METAL DETECTOR WITH IMPROVED DETECTION CAPABILITIES

TECHNICAL FIELD.

The current invention relates to metal detectors having improved detection capabilities, in particular the metal detector includes at least two coils capable of simultaneously receiving a signal from a target excited by a transmit coil and preferably wherein the two coils capable of receiving the target signal are of a different effective configuration.

The invention has particular relevance to any field where knowledge of the depth, size and orientation of the target in the ground is important, such as is the case when searching for dangerous military hardware (such as unexploded ordnance or land mines) or treasure.

BACKGROUND ART

It has been well established in the metal detector industry that objects can be identified based upon their physical composition due to phase or decay characteristics of receive signals produced when the object is in the field of the transmit and receive coils. There have also been disclosures that describe detectors with two or more receiving signal coils.

US patent 1812392 discloses a device for electromagnetically detecting and locating terrestrial conducting bodies having a plurality of receiving coils. This patent further discloses the step of measuring the ratio between the voltages induced in the receiving coils by the secondary field to determine the distance to the conducting body.

US patent 3471772 and similarly US patent 5721489 disclose a detector having a transmit coil and two receiving coils where the ratio of the two receive signals can be used to determine the approximate size and depth of the object to be detected.

US 4091322 discloses a detector for detecting pipe lines having three receive coils.

The metal detectors in the prior art that use two receive coils to process the received signals have at least three coils in the coil housing. (Two receive coils plus the transmit coil.) A typical embodiment from the art is shown in Fig. 1. Having three coils in the coil housing adds complexity, weight and cost to the coil housing design. In addition, although depth and hence target size information is available as previously disclosed from the ratio of signals from the two receive coils, the two sets of information from said receive coils has very similar form (on account of the similar effective configuration of the coils) and hence contains little new information regarding the target. For targets at depth - particularly those at depths larger than approximately the diameter of the largest of the coils - the magnetic field produced by the transmit coil is essentially vertical and can best excite eddy currents (or magnetic responses) in the target in the horizontal (x, y) plane, giving rise to a target dipole moment in the z (vertical) direction. Since both receive coils lie in the horizontal (x, y) plane, the manner in which the secondary magnetic field produced by the target eddy currents (or magnetic responses) will couple to each in a similar manner. Fig. 3 illustrates the calculated response of the first receive coil (RXl) and the second receive coil (RX2) as a coil arrangement, similar to that in Fig. 1, is moved over two targets (in this case located 15 cm below the plane of the coil arrangement). Target 1 is a target that can be excited to produce a dipole-like response only in the vertical (z) direction, such as a loop of wire with inductance L and resistance R which lies wholly in the horizontal (x, y) plane. It is well known among those skilled in the art, that a flat conductive disk such as a metallic coin similarly orientated to a loop, produces a response with similar spatial variation. Target 2 is a target that can be excited only to produce a dipole-like response only in the horizontal x direction (here defined to be in the direction of motion of the coil arrangement), such as a loop of wire or coin which

lies wholly in the vertical (y, z) plane. From Fig. 3 it is seen that the responses from RXl and RX2 to the said targets are very similar in form. It can be noted that the response to Target 2 is very similar to that which would be produced by two targets similar to and smaller in response than Target 1, but spatially separated at the same depth. The spatial variation of the response from RXl and RX2 provides, in itself, little new information about target structure or orientation.

There is another limitation to the existing art as it relates to measurement of target depth. For targets located at a distance farther than approximately the diameter of the larger receive coil in the coil arrangement, the ratio of the responses received from the target in the two receive coils RXl and RX2 varies only very slowly with distance of the target from the coils. This slow variation limits the amount of target depth information that is available, particularly in the presence of noise on the signal; the origin of the noise could be variation in the ground structure, external electromagnetic interference, or sources internal to the detector. The reason for the slow variation of the RX1/RX2 ratio with depth is that, far from the target, the secondary magnetic field produced by the target is approximately spatially uniform, so sampling by the two similarly-shaped, but differently sized (or similarly-shaped but different height), concentric receive coils produces a result which asymptotes to the ratio of the areas of RXl and RX2. (IfRXl and RX2 have a different number of turns, or different amplification gain, then the ratio is suitably modified.) The mono/mono curve in Fig. 4 shows the calculated variation of the ratio RX1/RX2 as a function of the depth of the target below a coil arrangement similar to that depicted in Fig. 1. The Dashed Curve in Fig. 4 represents the ratio (Area l)/(Area 2), to which the mono/mono curve asymptotes for large distances, as described above.

