A method for evaluating a property of buried objects

Abstract

A computer-implemented method is provided for evaluating at least one property of a plurality of buried objects, wherein the plurality of buried objects comprise a metallic material. The method comprises: receiving magnetic field data, wherein the magnetic field data comprises a plurality of magnetic field measurements, wherein each magnetic field measurement comprises a magnetic field value and three-dimensional measurement location information; determining a target set of buried objects, wherein the target set comprises estimated two-dimensional location information for each buried object in the target set; selecting the plurality of buried objects from the target set of buried objects based on the estimated two-dimensional location information; and performing a multiple-object inversion of a model of a magnetic field pattern affected by the plurality of buried objects, wherein the multiple-object inversion provides a respective value of the at least one property for each buried object of the selected plurality of buried objects that fits the received magnetic field data. Unlocking insights from Geo-Data, the present invention further relates to improvements in sustainability and environmental developments: together we create a safe and liveable world.

Description

A METHOD FOR EVALUATING A PROPERTY OF BURIED OBJECTSFIELD

[0001] The present disclosure generally relates to systems and methods for evaluating at leastone property of a plurality of buried objects using inversion of a model of a magnetic field patternaffected by the buried objects. Unlocking insights from Geo-Data, the present invention furtherrelates to improvements in sustainability and environmental developments: together we create a safe and liveable world. BACKGROUND

[0002] Magnetometry is a common survey technique for the detection of buried objectscomprising a metallic material, such as Unexploded Ordnance (UXO) objects including marine(or offshore) UXO objects. Metallic material interferes with magnetic fields to such an extent thatit is measurable by magnetometry sensors. In some examples, the metallic material isferromagnetic. For example, detection of an UXO object by a magnetometer survey is based onthe fact that most UXO objects are typically ferromagnetic as they are often made out of steel.UXO objects therefore have a magnetic susceptibility that is higher than its surrounding and createa detectable magnetic anomaly.

[0003] In the North Sea between Great Britain, Norway, Denmark, Germany, the Netherlandsand Belgium, there is known to be a large quantity of unexploded munitions from World War II (WWII). In addition to dropped aerial bombs during WWII, sea mines from the First World War and WWII and dumped ammunition after WWII still roam the North Sea. One estimate values the amount of dumped ammunition after WWII in the Baltic sea as 1.3 million tonnes, and in the North Sea as 300 thousand tonnes. Both are a mix of conventional and chemical munitions ranging insize from cases of rifle rounds to large aerial bombs. Detection of UXO objects is exemplified herewith respect to the North Sea and Baltic Sea but it remains a problem in different marine regions throughout the world.

[0004] Hazardous UXO objects are not the only objects which generate a magnetic anomalyin and below the seafloor. Scrap metal and other types of waste have the potential to generatemagnetic anomaly fields of a similar size as UXO objects. From a practical and legal perspective,confirming whether an object is, or is not, an UXO object is performed by visual inspection. Acertified UXO expert must view the object and make a determination as to its classification. MostUXO objects on the seabed are buried and must be dug out before a visual classification can bemade. Such visual classifications are expensive in terms of time and equipment use and also comewith safety risks for the operating personnel. In one estimate, over 90% of potential (candidate)UXO (pUXOs) objects turn out not to be UXO objects but rather false alarms (false positives),leading to high operation costs to find and examine object that are likely not even hazardous. Ofcourse, in particular when dealing with UXO objects, false negatives by dismissing an anomalywhich was in fact a real UXO object, is highly undesirable.

[0005] The purpose of the magnetic survey phase is to detect presence and location ofcandidate UXO objects on or in the seabed and to determine something about the nature of thecandidate UXO object. However, the interpretation of readings at a magnetometer is a complexprocess and can result in an unacceptably high error rate if insufficiently robust algorithms areused. Further, if a large survey area is used and / or there are multiple candidate UXO objects, theamount of data and complexity of the problem increases which introduces a higher error of rateand / or greater processing resources requirements (to avoid excessive runtime).

[0006] Although set out above in the context of detecting UXO objects, the same challengesare faced by methods of detecting other kinds of buried objects with magnetic field measurements. OVERVIEW

[0007] According to one aspect of the present disclosure, there is provided a computer-implemented method for evaluating at least one property of a plurality of buried objects, wherein the plurality of buried objects each comprise a metallic material. The method comprises receiving magnetic field data, wherein the magnetic field data comprises a plurality of magnetic field measurements. Each magnetic field measurement comprises a magnetic field value and three-dimensional measurement location information. The method comprises determining a target set ofburied objects, wherein the target set comprises estimated two-dimensional location informationfor each buried object in the target set. The method comprises selecting the plurality of buriedobjects from the target set of buried objects based on the estimated two-dimensional locationinformation. The method comprises performing a multiple-object inversion of a model of amagnetic field pattern affected by the plurality of buried objects, wherein the multiple-object inversion provides a respective value of the at least one property for each buried object of the selected plurality of buried objects that fits the received magnetic field data.

[0008] Using a multiple-object inversion produces a more accurate result, i.e., values of the atleast one property of the buried objects that are closer to the true values, compared to existingtechniques. The multiple-object inversion particularly improves accuracy when some of the buriedobjects are in close proximity, such as by reducing the false negative rate (e.g. when a smallerobject is not distinguished from a larger nearby object) and / or better distinguishing between twoburied objects, e.g. resolving their distinct locations.

