russia_flag_small

italy_flag_small

uk_flag_small

japan_flag_small

usa_flag_small

Sadly, mines have become harder to detect over the years as manufacturers sought to make them undetectable by simple metal detector techniques.  At right is an anti-personnel landmine, less than 70 mm in diameter, and made of plastic with only small bits of metal (pins and springs).

There are two main methods used today to search for buried mines.  Metal detectors are in common use, and luckily the highly lethal fragmentation mines, being metal, are relatively easy to detect (although much care must be used to avoid trip wires while searching).  Mine detection metal detectors are extremely sensitive, rugged, and reliable versions of the beachcomber-type machines that most people are familiar with.  As with beachcombing, the proximity of metal to the search coil triggers an audible alarm.  But, battlefields are notoriously full of metal scraps, and many locations have naturally metal-rich soil or rocks.  At any hint of a metal detector signal, the deminer must lie down and use a spike to locate the cause of the signal, and approximate its shape.  False alarms are ubiquitous, and it is dangerous work - with a casualty rate frequently cited as high as one deminer per approximately 1000 mines cleared.

pmamine
metal detector

Photo from
Landmine Monitor

Many other ingenious methods have been tried but none have found wide usage among de-miners in the field.  Any new method needs to satisfy several critical criteria:

(1) A high probability of detection for any type of mine
(2) A low false alarm rate even in the presence of clutter
(3) A low cost to be affordable by underfunded de-miners in war-ravaged countries
(4) User friendliness that allows usage by local non-expert personnel with minimal training.

Conventional impulse ground penetrating radar (GPR) does a reasonable job with the first two criteria above, but is expensive (at around 30,000 USD or more per system) and this has limited its use.  In addition, the raw data do not present an easily-interpreted image to any but highly-trained users.  Many clever data processing and target recognition algorithms, as well as friendly user interfaces, have been developed, but again at the expense of cost, speed, or compactness.

Microwave holography may be a technology that can achieve all four criteria.  It provides similar or better performance than GPR, at a fraction of the cost (only around 5000 USD per system), and provides rapid, real-time, visually interpretable holographic images of buried objects directly from the raw field data.

The physics are simple.  Suppose a continuous, oscillating, emitted signal at the source (on the left of the figure below) is described by:

a0(t) ~ cos (2pft + q0)

where f is the frequency, t the time, and q0 is some unknown phase angle.  The wave is reflected by some buried object, and will be going in the reverse direction.  At the location of the source, it will have a phase difference equal to twice the distance d to the object (because it goes there and back), divided by the wavelength, which is equal to V/2pf, where V is the velocity in the medium.  Thus arriving back at the source, the reflected signal is:

ar(t) ~ cos (2pft + 4pdf/V + q0).

The interference signal at the source is the product of the incident (transmitted) and reflected amplitudes averaged over time since it is a continuous signal.  The unknown phase angle q0 disappears and the resulting signal has amplitude A which depends on distance d and frequency f, but  not time t, and is given by:

A(d) = cos (4pfd/V)

A diagram of this interference function as a function of the obstacle distance d, for several different frequencies f is shown in the figure below. The wavelength of the oscillation with depth decreases as the frequency increases.

interference

The RASCAN system developed by the Remote Sensing Laboratory of Bauman Moscow State Technical University is the only production holographic radar system in the world.  The figure below (left) shows the small, lightweight scanning head being used on a sand bed which contains a buried aluminum plate, as illustrated in the adjacent cartoon (below, center).  The plate was inclined so that it dropped by 85 mm across 300 mm. This experiment, therefore, simulates the simple mathematical model described above for variation in the interference signal A with changing depth d.  As can be seen in the photo, the sand is covered with a microwave-transparent plastic sheet ruled with lines 10 mm apart.  The scan is performed manually by moving the head along each of these lines. A small wheel behind the head measures the distance traveled by the head along the scan line, and at 5 or 10 mm intervals, the interference signal is measured with two polarization directions, parallel and perpendicular to the scan line.  The polarizations can be very useful in identifying the nature of the reflecting object (e.g. for detection of trip wires, trigger wires, or hidden antennae), but will not be discussed here.

masscan

aldiagram

almcol

The holographic Image appears line-by-line as the scanning proceeds, and is built from simply the unprocessed interference signal plotted in gray scale.  Five discrete frequencies are recorded simultaneously to ensure that a target at arbitrary depth will be detected (since the phase difference between the incident and reflected signals continuously wraps around - producing “blind” depths where the phase difference is an integer multiple of 360o.  The false color image below shows a sum of a 4.0 GHz image in shades of red, a 3.9 GHz image in shades of yellow, a 3.8 GHz image in shades of green, 3.7 GHz image in shades of blue, and a 3.6 GHz image in shades of violet.  The edge of the inclined plate was about half way down the image, so that the upper half of the image is simply background.  The oscillatory signals from the inclined buried sheet are clearly seen in the lower half of the image. The image closely resembles the optical interference pattern produced by a slightly inclined glass plate.  We often call the characteristic pattern from inclined surfaces the "zebra effect".  By averaging several adjacent scans together, a more precise definition of the depth dependence of each of the five frequencies can be obtained as plotted at right, and it agrees well with the mathematical model described above, and redrawn here for comparison.

interference

aldscan

Theoretical Radar Response

Measured Radar Response

To test microwave holography for the specific application of detecting landmines, numerous images of various inert landmines and landmine simulants have been recorded in test beds at Bauman, the University of Florence, Enviroscan, Inc., and Franklin and Marshall College.  The pictures at the bottom of the page show an experiment in Florence in which several landmine simulants (left) were buried in a test bed containing slightly moist silty sand soil (middle).  The test bed was scanned manually using both microwave holography and also conventional impulse radar.  The holographic radar image is shown at right, and displays two clear yellowish circular targets representing the metallic mines.  In the top right corner of the image is a completely non-metal mock mine, and it is also clearly seen, and with a distinct bluish false colour indicating a different phase change.  In the lower right is a PMA-2 mine with its own distinctive phase change producing a purplish false colour.  A buried wire is clearly visible as well.  The oblong red shape is compacted sand from an experimenter’s knee.

minesphoto

florencemines

minesmcol

Please also visit our Gallery of Radar Holograms

For further technical details concerning the history and physics of holographic radar, please click below to read a preprint of a paper that has been accepted for presentation at the GPR2010 conference in Lecce, Italy.

Razevig_et_al

Looking for Landmines with Holographic Radar

sse2010_chartermark_sml_red
[Home] [Technical]