3D Digitization of Sub Surface Features Using 3D GPR

Ground Penetrating Radars (GPR) finds its applications in civil engineering, archeology, earth sciences, military and environmental studies. This paper proposes a framework for using GPR as a digitizing technique that can make 3D digitization/measurement of the buried objects. Currently, the 3D radar volume data is acquired by dense subset of parallel 2D radar profiles. However, it results in a very large amount data and related processing time, and needs to be avoided due to the significant computer resources and processing time required to process the data. In order to achieve simplified and faster processing as well as to reduce memory requirements, it is often useful to convert the original image data to a more compact representation. The conversion removes as much of the redundant information as possible. Then the data is processed by removing noise initially and then eliminating undesired presence of ground surface echo. Further, propagation losses are compensated followed by segmentation and skeletonizing. However the accuracy of the 3D measurement depends upon number of parameters such as antenna design, scan paths, scan spacing and frequency which needs to be evaluated in detail. The framework presented in this paper would help to measure the buried objects accurately using 3D GPR.