The calculations for the curves in Fig. 3 and mono/mono curve in Fig. 4 are for a coil arrangement as depicted in Fig. 1 with the diameters of the TX, RXl and RX2 coils set to 33 cm, 33 cm and 8 cm respectively. As noted above, the scale length for the variation of the RX1/RX2 ratio with depth is given by the diameter of the larger RXl receive coil. If the said coil arrangement is used together with a continuous wave

metal detector - as has been disclosed previously in the art - it is usually necessary for the receive coils to be arranged so that the net flux associated with the primary (transmitted) field is zero. Those skilled in the art are aware of the range of techniques that can be used to achieve this result. For the concentric arrangement disclosed in US patent 5721489, an additional TX coil (TX2), counter-wound with respect to the outer TX coil, is most often used. The presence of this additional coil does not significantly affect the discussion above (other than to complicate the detail of the transmitted magnetic field close to the coil arrangement), but it does require that RXl is somewhat smaller than the transmit coil TX. The smaller size of RXl compared with the largest dimension of the coil arrangement further limits the depth information that can be extracted from deep targets, and in particular limits the depth resolution available from this type of detector for deeper targets.

A further limitation to the existing art lies with the susceptibility of the detectors disclosed to external electro-magnetic noise. To obtain the desired depth resolution for targets, detectors described in the art require a signal to be derived from a two receive coils RXl and RX2, wherein RX2 is, in some embodiments, significantly smaller than RXl. Because signals from deeper targets will be small in RX2 for geometric reasons well understood by those skilled in the art, the system for receiving, amplifying and recording signals from RX2 must have both high gain (achievable by having a large number of turns on the RX2 coil, or high gain on the amplification circuit) and noise as low as achievable. If the RX2 coil is a simple circular coil, any external electromagnetic signals present in the local environment will induce voltage in the RX2 coil, which, because of the high gain required by the design, may cause noise problems in the detector which will further limit the depth resolution capabilities of the detector.

The current invention aims to improve on the ideas presented in these earlier disclosures.

SUMMARY OF INVENTION

In a first aspect of the invention there is thus provided a metal detector having a coil housing with at least two coils each capable of simultaneously receiving a respective signal from a target excited by a transmit coil and preferably wherein the two coils capable of receiving the target signal are of a substantially different effective configuration, and hence sensitive to different aspects of the secondary magnetic field from the target.

A different effective configuration includes coils that have a different physical configuration such as a circular coil versus an "8" shaped coil, and different effective configurations may also include coils having a similar physical configuration. Where the receive coils do have similar physical configurations, the receive signals from the respective coils may be formed either by addition or subtraction or some other mathematical manipulation to provide, in effect, different signal responses that would occur with coils of different physical configurations, thus having different effective configurations.

Flat, circular coils are examples of what are called "dipole" coils. If a source of electrical current is connected to a flat circular coil, effecting a current in a particular direction along the wire constituting the windings of the coil, a dipole-like magnetic field is generated. Field lines can be used to describe the field, pictorially, in a generally accepted representation. Field lines are continuous. In the case of a flat, circular coil, the field lines on one side of the plane, near the plane and within the boundary of the coil, point toward the plane, while just on the other side of the plane the field lines within the boundary of the coil point away from the plane. This is the dipole field of the dipole coil.

