NEURAL DETECTION OF PIPE THROUGH GROUND PENETRATING RADAR IMAGES

In this present study, a processing chain for the spatial analysis of the data recorded by ground penetrating radar (GPR) is presented. The detection and localization of buried pipes is carried out by making use of the knowledge that a buried pipe gives rise to a hyperbolic mark in GPR images. The image analysis is performed by a suitably trained simple neural detector after some pre-processing steps that aim toward the enhancement of the buried pipes. The algorithm has been tested on actual GPR images and compared with the information extracted by a trained human operator, and the agreement is extremely satisfying and the possible advantages to exploiting some sort of “spatial diversity” by combining the analysis of data simultaneously recorded by different antennas and are presented.