Ground penetrating radar is a high-resolution geophysical method which uses the reflections of radio waves emitted to the earth to depict structures and objects under the ground. All of the materials under the surface have certain electromagnetic properties that indicate their electromagnetic radiation permeability. The electromagnetic wave that returns from the earth contains information such as dielectric constant, magnetic permeability and electrical conductivity of objects inside the earth, which construct the frequency equation of the reflected wave. Changing the electromagnetic characteristics of different layers in the ground results in a change in the power returned to the radar receiver. These changes are visible in the radar receiver by changing the brightness of the constructed image. In this research, to investigate, the effect of different parameters of target such as shape, size and dimensions, depth and physical properties of the host environment on the receiver radar signal. T he GPR responses of many types of objects are evaluated in geotechnical applications. Also, the GPR responses are simulated in three software environments (Reflexw software, Gprmax and extended numerical of finite difference in MATLAB). Comparing these methods, we have shown that the extended finite difference numerical simulation method has more clear response to the ceil and floor of targets. Also using the hyperbolic mathematical equations, we have been extracted the relationships between geometric parameters of a horizontal cylindrical objective with parameters of GPR response. Hence , we have obtained the linear relationships between changes in diameter and depth of the cylindrical target and the dielectric passage rate of the host environment. This relationships can be used to determine the geometric characteristics of buried cylindrical targets under the ground surface. Finally, we examine the mathematical and theoretical results using the idea behind the Hough transform to detect hyperbolic response of voids and propose an algorithm which reduces the processing time comparing with the current methods which can accurately measure the radius, depth, and distance of targets. We also examined the proposed algorithm for real data, which yielded desirable results. The proposed algorithm is designed in such a way that all stages of the identification and estimation of the geometric parameters of buried objects are automatically estimated with minimal user intervention.Key Words : Ground penetrating radar (GPR) , Void , Hough transform , Forward modelin g, the geometric characteristics of cylindrical voids .