A coil wound as a Figureδ coil has two loops. Passing a current along the wire of the coil will produce a dipole field in one of the two loops, while the other loop also produces its own dipole field, but the sense of its field opposes that of the field of the first coil. When the sources of these two dipole fields are mutually proximate, in the

scale of the coils, the arrangement produces a total field called a "quadrupole" field, it having four poles as opposed to the two poles of the dipole field.

That it is the one coil, in this case, producing the quadrupole-like field entitles the coil to be nominated a quadrupole coil.

The application of the laws of electromagnetism, including electromagnetic reciprocity, can be used to infer how magnetic fields will induce electrical signals in coils with these configurations.

It is possible to produce a configuration of two dipole coils that provides an effective configuration of a single quadrupole coil. Connecting one wire-end of one coil to one wire-end of the other and positing them on parallel, or coincident, planes such that the senses of their windings, as referenced to their common connection, are in opposing senses, produces the effective configuration of a quadrupole coil. These coils can be described as being serially connected in opposition.

In another effective configuration, it is possible to produce an effective configuration of a quadrupole coil using two dipole coils, physically posited with respect to each other as described in the previous paragraph, whose windings are not electrically connected to each other, but are connected to respective receive circuits. The signals produced by the receive circuits can be, then, superposed in such a manner that the resulting electrical signal has the same form as if it were produced by an amplification of the single signal emanating from the two dipole coils serially connected in opposition.

This describes a configuration that is, effectively, a quadrupole configuration, but without the coils being wound, or even connected to each other, as to directly effect a quadrupole coil. Indeed, the situations in which the physical senses of the windings of the two coils are opposing, or in the same sense, can be accommodated in the electronic amplifiers by adding, in the case of coils with opposing senses, or

subtracting in the case of coils with similar senses, the two received signals. The resulting signal that is a result of the superposition of the signals produced by the receive circuits is a signal from an effective quadrupole coil

Conversely, it is also possible to superpose the signals from the same two receive circuits such that the resulting electrical signal has the same form as if it were produced by an amplification of the single signal emanating from the two dipole coils serially connected in concert. This superposition produces an effective configuration of a single dipole coil and the resulting signal is a signal from an effective dipole coil.

Both types of effective configuration can be applied to the signals for the two receive coils in the one detector; thus, the one pair of coils can be effectively configured, simultaneously, as both an effective dipole coil and an effective quadrupole coil.

Further along this theme, if the signals produced by the receive circuits are sampled and digitized, superpositions can be performed as arithmetic operations upon the digitized signals within a microprocessor or similar device, rather than with the electrical signals themselves. With this means, the one pair of coils can be effectively configured, simultaneously, as both an effective dipole coil and an effective quadrupole coil.

Li one embodiment of the invention there is provided a metal detector including; a first coil for transmitting a signal and receiving a first response signal, a second coil for receiving a second response signal, where said first and second coils for receiving first and second response signals are of different effective configurations and, where said first and second response signals are processed to determine at least the depth, location, approximate size, shape, and/or other useful characteristic of an object that is detected by said metal detector.

In another embodiment of the invention there is provided a metal detector including;

a transmit coil for transmitting a signal, at least two receive coils for receiving a first and second response signal, where said at least two receive coils are of different effective configurations and, where said first and second response signals are processed to determine at least the depth, location, approximate size, shape, and/or other useful characteristic of an object that is detected by said metal detector.

hi a further embodiment of the invention there is provided a metal detector, where at least one coil for receiving a first response signal is configured substantially as a dipole coil and, at least one coil for receiving a second response signal is configured substantially as a quadrupole coil.

In a further embodiment of the invention, the metal detector has more than two coils for receiving response signals.

