TIME DOMAIN METHOD OF GPR SIMULATION IN DISPERSIVE MEDIA

Abstract

Underlying object detection is an imperative and yet exigent problem that remains to be Solved. The goal of this research is to propose some techniques for revealing buried objects. Two sophisticated Feature based techniques are developed. One algorithm applies the clustering method Based on the spectral trait vectors formed by the energy density spectra of homecoming Sensor signals, the idea at the back is to find out whether there are some out of sight patterns 'Among the spectral feature vectors. The other one is the detector technique that utilizes the energy density spectra of homecoming signals directly. These techniques would be Tested in various testing data sets composed from the vehicle mounted ground Penetrating Radar to estimate their ability to perk up the detection result and diminish the False alarm Rates. Both of them would prove to be useful in civilizing the revealing of Targets like under lying utilities presently for these cases mass of engineers still opt for destructive testing due to unavailability of unswerving technique for investigation of data composed through non-destructive approach. GPR can be helpful non-destructive testing approach but is constrained by the fact that it requires proficiency professional for its data elucidation and majority of times it is done by hand without backing of computer. Therefore results obtained by manual or visual elucidation of GPR are often subjected to operator proficiency and sometimes may guide to mistaken interpretations. Without any data interpretation software elucidation of GPR data is highly time consuming. This research work attempts to trounce the drawbacks of non-destructive analysis using GPR for health monitoring and identification of type of buried objects and anomalies present in it using near to accurate mathematical modeling of the transmitted wave from ground coupled GPR systems and application of match filter and Fourier transform. Various image and signal dispensation techniques like FDTD have also been engaged for detection of different type of buried objects. The developed methodology would be validated using data collected on investigational setups and pits. While the percentage accuracy of fortitude of type of buried objects may be less than 9 %. More accuracy could be achieved if air coupled GPR system could be used for precise determination of dielectric constant of medium.