information obtained by 2D detection is less and the coverage area is small.So,it is difficult to accurately explain the spatial location,shape and trend of road diseases.In order to solve problems described above,this paper studies the automatic recognition technology of 3D GPR data changes based on face recognition for road detection.It automatically identifies when the road detection data of 3D GPR is mutated and realizes early warning of road collapse,then avoiding huge losses caused by road collapse.Firstly,we starting from the Maxwell equation group,the spatial distribution characteristics of the six components of the electromagnetic field are analyzed in detail by using the finite difference time domain method i n the full space uniform model.According to the analysis results,we choose the Ex-Ex energetically receiving mode,which has the strongest energy on the XOY plane,to conduct 3D detection simulation.Furthermore,common road cavity and crack disease models are constructed for three-dimensional detection simulation.The spatial advantages of 3D detection are analyzed in depth by combining with road detection experiments.The results show that 3D detection can obtain the spatial information of depth,strike and shape of the targets such as voids and cracks.The interpretation results are more accurate.Finally,aiming at the interpretation problem of the large volume of data obtained from 3D detection,we draw on the ideas and algorithms of face recognition.The adjusted principal component analysis algorithm is used to extract the main characteristics of the data and draw the curve of the ground penetrating radar data changes.When the detection data is abrupt,the data change curve will also appear abnormal peak,so as to realize the automatic identification of the data change of the ground penetrating radar and carry on the road cavitation catastrophe warning and the collapse prevention.