In another embodiment of the invention there is provided a metal detector including; a transmit coil for transmitting a signal, at least three receive coils for receiving a first, a second and a third response signals, where the said first receive coil is configured substantially as a dipole coil and the said second and third receive coils are configured substantially as an effective quadrupole receive coil through a superposition of emissions of a first and a second receive circuits, the second receive coil being connected to an input of the first receive coil and the third receive coil being connected to an input of the second receive circuit and, where the said first response signal and the said superposition of emissions of the first and second receive circuits are processed to determine at least the depth, location, approximate size, shape, and/or other useful characteristic of an object that is detected by said metal detector.

hi another embodiment of the invention there is provided a metal detector including; a transmit coil for transmitting a signal, at least two receive coils for receiving two response signals,

where said at least two receive coils being effectively configured, simultaneously, as an effective dipole coil and as an effective quadrupole coil, where a signal from the effective dipole coil and a signal from the effective quadrupole coil are processed to determine at least the depth, location, approximate size, shape, and/or other useful characteristic of an object that is detected by said metal detector.

In a further embodiment of the invention at least one coil for receiving a response signal is of a circular configuration and at least another coil for receiving a second response signal is of an "8" shaped configuration. Examples of this embodiment are shown in Fig. 2 in the drawings.

In a further embodiment of the invention at least one coil for receiving a response signal is of an effective circular configuration and at least another coil for receiving a second response signal is of an effective "8" shaped configuration. Examples of this embodiment are shown in Fig. 7, Fig.8, and Fig. 9 in the drawings.

In a further embodiment of the invention the coils of the metal detector may be arranged as an array. Preferably, the metal detector includes pairs of coils of different effective configurations arranged as an array. An example of the coils arranged as an array is shown in Fig. 9.

It has been found that when the first receive coil is of an effective circular configuration and the second receive coil is of an effective "8" shaped configuration the magnetic fields created by the two configurations allow for complementary information to be obtained. In addition using the effective "8" shaped configuration in conjunction with a circular coil allows for more accurate processing of the signal. In addition the effective "8" shaped configurations are relatively quiet having low electromagnetic interference. In addition the effective "8" shaped configured coils offer excellent pinpointing of concealed targets.

The metal detector of the current invention may be a pulse induction (PI) type metal detector or a continuous wave (CW) metal detector, both types being well known in the metal detector industry.

Where the metal detector is a continuous wave metal detector, it has previously not been possible to have the same coil for transmitting and receiving the signals. The Applicant has been able to achieve a single coil capable of both transmitting and receiving a continuous wave signal using novel signal processing techniques. This invention is disclosed in co-pending Australian provisional patent application, 2007901083, that is hereby incorporated by reference.

DETAILED DESCRIPTION OF THE INVENTION

Fig. 1 illustrates a typical embodiment of the prior art having two receive coils and one transmit coil, all of the same circular physical configuration.

Fig. 2 illustrates one embodiment of the invention where one coil of circular configuration acts as the transmit coil and a receive coil, and a second coil of a fig. "8" configuration acts as a second receive coil.

Fig. 3 depicts a response curve using a coil configuration as depicted in Fig. 1 when moved over two targets.

Fig. 4 shows the calculated variation of the ratio of the two receive coils as a function of depth using a coil arrangement similar to that depicted in Fig. 1 and Fig 2.

Fig. 5 illustrates the calculated response of the two receive coils using a coil arrangement as shown in Fig. 2

Fig. 6 shows results obtained when passing coils having the arrangement shown in Fig. 2, over a coin target.

Fig. 7 illustrates a coil arrangement where the two receive coils have a similar physical configuration.

Fig. 8 illustrates a coil arrangement where the first coil is used to transmit and receive signals, and the second coil is used to receive a second receive signal.

Fig. 9 illustrates a coil arrangement of the current invention as an array.

Fig. 10 shows the experimental data using an array illustrated in Fig. 9.

