A simple Approach for radargram Pattern Recognition: Identification of Metal and Non Metal Object

Ground Penetrating radar is an active geophysical methods which is mostly used to interpret objects buried under the surface or to determine the distribution of shallow subsurface sediments. Radargram as one of digital recording signals forms on GPR is often analysed using variety of mathematics and physics attributes to determine subsurface objects. The recorded signal is the reflection of the electromagnetic wave with a specific parameter that depends on the antenna frequency. Metals such as steel (Fe3C), iron (Fe2O3) and aluminium (Al2O3) can be visibly distinguished from non-metal on the surface, but when the object lies below the surface, it is very difficult to perform identification for each object, this research is done to see changes in patterns on radargram (Pattern Recognition), by performing mathematics and physics approaches to identify changes of frequency, amplitude, velocity and density of the recorded signal wiggle, this study uses survey data with objects containing metal buried deep as 1-2 m and non-metal objects (sedimentary rock) that are in the same relative depth. Each wiggle processed by time-shifting (MST), de-wow, Gain, Filtering, background removal until the object become clearer than before (until the best resolution obtained). Visual interpretation of peak-trough which formed on the wiggle function of distance and depth can help accelerating the identification process so that the distinguish process between metal and non-metal can be easily performed. The results appear on the hyperbole pattern radargram for metal objects but not for non-metal objects