Recovering the structure of a layered soil, including layer thickness and dielectric permittivity, using the interfaces and objects backscatter detected in GPR B-scans
M. Ardekani, P. Druyts, S. Lambot, A. De Coster, X. Neyt
{"title":"Recovering the structure of a layered soil, including layer thickness and dielectric permittivity, using the interfaces and objects backscatter detected in GPR B-scans","authors":"M. Ardekani, P. Druyts, S. Lambot, A. De Coster, X. Neyt","doi":"10.1109/ICGPR.2014.6970454","DOIUrl":null,"url":null,"abstract":"It is well-known that point scatterers appear as hyperbolas in ground-penetrating radar (GPR) B-scans and that the layer interfaces appear as horizontal lines. In this paper the shape and location of the hyperbolas, together with the location of the layer interfaces, are used to estimate the soil dielectric permittivity for a layered soil. For this, a procedure composed of following steps is used: (1) reflection detection, (2) hyperbola detection, (3) refinement of hyperbola parameters and estimation of the corresponding scatterer location and soil effective dielectric permittivity, and (4) computation of scatterer depth and layer permittivity taking into account the properties of the upper layers. The reflection detection step takes the GPR B-scan as input and produces a `reflection binary image' as output. The binary image highlights reflections of interest, which includes the hyperbolas and the soil layer interfaces. The effective soil dielectric permittivity is estimated by fitting a theoretically computed hyperbola to the `reflection binary image' for each reflection detected. Then, hyperbola parameters are refined by optimizing a cost function which is computed on the original Bscan for each detected hyperbola. Finally, the soil layer dielectric permittivity and scatterer depth are derived from the hyperbola parameters, taking into account the properties of the upper layers. The procedure is applied to simulated data, showing good accuracy in soil dielectric permittivity estimation and high computational efficiency.","PeriodicalId":212710,"journal":{"name":"Proceedings of the 15th International Conference on Ground Penetrating Radar","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2014-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 15th International Conference on Ground Penetrating Radar","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICGPR.2014.6970454","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
Abstract
It is well-known that point scatterers appear as hyperbolas in ground-penetrating radar (GPR) B-scans and that the layer interfaces appear as horizontal lines. In this paper the shape and location of the hyperbolas, together with the location of the layer interfaces, are used to estimate the soil dielectric permittivity for a layered soil. For this, a procedure composed of following steps is used: (1) reflection detection, (2) hyperbola detection, (3) refinement of hyperbola parameters and estimation of the corresponding scatterer location and soil effective dielectric permittivity, and (4) computation of scatterer depth and layer permittivity taking into account the properties of the upper layers. The reflection detection step takes the GPR B-scan as input and produces a `reflection binary image' as output. The binary image highlights reflections of interest, which includes the hyperbolas and the soil layer interfaces. The effective soil dielectric permittivity is estimated by fitting a theoretically computed hyperbola to the `reflection binary image' for each reflection detected. Then, hyperbola parameters are refined by optimizing a cost function which is computed on the original Bscan for each detected hyperbola. Finally, the soil layer dielectric permittivity and scatterer depth are derived from the hyperbola parameters, taking into account the properties of the upper layers. The procedure is applied to simulated data, showing good accuracy in soil dielectric permittivity estimation and high computational efficiency.