Fig. 11 shows the experimental data using an array illustrated in Fig. 9 where the array has a forward and reverse pass of the target.

hi one embodiment of the invention, the transmit coil TX and the first receive coil RXl are one and the same coil which has, for example, a circular configuration, while the second receive coil RX2 is an "8' shaped coil located coplanar with the TX/RX1 coil. This configuration, which is only one embodiment of the present invention, is shown schematically in Fig. 2. Fig. 5 illustrates the calculated response of the first receive coil (RXl) and the second receive coil (RX2) as the Fig. 2 coil arrangement is moved over two targets, in this case located 15 cm below the plane of the coil arrangement. Target 1 is a target that can be excited to produce a dipole-like response only in the vertical (z) direction, such as a loop of wire with inductance L and resistance R which lies wholly in the horizontal (x, y) plane. It is well known among those skilled in the art that a flat conductive disk such as a metallic coin similarly orientated to a loop produces a response with similar spatial variation. Target 2 is a target that can be excited only to produce a dipole-like response only in the horizontal x direction (here defined to be in the direction of motion of the coil arrangement), such as a loop of wire or coin which lies wholly in the vertical (y,z) plane. In contrast with the case for coil arrangements of the prior art (Fig. 3), it is seen in Fig. 5 that the responses from RXl and RX2 to the said targets are different in form, owing to the

complementary information provided by the two different arrangements of RXl and RX2. Fig. 6 shows experimental results obtained when passing a common coin over a detector of the type described in this invention, with the coil arrangement of Fig. 2, at a distance of 15 cm and with the orientation for each pass coinciding with Target 1 and Target 2 above. The shapes and relative sizes of the RXl and RX2 response for each of the two coin orientations are in excellent agreement with the calculations of Fig. 5, and illustrate the rich target information that is inherent in the present invention. Processing of the signals in Fig. 5 using well-known techniques of inversion could be carried out in an on-board microprocessor to return both target depth and orientation for each of the signals depicted.

The mono/Fig 8 curve in Fig. 4 shows the calculated variation of the ratio RX1/RX2 as a function of the depth of the target below the coil arrangement depicted in Fig. 2. In contrast with the case for coil arrangements in the prior art (mono/mono curve), it is seen the RX1/RX2 ratio varies almost linearly with depth across a wide range of depths. For this coil arrangement, there is only one scale length, and it is governed approximately by the diameter of the outer RX coil (in this case, but not always, the same as the TX coil). Thus for any size coil arrangement, the RX1/RX2 ratio will vary approximately linearly with depth, with smaller coils varying more quickly. The linear variation of the ratio RX1/RX2 with depth means that there is no intrinsic loss in depth resolution for deeper targets, as is the case for prior art arrangements.

The interference-canceling properties of the "8" shaped receive coil configuration are well known among those familiar with the art, with most sources of external electromagnetic interference inducing approximately equal and opposite voltages in the opposite lobes of the "8" shaped coil, resulting in a net zero signal from the coil. As can be seen from mono/Fig 8 curve in Fig. 4, the signal from the "8" shaped RX2 coil in the present arrangement is most often significantly smaller than that from the mono-loop RXl coil, and as such the presence of external electro-magnetic noise has a potentially larger effect on the smaller RX2 signal compared with the larger RXl signal. The interference-canceling properties of the "8" shaped RX2 coil provide a

significant benefit over mono-loop RX2 configurations by providing an RX2 signal with much improved signal-to-noise ratio, allowing more accurate measurements of the spatial evolution of the signal and therefore providing better information on deep targets.

The pinpointing properties of the "8" shaped receive coil configuration are well known among those familiar with the art. For all simple target geometries, the signal output from the "8" shaped receive coil crosses through zero when the centre of the coil arrangement passes directly over the centre of the target, and this zero-crossing has been used to good advantage in signal processing to indicate the position of the target beneath the coil.

The discussion above is based for convenience on the coil arrangement depicted in Fig. 2. There is a number of other ways in which all the desirable properties of this coil arrangement can be replicated in configurations that exhibit the essence of the invention. Two examples of suitable configurations are shown in Fig. 7 and Fig. 8. In Fig. 7, the outer coil TX is used to transmit the primary magnetic field, while two receive coils RXl and RX2 are arranged inside TX and connected to independent receive electronics. Two receive signals required in the current invention can be generated as two effective receive coils RXl' and RX2' of different effective configuration by forming (in analogue electronics or by digital addition of digitized signals in the detector electronics) by a sum and a difference, respectively of the RXl and RX2 signals. RXl ' = RX1+RX2 gives an effective "mono-loop" response, while RX2' = RX1-RX2 gives an effective "8" shaped response. The signals RXl ' and RX2' can then be analyzed to deliver all the outcomes described for the current invention. In Fig. 8, the coil on the left is used for both TX and RXl, and the coil on the right is used for RX2. Effective signals RXl ' and RX2' can be formed and analyzed as described above to give the desired benefits of the present invention, despite some additional complexity in analysis required to compensate for the lack of symmetry between the transmit and receive coils in this geometry.

