Development of a Discrete Algorithm for Interpreting Ground-Penetrating Radar Data in Vertically Heterogeneous Media


Iskakov K. Tatin A. Glazyrina N. Kussainova A. Uzakkyzy N. Sagindykov K.
July 2025Multidisciplinary Digital Publishing Institute (MDPI)

Applied Sciences (Switzerland)
2025#15Issue 13

This study presents the development of a discrete algorithm for interpreting ground-penetrating radar (GPR) data in vertically inhomogeneous media for the diagnostics of road structures. Experimental data were obtained using an OKO-2 GPR system, followed by primary radargram processing using the CartScan software. This included noise and interference filtering, as well as the initial estimation of the dielectric permittivity of detected layers. The resulting dataset was used to validate numerical algorithms for solving the forward and inverse problems of geolectrics. The proposed approach is based on minimizing a quadratic misfit functional between the calculated and observed values of the horizontal component of the electromagnetic field. The gradient of the functional required for optimization is obtained via the numerical solution of an adjoint problem. A discrete version of this problem was developed, which satisfies the properties of conservativeness and uniformity according to finite difference theory. The inverse problem reconstruction of dielectric permittivity is considered a non-destructive method for radargram interpretation. Assuming a piecewise-continuous medium structure eliminates the need for computing gradients at material interfaces. The proposed methodology enhances the accuracy and reliability of pavement condition assessment and holds practical value for road infrastructure monitoring.

dielectric permittivity reconstruction , digital signal processing , discrete method , ground penetrating radar (GPR) , inverse modeling , inverse problems , optimization methods , pavement diagnostics , radargram interpretation

Text of the article Перейти на текст статьи

Department of Computer and Software Engineering, Faculty of Information Technology, L.N. Gumilev Eurasian National University, Astana, 010000, Kazakhstan

Department of Computer and Software Engineering

10 лет помогаем публиковать статьи Международный издатель

Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026