[0009] As the buried objects each comprise a metallic material, the buried objects interact withmagnetic fields such that their influence on the measured magnetic field at a magnetometry sensoris detectable. As such, the buried objects can be said to affect a magnetic field pattern, i.e., to causea magnetic field pattern that is different to what the magnetic field pattern would be if the buried object were absent. This is sometimes referred to as producing an anomaly in the magnetic field.The buried objects may directly produce a magnetic field and pattern of their own, i.e., have apermanent magnetic moment, but more commonly the buried objects affect a magnetic fieldpattern of an external magnetic field applied thereto. For example, buried objects may affect amagnetic field pattern of Earth’s magnetic field or of a probe having its own a magnetic moment.The metallic material may be ferromagnetic. The metallic material may be paramagnetic.

[0010] In general, inversion is the process of taking an equation or algorithm that transformsan input into an output and ‘inverting’ the equation to determine the input from a measured output.In particular, inversion is useful when the relationship between the input and output is complex such that there is no direct calculation that can be performed to derive the input from the output. In the context of magnetometry, the input is typically the location and / or a magnetic moment ofan object and the output is an observed magnetic field at a given location. Therefore, inversionfinds the location and / or magnetic moment of a candidate buried object from a measured magnetic field. The methods described herein use multiple-object inversion, wherein values of at least oneproperty (e.g., location, magnetic moment) of multiple buried objects are determined from oneinversion of the magnetic field measurements.

[0011] The estimated two-dimensional location information may be a position on Earth’ssurface, e.g., in the form of a northing value and an easting value. This may be relative to a localreference point or may be a latitude value and a longitude value. The estimated two-dimensional location information may be from an initial low-accuracy calculation based on the magnetic fielddata. The selecting the plurality of buried objects from the target set of buried objects may be basedonly on the estimated two-dimensional location information or based at least in part on theestimated two-dimensional location information and one or more additional parameter.

[0012] The at least one property of the plurality of buried objects may include at least one of:three-dimensional location; three-dimensional magnetic moment; magnetic moment magnitude;and buried depth. The at least one property may include any combination of two, three, or four of:three-dimensional location; three-dimensional magnetic moment; magnetic moment magnitude;and buried depth. The at least one property may be buried depth and the methods described hereinare particularly accurate at determining buried depth, which is a major factor in the complexity,time, and cost for extracting a buried object. In an example, the at least one property includesthree-dimensional location and three-dimensional magnetic moment.

[0013] The method may comprise determining size and / or orientation information about theplurality of buried objects, either directly (i.e., these being one of the at least one property of theplurality of buried objects) or by calculation based on the values of at least one property foundusing inversion. For example, the orientation of a buried object may align with the determined magnetic moment of the buried object. The size of the buried object may be determined from the magnetic moment magnitude.

[0014] The method may comprise performing a plurality of preliminary inversions, whereineach preliminary inversion is performed for one buried object of the selected plurality of buried objects. A result of each preliminary inversion may be used as a starting value for the at least one property for the multiple-object inversion. The multiple-object inversion may be sensitive to astarting value used for the inversion process. Therefore, using preliminary inversions increases theaccuracy, e.g., likelihood of reaching a global residue (the difference between observed andmodelled magnetic field) minimum rather than local residue minimum and / or increase the speedof the multiple-object inversion, e.g. reducing the time or number of iterations until an endcondition has been met.

[0015] Each preliminary inversion of the plurality of preliminary inversions may use arespective plurality of starting values of the at least one property. This increases the accuracyand / or speed of the preliminary inversions, which in turn improves the accuracy and / or speed ofthe multiple-object inversion. Each preliminary inversion using a respective plurality of startingvalues means, in general, that there is a different designated plurality of starting values used for each of the preliminary inversions. However, in some examples, the plurality of starting valuesmay be the same for some or all of the preliminary inversions. Each starting value may be acombination of a set of depth values and a set of magnetic moment values. The magnetic momentvalues may be expressed in cartesian coordinates (e.g., as mx, my, mz) or in polar coordinates (e.g.,the magnetic moment values having a range of magnetic moment magnitude values, magneticmoment inclination values, and magnetic moment declination values, wherein inclination is the polar angle between 0 and 180 degrees from a vertical axis perpendicular to the ground and declination is the azimuthal angle between 0 and 360 degrees clockwise from North), or any other suitable coordinate system.

[0016] The multiple-object inversion may use a plurality of starting values of the at least oneproperty. Each starting value may be a combination of a set of position values and a set of magneticmoment values for each buried object selected for the multiple-object inversion.

[0017] The selecting the plurality of buried objects in the method may comprise clustering thetarget set of buried objects into at least one cluster, wherein locations of the buried objects in each cluster are separated by no more than a threshold separation distance from other buried objects inthat cluster. Clustering buried objects before multiple-object inversion limits the number of buriedobjects in the multiple-object inversion to buried objects within a certain distance, increasingcomputational efficiency by utilising multiple-object inversion for buried objects that share aregion of influence on the magnetic field measurements. The clusters may each comprise at least two buried objects. The threshold separation distance may be 50 metres. The threshold separation distance may be 100 metres, 80 metres, 60 metres, 40 metres, 20 metres, or 10 metres. In general, the threshold separation distance may depend on the expected size and / or magnetic moment of theburied objects (and therefore their range of influence on magnetic field). The threshold separationdistance may be determined from a user input. The method may comprise performing the multiple-object inversion for each cluster. That is, the process of multiple-object inversion may be performed collectively on all the buried objects of a first cluster and then likewise for each other cluster in turn. The method may comprise collating the results of the multiple-object inversionsinto a survey output including the values of the at least one property for all the buried objects in aregion of interest.