In a further extension to the invention, a coil configuration such as that shown in Fig. 2 could be complemented by addition of a third receive coil RX3, of similar geometry to RX2, orientated in the same plane but rotated 90 degrees relative to RX2. If connected to a third receive circuit, the information from RXl, RX2 and RX3 would provide a complementary set of information primarily sensitive to the three perpendicular excitation axes of a concealed target.

The configurations discussed above are well suited for use in a hand-held metal detector, since such a unit is typically swung from side to side to produce time- evolving signals such as those shown in Fig. 5. The metal detecting array has found a number of applications, in areas as diverse as treasure hunting, landmine or UXO detection, and geophysical mapping. Such arrays are typically designed to cover a relatively wide area while being propelled in the forward direction only: either fixed to the front of a vehicle, or mounted on a pushcart, or carried by hand. The present invention can be applied readily to metal detecting arrays, provided the coils are arranged so that the two receive coils of different effective configuration cross the target in the appropriate direction. One possible embodiment is shown in Fig. 9, which can be thought of as a multiple-receive-coil implementation of configuration of Fig. 7. Other embodiments are possible based on the configurations of Fig. 2 and Fig. 8. While Fig. 9 depicts a nine-receive-coil array, this is for illustration only as any number of receive coils from two or more could be used. In Fig. 9, the "forward" direction is indicated by the direction of the arrow. As is well known in metal detector array art, the cross-track position of a target can be deduced from the relative amplitudes of the responses in the various receive coils as the array passes over the target. Consider a target that passes under the array along the dashed line in Fig. 9. Such a target produces first a response in coil RX3, followed by a slightly delayed but similar-amplitude response in coil RX7, and from this information the cross track location of the target can be inferred. Fig. 10 shows experimental data from a test array of this type, with the array passing the target first in the forward direction (RX3 leading), then in the reverse direction (RX3 trailing). Because the coils RX3 and RX7 have similar physical configuration and size, these data exhibit no information that

readily relates to the depth of the target. Using the present invention, we can form two coils of different effective configuration, first an "effective" mono-loop-type response (call it RXl ') is formed from the sum of the two coils RXl ' = RX3+RX7, and an effective "8" shaped response (call it RX2') is formed from RX2'= RX3-RX7. Fig. 11 shows the RXl ' and RX2' responses for the forward and reverse pass, and as described in detail above, the depth of the target can be inferred from comparison of the ratio RXl '/RX2' with calculation. Experimental measurements with a test array of the type shown in Fig. 9, but with a smaller number of receive coils, show good agreement between actual and inferred depth. For targets that produce a response across a number of array elements, depth measurements can be improved by forming linear combinations of the receive coil outputs to produce effective receive coils.

In some of the coil configurations discussed above, one or more of the receive coils are induction-balanced with respect to the transmit coil. Examples are RX2 in Fig. 2, RX2 in Fig. 8 and the effective receive coil RX2' = RX1-RX2 in Fig. 7. Because the primary transmitted field induces no voltage in these induction-balanced receive coils, they may be used to receive signals while the transmitter is in operation, as is common in the continuous-wave (CW) type of metal detector well known in the art. Information received from the receive coils during transmission provides a useful means of discriminating between ferrous and non-ferrous targets, because the magnetic properties of ferrous targets induce a characteristic phase shift in the received signal. In the exemplars depicted in Figs. 2, 7 and 8, the first receive coils RXl (and the effective receive coil RXl ' = RX1+RX2 in Fig. 7) are not induction balanced. This does not, however, preclude the construction of a CW metal detector that embodies the invention. In one possible embodiment, a means of receiving a signal from a CW metal detector using a single mono-loop coil for both transmitting and receiving such as that disclosed by the applicant in a co-pending provisional patent application 2005901108, could be employed to allow the RXl to be received and analysed from a coil which is not induction-balanced with respect to TX. In a further embodiment, where additional complexity is not a consideration, RXl could be inductively balanced with respect to a non-simple transmit coil by using techniques

