The Research Of Different Target Detection And Data Processing Based On Ground Penetrating Radar

As a geophysical detection technology, Ground Penetrating Radar(GPR) detectiontechnology is applied and developed widely in the underground target detection field becauseof its non-destruction, rapidness, high-precision, easy operation and so on. GPR is oftenunable to standardize for the detection target and the detection environment quantitatively inpractical work, so it is usually empirical for the echo imaging recognition, data processingand data interpretation of GPR.It has been brought many difficulties for the GPR workers due to the diversity oftargets and the complexity of detection environment, in the different target echo imagingrecognition and data processing. The underground different target detection analysis and echodata processing are researched specifically in the paper in connection with the existingproblems of GPR work.In the different target detection research, on the basis of the analysis and studyof GPR principles and detection methods, the five different target circumstances such asthe underground medium with or without target, target different depth of the target,different size target, the target of different materials and the target of different undergroundmedium are have been analyzed exhaustively, and besides the target model has been analyzedin detail and large number of physical target has been detected in each target case. Thecombination of the theory and practice, afterwards, the specific characteristics of the radarimage of the target and the degree of anomaly have been analyzed and summarized , and thecorresponding conclusions are given. The conclusions have a important guidingsignificance of GPR target recognition, All of this provides a reliable basis for GPR workersin target extraction.In the data processing research, there are some interference wave existed inevitably because of the complexity of the detection environment, on the basis of the analysis of the interferencecomponents of the echo data. The echo data preprocessing and the de-noising processing havebeen researched, and the mean filter method and median filtering method and also wavelettransform have been separately used for de-noising the measured GPR image data. It isproved that wavelet transform is better than the other two methods by experiment, which notonly achieves the purpose of suppression of clutter, but also retains the targetfeature information. On this basis, an energy statistical monitoring method has been adoptedfor target feature extraction, and the extraction effect is obvious. So it lays a scientificfoundation for ground penetrating radar target data interpretation.