[0018] The plurality of respective values of the at least one property for each buried object ofthe selected plurality of buried objects may fit a local region of the received magnetic field data, wherein the local region is defined by a threshold local distance from the locations of the selectedplurality of buried objects. This increases computational efficiency because only the magnetic fielddata of the local region that is influenced by the buried objects (or cluster) is used to determine theat least one property of the buried objects. The threshold local distance may be 10 metres. Thethreshold local distance may be 50 metres, 30 metres, 20 metres, 5 metres or 2 metres. In general,the threshold local distance may depend on the expected size and / or magnetic of the buried objects moment (and therefore their range of influence on magnetic field), and / or the level of noise in anarea of interest (wherein a smaller threshold local distance can be used to not capture too muchnoise).

[0019] The determining the target set of buried objects in the method may comprise receivingthe target set from a user input. The user input may be a file (e.g., uploaded by the user) describingthe target set or may be a signal from a user interface (e.g., into which the user defines the targetset). The target set of buried objects may include estimated locations of known buried objects. Forexample, the target set may be the results from an earlier map of buried objects. Using a target setof buried objects increases computational efficiency because the method does not spend resourcesexploring the possibility of a buried object in a region that has previously been found to not containa relevant buried object, or assessing if candidate buried objects really exist or not. It also meanscomputation resource can be not spent assessing the possibility of buried objects (or theirproperties) in an area of no interest, e.g., where the presence of buried objects would not affect theuser’s construction plans, by not including buried objects in that area in the target set.

[0020] The method may comprise displaying, at a user display device, the respective values ofthe at least one property for each buried object of the selected plurality of buried objects and / orsaving, to a memory, the respective values of the at least one property for each buried object ofthe selected plurality of buried objects and / or sending, to a server or a user device, the respectivevalues of the at least one property for each buried object of the selected plurality of buried objects.

[0021] The three-dimensional measurement location information describes the position wherethe magnetic field measurement was measured, which may comprise a northing value, easting value, and either an altitude value or a depth value. The location information may be relative to a local reference point, or may be a latitude value, a longitude value, and a number of metresabove / below sea-level / seafloor / ground-level value.

[0022] The received magnetic field data may include a measurement time and measurementgeographic coordinates. The measurement geographic coordinates provide a reference todetermine the direction and strength of Earth’s magnetic field, which may be used as the externalmagnetic field affected by the buried objects (e.g., the buried objects produce an anomaly in,interference with, or disturbance in, the external magnetic field). Likewise, the measurement timeprovides a reference to determine the direction and strength of Earth’s magnetic field. Thisincreases the accuracy of the method because a more precise value for the external magnetic fieldcan be determined. However, in some examples, a less precise value for the external magnetic field is adequate. The measurement time may be a calendar date, or a value of coordinated universal time. As Earth’s magnetic field varies relatively slowly compared to magnetic field data acquisition, the measurement time may be a single value that is applicable to all the magnetic field data. If the magnetic field data was taken across a longer time period, or in multiple sessions, themeasurement time may include a set of time values each associated with the respective magneticfield measurements taken at or around that the relevant time.

[0023] The magnetic field measurements may comprise magnetometer readings. Themagnetometer readings may be taken by a first magnetometer at multiple locations, corresponding to the three-dimensional measurement location information. The magnetometer readings may be taken by an array of magnetometers, wherein the location of each magnetometer in the array corresponds to the three-dimensional measurement location information associated with the magnetic field value taken by that magnetometer. The first magnetometer and / or the array ofmagnetometers may be pulled by a vehicle, e.g., a boat, to take readings across a region of interest.

[0024] The magnetic field values may be scalar magnetic field values. This reduces thecomputational resources of the multiple-object inversion because there are fewer parameters of the magnetic field data to fit to. If the magnetic field data comprises three-dimensional magnetic field data, the method may comprise determining the scalar field based on the three-dimensionalmagnetic field data and multiple-object inversion provides respective values of the at least oneproperty that fits the scalar magnetic field values.

[0025] The buried objects of the plurality of buried objects may be unexploded ordinance(UXO). The buried objects of the plurality of buried objects may be a pipeline, or other metallicdebris, such as, e.g., anchors or anchor lines, wreckage, chains, equipment, drums, or the like.

[0026] The magnetic field pattern affected by the plurality of buried target objects may be dueto an external magnetic field applied to the plurality of buried objects. The external magnetic fieldmay be Earth’s magnetic field.

[0027] According to another aspect of the present disclosure, there is provided a systemcomprising at least one processor and at least one memory having stored thereon computer readable instructions configured to cause the at least one processor to perform any of the methods described herein.