well known in the art, such as splitting the windings of the TX coil into a main coil and a counter-wound bucking coil configured to contribute zero net flux through either of RXl or RX2.

Many metal detectors - be they CW or PI - have the ability to provide the user with information regarding the characteristic frequency of a concealed target (or equivalently its time constant), and its ferrous content. In a CW detector, this information resides in response measured both in-phase and in-quadrature with the transmitted signal, while in a pulse induction detector time constant information comes from the shape of the characteristic decay curve of a target and ferrous information by measurements made during the transmit pulse. As described above, the invention disclosed here can be implemented with either CW or PI technologies, and the additional target information supplied by the detector technology can be used to complement the invention. As an example, two buried targets may have markedly different characteristic frequencies of ferrous content. If this information is extracted from the outputs of RXl and RX2, it is possible to separate the data stream into two complementary data streams, one for each target. From these data streams, further information can be extracted relating to the relative contributions of RXl and RX2 and hence the depth and position of the two targets as described above.

Those experienced in detecting buried metallic objects with a metal detector are able to estimate the depth of a target by the spatial distribution of the target signal. When using a particular detector, fitted for example with a mono-loop coil, the breadth of the target response signal (quantified as the full width at half maximum (FWHM) of the receive signal) provides direct information about the depth of a target, since beyond a depth of approximately one coil diameter the FWHM increases approximately linearly with target depth. It is clear, then, that correlating target response with the position of the detector head has the potential to offer increased information about target depth. In practice, this correlation is presently impractical for hand-held metal detectors in the absence of a means of accurately estimating the position of the sensor head, since the temporal FWHM (which can be easily measured

in the detector) is related to the spatial FWHM through the detector speed, which is relatively uncontrolled. Disclosed techniques for range-finding using the response of two or more receive coils in an electro-magnetic induction systems have been based solely on the ratio of the amplitude of signals from the receive coils. By using receive coils of different effective configurations, additional information is provided that can be used for range-finding. In the present invention, both RXl and RX2 provide information that varies spatially, and correlating this information with position would greatly enhance the available target information, particularly concerning depth. As noted above, the spatial FWHM of the RXl response is related to target depth, as is the curvature at the peak of the RXl response, and calculations and measurements show that so, too, is the separation between the positive and negative peaks of the RX2 response (see Target 1 in Fig. 6, for example), and the slope of the RX2 response as it passes through zero between the positive and negative peaks. Combinations of these quantities can be formed that can provide additional information on target depth, independently of the speed of the detector head over the target.

Testing with a detector built along the lines of the invention indicates that many targets - arguably the majority of commonly encountered metallic items buried in the ground - produce responses that are similar to those calculated from simple models. For these targets, comparison of the observed evolution of the RXl and RX2 signals with model-based predictions allows various parameters relating to the target (depth, size, orientation) to be estimated, together with an estimate of the goodness-of-fit between the observed and expected response. Additional information on the target composition is potentially available depending on the base technology used in the detection process (CW or PI). For many applications of a metal detector (hunting for gold, treasure, unexploded landmines, UXO) such information provides powerful discrimination capabilities. If a response can be related with a high confidence level to a particular target depth, orientation, size and composition, it may be possible in certain circumstances to rule out targets of interest. For example in a minefield a small pieces of ferrous shrapnel may be readily distinguished from a landmine, and the user provided with a confidence level indication for the discrimination process.