A new algorithm for imaging ground‐penetrating radar (GPR) data follows from the theory of the generalized Radon transform (GRT), which was developed and has been used extensively for seismic imaging. The algorithm enables separate reconstructions (to first‐order accuracy) of subsurface permittivity and conductivity images. A pseudovector inverse adapts the original scalar Radon transform theory for vector electromagnetic data. Synthetic examples show that the algorithm can image plastic and metallic pipes buried in a half‐space. In the examples, the half‐space is 100 ohm-m with a relative permittivity of 10. The pipes are 2 cm thick and 15 cm in outer diameter and are buried at a depth of 2 m. The plastic pipes are assumed to be pure insulators, whereas the metallic pipes are assumed to be pure conductors with a conductivity value of 106 S/m. The transmitter waveform has a peak frequency of 200 MHz with a 500-MHz bandwidth. The results show that quantitative images are obtained for plastic pipes, whose scattering is weak (so the Born approximation is accurate). Good images are also obtained for metallic pipes. As with most imaging algorithms, a sharp image requires good estimates of the properties of the background medium.
Three-dimensional passive seismic waveform imaging around the SAFOD site, California, using the generalized Radon transform