[0028] According to another aspect of the present disclosure, there is provided a computer-readable medium comprising instructions, that, when executed by at least one processor, cause theat least one processor to perform operations comprising any of the methods described herein. Thecomputer-readable medium may be non-transitory. The computer-readable medium may be tangible. Alternatively, the computer-readable medium may be transitory. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to describe the manner in which the above-recited and other advantages andfeatures of the disclosure can be obtained, a more particular description of the principles briefly described above will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings. Understanding that these drawings depict only exemplary implementations of the disclosure and are therefore not to be considered to be limiting of its scope, the principles herein are described and explained with additional specificity and detail by way of example to illustrate aspects of the disclosure and with reference to the accompanying drawings, in which:

[0030] FIG. 1 is a schematic view of apparatus for measuring magnetic field;

[0031] FIG. 2 is a diagram of a coordinate system for detecting a buried object having amagnetic moment;

[0032] FIG. 3 schematically illustrates a method for evaluating at least one property of aplurality of buried objects;

[0033] FIG. 4 illustrates schematically illustrates an example of the method of FIG. 3;

[0034] FIG. 5 illustrates an observed magnetic field;

[0035] FIG. 6 illustrates a modelled magnetic field of buried objects determined by a single-object inversion method;

[0036] FIG. 7 illustrates a modelled magnetic field of buried objects determined by a multiple-object inversion method;

[0037] FIG. 8 illustrates a comparison between the accuracy of results between a single-objectinversion method and a multiple-object inversion method;

[0038] FIG. 9 is a schematic diagram of a computing device suitable for performing themethods described herein. DETAILED DESCRIPTION

[0039] The following is a description of certain embodiments of the present disclosure, givenby way of example only and with reference to the drawings.

[0040] Various implementations of the disclosure are discussed in detail below. While specificimplementations are discussed, it should be understood that this is done for illustration purposes only. A person skilled in the relevant art will recognize that other components and configurations may be used without parting from the spirit and scope of the disclosure. Thus, the following description and drawings are illustrative and are not to be construed as limiting. Numerous specific details are described to provide a thorough understanding of the disclosure. However, in certain instances, well-known or conventional details are not described in order to avoid obscuring the description. A reference to an implementation in the present disclosure can be a reference to the same implementation or any other implementation.

[0041] The terms used in this specification generally have their ordinary meanings in the art,within the context of the disclosure, and in the specific context where each term is used. Alternative language and synonyms may be used for any one or more of the terms discussed herein, and no special significance should be placed upon whether or not a term is elaborated or discussed herein. In some cases, synonyms for certain terms are provided. A recital of one or more synonyms does not exclude the use of other synonyms. The use of examples anywhere in this specification including examples of any terms discussed herein is illustrative only and is not intended to furtherlimit the scope and meaning of the disclosure or of any example term. Likewise, the disclosure is not limited to various implementations given in this specification. Introduction

[0042] The present disclosure describes computer-implemented methods for evaluating atleast one property of a plurality of buried objects comprising performing a multiple-objectinversion of a model of a magnetic field by the buried objects. By solving the inverse problem formultiple buried objects at a time, i.e., finding a combination of values for the properties of multipleburied objects, a more accurate method is provided. The false negative rate can be reduced, particularly for smaller buried objects near larger buried objects, and the position of multiple buried objects can be found more accurately than treating each buried object individually.

[0043] In the following, an example system for producing magnetic field measurements thatcan be used in the methods according to the present disclosure is described with reference to FIG.1. A coordinate system for describing the location of a buried object is defined with reference to FIG. 2. A general method for evaluating at least one property of a plurality of buried objects is then described with reference to FIG. 3. An example of the methods described herein is then provided with reference to FIG.4 and illustrative results of the method shown in FIG.5 to FIG.8.A computing device suitable for performing the methods described herein is then described withreference to FIG.9. System for producing magnetic field measurements

[0044] With reference to FIG. 1, a system 200 for producing magnetic field measurementsincludes a magnetic detection device 10 and a processing system 100. The magnetic detectiondevice 10 is for measuring magnetic field below the surface of the ground in marine environments(offshore). The processing system 100 receives the measured magnetic field measurements (e.g.,in vector form) from the magnetic detection device 10 and performs various processing steps thereon in order to produce magnetic field data for evaluating at least one property of a plurality of buried objects. The magnetic detection device 10 comprises at least one sensor housing 6, the sensor housing 6 defining a housing space 60. The sensor housing 6 is arranged to prevent theingress of water into the housing space 60 such that the magnetic detection device 10 can be towedbehind a vessel without damaging the internal components in the sensor housing 6.

[0045] The magnetic detection device 10 of the illustrated implementation comprises at leastone optically pumped magnetometer 1 provided in the housing space 60 of the sensor housing 6. In other implementations, any suitable magnetometer may be used, not limited to optically pumpedmagnetometers, e.g., fluxgate sensors, Overhauser magnetometers, Cesium vapourmagnetometers, diamond nitrogen vacancy magnetometers etc. The optically pumped magnetometer 1 is an optically pumped vector magnetometer arranged to provide information on a magnetic field in three substantially orthogonal directions. The use of an optically pumped vector magnetometer 1 is advantageous as it is able to provide information on a magnetic field in three substantially orthogonal directions at a single point in time. That is, the optically pumped vector magnetometer 1 provides information about the total magnetic field and on the individual vector components in three orthogonal directions. The magnetometer 1 provides this information at a substantially single location and at a substantially single point in time, to provide an improved information quality and density.

[0046] The magnetic detection device 10 comprises an electronic control unit 2 and acommunications board 3. These operate the magnetometer 1 and arrange the data transmissionfrom the magnetometer 1 to a tow vessel (not shown) or another vehicle. The electronic controlunit 2 is arranged to instruct the optically pumped vector magnetometer 1 and to instruct when the magnetometer 1 takes the measurements.

