Interpolation techniques to visualize underground structures with high accuracy using fewer GPR data

Abstract

Summary Ground Penetrating Radar (GPR) is widely used to collect the data from near surface depth. GPR data allow the users to visualize and interpret the underground structures quickly and with high accuracy. The obtained accuracy depends on such data collection parameters as sampling rate, trace and profile intervals. In order to visualize the underground structure with high accuracy, the sampling rate should be selected high and the trace and profile intervals should be selected small. Due to the high sampling rate and small trace and profile intervals the size of the collected GPR data becomes bigger. However, collecting the big data will increase both cost and time loss. In this paper, a methodology and Mean interpolation technique are proposed to reduce the cost and time loss by decreasing the size of collected data as much as possible and to increase an accuracy of underground structures by applying the interpolation techniques to the collected raw data. The proposed Mean and standard Cubic and Linear interpolation techniques were implemented on real GPR data. The obtained results showed that the similarity ratio between original and interpolated GPR data is about 95–98%.