3D Reconstruction Hyperbolic Image of Buried Object Using GPR B-scan Images for Mapping Purposes

Abstract

Ground Penetrating Radar (GPR) is a non-destructive evaluation and imaging tools for shallow subsurface investigation that based on the transmitting and receiving of electromagnetic when moving along the antenna along the surface. For an unexperienced GPR user, it may have difficulties to understand the raw data of the GPR images due to the data complexity and resemblance of the hyperbola pattern. Thus, this aim of the study is to reconstruct 3D representation of hyperbola images of buried object for underground mapping application. In this framework, Synthetic Aperture Focusing Technique (SAFT) will be used to reconstruct 3D images. Firstly, the raw data of B-scan images undergo zero-time correction and background removal. Then, the projection of each B-scan images to reconstruct 3D image. The obtained 3D images then were stacking together. Finally, the 3D interpolation technique of the stacked images were implemented to obtain 3D reconstruction of hyperbola images. A series of experiments has been conducted on the collected B-scan image of sphere metal ball. The obtained results shows that the SAFT technique not only able to reconstruct the 3D images however it also reveal the location of the buried object that represent by voxel point of 3D images. Thus, the proposed method using SAFT method shows promising 3D reconstruction technique of hyperbola images of buried objects.