[0047] The electronic control unit 2 is arranged to control the heating of a vapour cell of theoptically pumped vector magnetometer 1 and is arranged to monitor characteristics of the optically pumped vector magnetometer 1 through one or more sensors provided therein. For example, the electronic control unit 2 may monitor the temperature of the vapour cell.

[0048] The electronic control unit 2 may further control a laser state and frequency of the laserin the optically pumped vector magnetometer 1, as well as detect and control the operational status of the one or more sensors in the optically pumped vector magnetometer 1. Furthermore, the electronic control unit 2 may sample the resulting data stream received from the optically pumped vector magnetometer 1 and may apply appropriate filtering to the incoming sensor data. A user may input the appropriate data filter. Alternatively, or additionally, the appropriate data filter may be determined by the system itself. For example, a noise filter may be applied to the data from the optically pumped vector magnetometer 1.

[0049] The communications board 3 then transmits the measurements and / or measurementconfirmation through to the tow vessel. The communications board 3 may also communicateinformation from a motion sensor to the tow vessel, so that measurement results from the magnetometer 1 can be correlated to positional data of the magnetic detection device 10.

[0050] The communications board 3 receives the output of the sensor from the electroniccontrol unit 2 and is arranged to package the data for transmission in an appropriate format. Such format may e.g., be provided through the application of a USB protocol or RS232 protocol for use in a topside recording device such as a computer hard drive. The communications board 3 may also incorporate data from other sensors, such as e.g., the motion sensors, timing systems,altimeter, temperature sensors, or the like, in the data to be transmitted from the communications board 3.

[0051] The shown magnetic detection device 10 further comprises a bulkhead connector 4.The bulkhead connector 4 is arranged to provide a leak-proof transmission from the housing space 60 to the outside, allowing for data transmission between the magnetometer 1 and a tow line 7 arranged to transmit data from the magnetometer 1 to e.g., a tow vessel. The tow line 7 is arranged to transfer a towing force to the sensor housing 6. The tow line 7 may also be a data and / or power cable. If the tow line 7 is also a power cable, a power 30 converter 5 may be provided in the tow line 7. The power converter 5 may also contain a synchronisation pass-through function. The power converter 5 may be arranged to allow for conversion of transmitted power from e.g., a towvessel to the sensor housing 6. The magnetic detection device shown in FIG. 1 also comprises aremovable endcap 8 and a fixed endcap 9. These are arranged to keep the sensor housing 6 watertight to ensure no water leaks into the housing space 60. The removable endcap 8 and the fixed endcap 9 may be any suitable endcap and have any suitable geometry to attain the required depth rating of the sensor housing 6.

[0052] The magnetic field measurements provided by the magnetometer 1 and communicatedthrough the communications board 3 is combined with global positioning data of a motion sensor(e.g., a Global Positioning System receiver). This results in georeferenced magnetic field datadescribing magnetic field measurements in three orthogonal directions for each of a plurality ofsample locations. The sample locations may also be defined in three orthogonal dimensions – x, yand z coordinates. The vector magnetic field data is received by the processing system 100 for further processing to produce magnetic field data and either to send the magnetic field data to another processor for using in a method evaluating properties of buried objects, as described further herein, or to perform the method itself.

[0053] The system 200 described with reference to FIG. 1 is one way to produce magneticfield measurements, although any known form of measuring the magnetic field is also suitable.For example, in some arrangements, the magnetic field measurements may be scalar measurementsrather than vectors. The system may have a location sensor (for sensing location relative to a localreference point or for determining location using a Global Navigation Satellite System) and themeasurement location information may be appended to the magnetic field value. Themagnetometer may not be in a marine setting, i.e., may not be submerged or above water and maybe above land or underground. Accordingly, the vehicle may be a land vehicle rather than a towingvehicle. Alternatively, the magnetometer may be part of an array of magnetometers in stationarypositions. Any suitable magnetometer may be used, including magnetometers other than opticallypumped magnetometers.Coordinate systems

[0054] With reference to FIG. 2, a coordinate system for defining the position of a buriedobject having a magnetic moment includes three dimensions x, y, and z. The x direction isNorthing, the y direction is Easting, and z direction is depth (or altitude). Accordingly, a location of a buried object in this coordinate system can be defined as (x0, y0, z0), where the x, y, and zvalues are a distance in metres between the projection of the position onto the respective axes andthe origin. The magnetic moment ^^^⃗ of a buried object is either an intrinsic magnetic moment, ifthe object is a permanent magnet, or a magnetic moment induced by an external magnetic fieldsuch as Earth’s magnetic field. The magnetic moment can also be defined as a magnetic momentcomponent in each direction x, y, z. Methods for evaluating property of buried objects

[0055] With reference to FIG. 3, in general, a method 300 for evaluating at least one propertyof a plurality of buried objects has the following features. The plurality of buried objects eachcomprise a metallic material. The method comprises receiving 302 magnetic field data, wherein the magnetic field data comprises a plurality of magnetic field measurements, wherein each magnetic field measurement comprises a magnetic field value and three-dimensional measurement location information. The method comprises determining 304 a target set of buried objects, wherein the target set comprises estimated two-dimensional location information for each buried object in the target set. The method comprises selecting 306 the plurality of buried objects from the target set of buried objects based on the estimated two-dimensional location information. The method comprises performing 308 a multiple-object inversion of a model of a magnetic field pattern affected by the plurality of buried objects, wherein the multiple-object inversion providesa respective value of the at least one property for each buried object of the selected plurality ofburied objects that fits the received magnetic field data.

