Study On Detection And Recognition Technology Of Road Underground Cavity Based On GPR

As a rapidly developing nondestructive testing technology,ground penetrating radar detection has been widely used in urban road detection because of its convenience and high efficiency.It not only realizes the data collection of road collapse disaster detection and daily maintenance detection,but also meets the demand of urban road rapid detection.But most of the current research is limited to underground pipelines,voids,blind objects and others.In the qualitative interpretation of defects,there are few quantitative studies on the size of underground voids,which can not effectively evaluate the severity of underground diseases,so it is of great significance to use ground penetrating radar to carry out quantitative analysis and research on road underground cavity diseases.Starting from the model test of the cavity disease,this paper introduces the digital image processing technology into the processing of the radar three-dimensional detection image,and studies the radar image area of the cavity disease with MATLAB program.The main research contents and achievements are as follows:(1)The physical model of the cavity in the roadbed is designed and made,and the three-dimensional detection is carried out by GPR,and the radar data obtained is processed by radan7 software,the radar image slice map of the measured target body is obtained,the radar image boundary curve is drawn by Origin software and the area is calculated,the results show that GPR is three-dimensional Detection can effectively determine the distribution,shape and size of cavities.(2)The influence of the embedded depth,antenna frequency and dielectric material on the radar image characteristics is analyzed.At the same embedding depth,the high-frequency radar antenna can more truly reflect the shape information of the cavity due to its high resolution;at the same antenna frequency,because the electromagnetic wave signal gradually weakens with the increase of depth,noise interference and filtering means and other factors,the shallower the embedding depth of the test piece is,the closer the radar image contour is to the actual situation.Because the signal strength of the electromagnetic wave is determined by the dielectric constant difference between the test piece and the surrounding medium,and the dielectric constant difference between the metal medium and the sand medium is very large,the reflected wave signal received by the receiving antenna is stronger and wider,so theimage area of the metal hollow sphere under the same conditions is larger than that of the plastic hollow sphere,which is different from the actual cross-sectional area.(3)The automatic recognition and calculation of radar image area with different shape cavities are realized by MATLAB programming.The basic principles and algorithms of image binary processing,threshold selection,morphological image processing and other methods are introduced.GUI interface is made to show the processing process and calculation results of the image directly,which is verified by the recognition and calculation of radar images of semi grouted prestressed pipeline The program is efficient and accurate.The research shows that the automatic recognition technology of radar hole disease image based on Matlab can recognize the area of hole image accurately,which not only has important guiding significance for the evaluation and reinforcement scheme of the disease under the road,but also is suitable for the recognition and quantitative calculation of the radar image of various shapes and depths.