Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Automatic Pipe Detection Using Fully Polarimetric GPR

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  Paper number  032343,  2003 ASAE Annual Meeting . (doi: 10.13031/2013.14036) @2003
Authors:   H. Youn, C. Chen, L. Peter, Jr.
Keywords:   Agricultural drainage pipe, Automatic detection, Neural network, GPR

An automatic pipe detection algorithm using a two-step Artificial Neural Network (ANN) scheme on ground penetrating radar (GPR) data is presented. The Step-1 ANN is trained to recognize specific time-domain waveform associated with a buried pipe. Once the target waveform is detected by the Step-1 ANN, the Step-2 ANN, is then applied to compare the spatial variation pattern as a function of antenna position with feature patterns. The detection performance of these algorithms in the presence of different signal-to-noise and signal-to-clutter ratios was investigated via Monte Carlo simulations. The estimation of pipe orientation using fully polarimetric GPR data will also be demonstrated. The fully polarimetric feature is useful in discriminating response from a point scatter such as a rock from a transverse pipe in a single GPR scan. This orientation information is utilized to normalize the spatial variation pattern before applying the Step-2 ANN when the scan direction is not perpendicular to the pipe. This helps discriminate pipe scattering from that caused by a subsurface layer since the single-polarized GPR data from a layer may appear to be similar to that from a pipe oriented parallel to the scan direction. Finally, The developed ANN algorithm was applied to detect drainage pipes from actual GPR data collected at farmlands with a variety of soil types. The results indicated that the said 2-Step ANN algorithm was able to effectively detect most of the pipes buried up to one meter

(Download PDF)    (Export to EndNotes)