[0056] A detailed example of the method 300 shown FIG. 3 is described below with referenceto FIG.4. Overview of example method

[0057] Multiple-object inversion is an algorithm designed to retrieve relevant properties, suchas the centre location and magnetic moment vector, of a set of dipole point sources. An examplemethod 400 of multiple-object inversion uses a magnetic field data set, including sensorcoordinates in 3D space and scalar magnetic data, combined with a set of estimated Northing andEasting locations of targets. First, the target list is clustered with the aid of the DBSCAN clusteringalgorithm. DBSCAN is an acronym for “Density Based Spatial Clustering of Applications with Noise”. This algorithm relies on density-based notion of clusters rather than shape specificclustering. DBSCAN is an unsupervised clustering algorithm which only relies on two inputparameters: 1) ^, radius that defines the neighbourhood of a point, and 2) MinPts, the minimumnumber of points required to form a dense region (including the point itself). Each target,regardless if it is a single target or if it belongs to a cluster, goes through a single-object inversionusing a Gauss-Newton method with multiple starting models. The targets that belong to a clusterthen go through a multiple-object inversion using the results from the single-object inversion toget to the final inversion results. Single-object inversion

[0058] Magnetometry is a survey techniques for the detection of buried objects comprising ametallic material, such as Unexploded Ordnance (UXO) objects. The buried objects typically donot create a magnetic field on their own, but affect a magnetic field pattern of an inducing field.In the case of passive magnetometer surveys, the inducing field is Earth’s magnetic field. The buried object creates a small perturbation, up to a maximum of a few hundred nT in Earth’smagnetic field, which is in the order of 104 nT. To be able to extract information about the objectfrom the anomaly field, Earth’s inducing magnetic field at the survey site (and time) is known.

[0059] Assuming that the distance between the sensor and the object is at least three times thelargest dimension of the magnetic object, one can consider it as a magnetic dipole. The dipole fieldof such a magnetized object can be calculated as (equation 1):where B is the magnetic flux density in nT, r is the distance vector in metres between the magneticobject at (x0, y0, z0) and the sensor at (xs, ys, zs), R is the norm of the distance vector of r, R = |r|,and ȓ is its unit vector. The magnetic moment of the dipole source is m, which is a three-componentvector that describes the orientation and strength of the magnetic field created by a magneticobject. It depends on the volume of the ferromagnetic object, its magnetic susceptibility and, incase of objects that are not perfectly symmetric, its orientation.

[0060] Equation 1 returns the 3D vector of the magnetic field, but some surveys usemagnetometers which measure the Total Magnetic Intensity (TMI), also known as scalar and totalmagnetic field. In short, this is the norm of the anomaly field Baplus Earth’s inducing field B0. Toextract the anomaly field from the TMI field, a Taylor series approximation can be used (equation2):^^^ = ^^^^ ⋅|^^|

[0061] Earth’s inducing field varies over time and space, but slowly enough to deem that it isconstant for the whole survey. It is known and can be found via publicly accessible online services.

[0062] In other examples, an external magnetic field may be a provided by a magnetic probeor other non-natural source, e.g., cables carrying an electric current.Multiple-object inversion

[0063] The multiple-object inversion is an algorithm using given a dataset of magnetic fielddata comprising a plurality of magnetic field measurements, each comprising a magnetic fieldvalue B with three-dimensional measurement location information, e.g., a 3D location (x, y, z),and a target set of buried objects which give an initial guess on at least the targets locations in 2D-plane (xt, yt).

[0064] With reference to FIG. 4, an example method 400 using multiple-object inversion startswith an input 402 of magnetic field values B and associated 3D location information (which is anexample of magnetic field data as received 302 in the general method 300 described with referenceto FIG.3) and a target set of buried object locations (which is an example of determining 304 in the general method 300).

[0065] The features of the example method 400 can be summarised as:1. Clustering 404 target buried objects together using DBSCAN algorithm, to producesingle targets 406 and clusters 407 (clustering 404 is an example of the selecting 306 in the generalmethod 300); 2. Running a single-object inversion 408 with different starting parameters for each targetburied object; and 3. Running a multiple-object inversion 409 for the targets that are clustered together usingthe best results from the single-object inversion for those targets as initial guess (which is anexample of performing a multiple-object inversion 308 in the general method 300).

[0066] The output 410 of the example method 400 is then a combination of the results of thesingle-object inversion 408 for single targets 406 and the multiple-object inversion 409 for the clusters 407. Clustering

[0067] To determine which targets should be grouped together for multiple-object inversion,a DBSCAN clustering algorithm is used. DBSCAN is an acronym for “Density Based Spatial Clustering of Applications with Noise”. This algorithm relies on density-based notion of clustersrather than shape specific clustering. DBSCAN is an unsupervised clustering algorithm which only relies on two input parameters: 1) ^, radius that defines the neighbourhood of a point, and 2)MinPts, the minimum number of points required to form a dense region (including the point itself).

