DeTeC  Demining Technology Center

GPR basic principles
Description of the GPR system used by DeTeC
Commercial GPR products
Some publications on GPR
Link to GPR related Web pages

The prototype DETEC-2

GPR basic principles

Any dielectric discontinuity in a propagating media, as for example the presence of an object, will cause a reflection; its intensity will be higher with increasing difference between the dielectric coefficients. Typical soils permittivity vary from about 4 to 40, wet soils having a higher permittivity than dry ones (water has a permittivity of 80). Permittivity of plastic objects is in the range of 2 to 6.

Typical pulses have a width of the order of a nanosecond or less, with rise time of some hundred of picoseconds, which correspond to a frequency spectrum of some hundreds of MHz to 1-2 GHz.

High frequencies are needed to achieve a good spatial resolution, but penetration depth of electric fields being inversely proportional to the frequency, too high frequencies are useless after some centimeters. Hence the choice of the frequency range is a tradeoff between resolution and penetration depth. A system working with an antenna whose center frequency is at 1 GHz is considered by experts as beeing a good compromise for anti-personnal mines, allowing to work at depths of 50 cm to 1 m in most soils with a resolution of the order of some centimeters.

Penetration depth will also depend on the nature of the soil which have different attenuation. For example desert sand has an attenuation of about 1 dB/m for 1 GHz frequency, clay having attenuation of 100 dB/m at the same frequency.

The reflected wave is sampled and digitized by an A/D converter. The region of interest is quite short, for example in sand, 1 meter depth corresponding to about 20 ns. Typically 512 points are taken through the region of interest, which correspond to a sampling rate far too high for standard converters. The solution is to repeat the generation of the pulse (typical repetition rates range from some tens of KHz to 1 MHz) and to acquire only one sample in each reflected wave, the sampling time being shifted by some tens of ps for each pulse. This allows to use standard and relatively low cost converters.

By moving the antenna along a line and taking regularly spaced acquisitions, it is possible to construct an image representing a vertical slice of the ground. These images show hyperbolas whose apex is located at the objects positions. These hyperbolas result from the fact that antennas have a certain aperture and capture all reflections coming from a cone-shaped area below them. These hyperbolas can be focused by software (migration algorithms) to obtain the real image of the objects.

An alternative to the most used (real pulse) GPR is to work with so called synthetic pulses. The response to pure sine waves of discrete frequencies of the spectrum of a theoretical pulse are measured. Using the appropriate signal processing algorithms the response to the theoretical pulse can be reconstructed.

That technology has the following advantages over the real pulse GPR :

- it is possible to have a better control on the shape of the theoretical pulse. Its bandwidth can for example be quite easily modified by software if needed by using a larger range of frequencies.

- antennas do not have a flat gain in their bandpass. If the response diagram of the antenna is known, it is possible to correct the amplitude of attenuated frequencies to compensate the loss in the antenna.

- if the GPR works in a noisy environment, it is possible to skip the specific frequency of the disturbing element (a portable phone for example).

The main disadvantage of this technology is that controlled oscillators able to change frequency quickly in the GHz domain are still quite expensive, compared to fixed pulse generator. But the use of new integrated components developed for the mobile communication domain will allow to reduce costs in the near future.

Some interesting publications on GPR

Books

Papers

Daniels D.J., Gunton D.J., Scott H.F., "Introduction to subsurface radar", IEE Proceedings, Vol. 135, Pt. F, No. 4, August 1988, p. 278-320

Daniels D.J., "High resolution radar for the detection of buried anti-personnel mines for humanitarian clearing operations", Proceedings of ISMCR'96 (6th International Symposium on Measurment and Control in Robotics), Brussels 9-11 may 1996, p. 542-551

Peters L.Jr, Daniels J.J., Young J.D., "Ground penetrating radar as subsurface environmental sensing tools", Proceedings of the IEEE, Vol. 82, No. 12, December 1994, p. 1802-1822

Some of our material (please check also our Home Page)

C. Bruschini, B. Gros, F. Guerne, P.-Y. Pièce, O. Carmona, "Ground Penetrating Radar and Induction Coil Sensor Imaging for Antipersonnel Mines Detection", GPR'96, Sendai (Japan), 30 September-3 October 1996, pp. 211-216.
(PDF version (304 kB); PostScript compressed (1.16 MB) as gzip or pkunzip)
Published as "Ground penetrating radar and imaging metal detector for antipersonnel mine detection" in Journal of Applied Geophysics 40 (1998) 59-71. Special Issue: GPR'96, ed. by M. Sato and R. Versteeg.

F. Guerne, B. Gros, M. Schreiber, J.D. Nicoud "DETEC-1 and DETEC-2: GPR Mine Sensors for Data Acquisition in the Field", SusDem'97, Zagreb (Croatia), 29 September-1 October 1997, pp. 5.34/5.39.