[0068] The algorithm initiates by picking a random target t in the dataset, and retrieves alltargets q that are within distance ^ from t, (equation 3):^^^^^^^^(^, ^) ^ >Exclude from cluster<Add to cluster

[0069] Point t and points q form the bases of a new cluster C under the condition that thenumber of points is equal to or more than MinPts. If this condition is not met, points t and q willbe labelled as “noise”. The process is then repeated as another datapoint is chosen as point t. If thiscondition is met, the search is extended, (equation 4):> ^^^^^^ Continue forming cluster< ^^^^^^ Restart with a new target ^

[0070] Given that Cluster C contains the minimum number of points, the algorithm labelstargets t and q in the cluster as “visited”. The algorithm repeats equation 3, for unvisited pointsthat are within distance ^ away from a visited point. This process repeats itself until there are nomore points in the dataset that are added to the cluster and the current cluster is complete.

[0071] After completing a cluster, DBSCAN selects another datapoint that is not part of anyclusters yet and repeats the procedure described above with the unvisited points. The algorithm isdone when each datapoint has been visited and belongs either to a cluster or is marked as “noise”.

[0072] The clusters (C1, ..., Ci) include objects that should go through multiple-objectinversion and the so-called noise points (n1, ..., ni) are objects that should go through single-objectinversion only. The multiple-object inversion is performed per cluster and the single-objectinversion is performed for the remaining targets that did not fit in a cluster, i.e., the “noise” points.Inversion method

[0073] The Gauss-Newton method is an iterative non-linear least-squares solver which is wellsuited for low-level non-linear functions such as magnetic field data. The method applies a linearization on the model function with a preliminary estimate of the unknowns with the aid of the first-order Taylor series expansion. The second derivative is ignored, leaving a quadratic programming problem.

[0074] The objective function ψ to be optimized in the Gauss-Newton method can be writtenas, (equation 5):where p are the model parameters, yi is the observed signal, fi(p) is the analytically calculatedsignal based on model parameters p and resi is the residual between the y and f(p) summed overeach datapoint, denoted by subscript i.

[0075] The model parameters are iteratively updated until theoretically the objective functionconverges to its minimum. In practice, reaching the minimum is typically not necessary, so athreshold will be set to define the stopping criterion; when ψ(p) drops below a certain value or ifit does not change for a set of consecutive iterations.

[0076] The model parameters are updated as (equation 6):^^^^ = ^^ + Δ^^^^where k stands for the iteration number; when k = 0, the initial estimates of the model parametersare used and updated accordingly to k + 1. ∆p represents the step size which depends on theJacobian matrix J and the residual vector res, (equation 7):

[0077] The linearization can be visualised in this update step since the first derivatives withrespect to the model parameters are contained within the Jacobian matrix. The model equation isthen updated accordingly (equation 8):^(^^ + Δ^^^^) ≈ ^(^^) + ^^ ⋅ Δ^^^^

[0078] The Jacobian contains as many columns as there are model parameters and as manyrows as there are data points. This leads to the form (equation 9):where p(1), ..., p(n) represents the model parameters and xs(0), xs(n) the location of the data point in the 3D x,y,z-space.

[0079] To summarise, the algorithm follows the following process:1. Fill in (first guess of) the mode parameters in f(p). 2. Calculate the residual of y − f(p).3. Calculate the elements of the Jacobian matrix (equation 9), either with the aid of the firstorder Taylor series or an exact derivative if possible. 4. Update the model parameters with equation 7 followed by equation 6.5. Repeat steps 1 to 4 until the objective function reaches a self-defined criterion.

[0080] In single-object inversion the model parameters are the 3-component magnetic momentm = (m1, m2, m3) and the centre location (x0, y0, z0), which leads to a total of six model parameters.The model parameter p is defined as (equation 10):where m1 corresponds with x-direction, m2 with y-direction and m3 with z-direction of the magnetic moment vector. The inversion algorithm itself uses m1, m2, m3, but will convert these Cartesian parameters to spherical coordinates at the end of the inversion.

[0081] When using the inversion techniques described above for multiple-object inversion,equation 10 is extended to n targets that are used for the inversion. As an example, having threetargets a, b, c, the model parameter p can then be written as (equation 11):

[0082] Equivalently, equation 1 will turn into a summation of all the n dipole targets (equation12):

[0083] The size of the Jacobian matrix written in equation 9 depends on the length of modelparameter p. This means that the single-object inversion has a Jacobian of width six entries, whilethe multiple-object inversion Jacobian has a width of n × 6.Inversion starting values

[0084] The Gauss-Newton inversion is an optimization method that, like most inversionmethods, is sensitive to the initial guess of the parameters. To make the inversion algorithm morerobust against converging towards a local minimum rather than to the global minimum, multiple starting models are used in search of the best fit.

[0085] In the example method 400, all targets go through a single-object inversion, even thetargets which are grouped together as part of a cluster. This is because inputting multiple startingpoints in the multiple-object inversion for each target will greatly increase computational time.Therefore, using a preliminary inversion for a single object is more computationally efficient. Theoutput of the single-object inversions (i.e., the preliminary inversions) will be used as input for themultiple-object inversion.

[0086] In the inversion, all three components in the magnetic moment vector as well as theburial depth are varied. Note that the Easting and Northing locations of the targets (i.e. theestimated two-dimensional location information in the determined 304 target set) will remain the same for each starting model, meaning there will be different starting values for 4 out of 6 parameters.

[0087] Table 1 shows the values that are used in the starting models for the example method.Since there is no use varying the declination when the inclination is (almost) vertical - in this caseat 80 and 260 degrees – the result is 96 different starting models. In alternative implementations,a different number of starting values, and different values themselves, may be used, and these may be default parameters or user-defined parameters. Table 1: the set of different starting values for each of the varied parameters Parameter Starting values UnitsMagnetic moment strength 7, 14, 21, 28 Am2Dipole inclination 10, 80, 260 DegreesDipole declination 11, 101, 191, 281 DegreesBurial depth 0, 1, 2, 3 m

[0088] To pick the best results, the distribution of the results is reviewed to find out whichresults occurs the most, by putting the 96 results in a histogram with 24 bins and taking only those results in the most occurring bin. From this subset, the best fitting results are picked. Data input

[0089] The example method 400 uses magnetic field data that describes the observed field.This means both the scalar magnetic field as well as the observation locations in 3D space (x, y,z). The algorithm takes as input a n × 4 array as input, where n is the number of datapoints and 4represents position and field strength (x, y, z, B). Note that, to match Earth’s inducing fielddirection, x is positive towards North, y is positive towards East, and z is positive downwards with0.0 at the seafloor.

[0090] Another piece of information about the observed field that the example method 400uses is the location and date of the survey. Magnetometers measure anomalies in Earth’s magneticfield, however Earth’s magnetic field changes over time and space. The UTM (UniversalTransverse Mercator) coordinate system can be used in combination with the mean Northing andEasting values of the survey to define mean the location and survey date to calculate Earth’sinducing field components. Alternatively, a mean of the latitude and longitude values of thesurvey, and the survey date(s), can be used.

[0091] The inversion of the example method 400 also uses some initial guesses about thetargets. The 6 parameters to be retrieved from the inversion are the three components from the magnetic moment and the location of the target in 3D space. Since the algorithm uses differentstarting models, as described above, a guess of the magnetic moment and burial depth is notrequired. Accordingly, only the initial estimate for the x, y-location of target set of buried objectsis part of the input. Parameters

[0092] The example method 400 is built to have little subjective user input. This meansproviding default values for parameters controlling the method. For example, the clustering feature(using the DBSCAN clustering algorithm) depends on two parameters: the distance between twotargets and how many targets should be at minimum in one cluster. The second parameter can beset to default as MinPts = 2. This is because any two dipoles close to each can disrupt the magneticfield resulting from the other. The choice of a distance parameter ^ default is dependent onimplementation. For example, some dipoles are stronger than others and thus can disrupt themagnetic field at a further distance than smaller dipoles. Without introducing a dynamic distanceand making the clustering more complex, ^ should be taken to balance covering a reasonable rangeof influence between dipoles without creating clusters (and therefore multiple-object inversions)containing too many buried objects. The inventors have identified that the distance 50 m providesa good balance.

[0093] The inversion part of the example method 400 also uses some controlling parameters.Both the multiple-object inversion and single-object inversion may use the default parameters ofSciPy’s least squares solver (SciPy (2021) scipy.optimize.least.squares.). These parametersdescribe the conditions for terminating the inversion iterations. Alternatively, any suitable endcondition may be set, according to the accuracy and runtime desired for the implementation.

[0094] Another inversion parameter that may be controlled is the window size, also referredto herein as threshold local distance. The window size is the area around the target that is used asdata input for the inversion. This parameter is variable depending on the size of the anomaly. However, to standardise the process and avoid user bias, the multiple-object inversion may use adefault window size of 10 m, meaning that the data within 10 m radius of the target is used forfitting. Results

[0095] A set of results obtained using the example method 400 are shown in FIG. 5 to FIG. 8.FIG. 5 shows an observed field of magnetic field measurements from which the location andmagnetic moment of buried objects is to be found. The observed region is approximately 125 metres by 125 metres and contains several anomalies and some background noise.

[0096] Techniques using only single-object inversion, i.e., comparative examples, can derivea set of locations and magnetic moments of detected buried objects. The modelled magnetic fieldpattern affected by the buried objects found using these techniques is shown in FIG. 6. The resultshows that two buried objects are identified, and the magnetic moments indicated by the size,shape, and orientation of the lobes of magnetic field. A first detected buried object 610 is locatedat a coordinate of approximately (70 m, 85 m) and a second detected buried object 620 is locatedat a coordinate of approximately (80 m, 40 m). The first detected buried object 610 has an orientation approximately aligned East-West.

[0097] In contrast to FIG. 6, the example method 400 using multiple-object inversion,produces a more accurate detection of buried objects. The modelled magnetic field pattern affectedby buried objects as found by the example method 400 is shown in FIG.7. The result shows thatthree buried objects are identified, and the magnetic moments are indicated by the size, shape, andorientation of the lobes of magnetic field. A first detected buried object 610A is located at acoordinate of approximately (70 m, 85 m) and a second detected buried object 620A is located ata coordinate of approximately (80 m, 40 m). Additionally, a third detected buried object 615 isdetected at a coordinate of approximately (55 m, 85 m) that was not detected in the results of FIG.6, perhaps due to its proximity to the first detected object and its weaker magnetic momentcompared to the first detected object. Additionally, the first detected buried object 610A has anorientation approximately aligned Northeast-Southwest, whereas the orientation in FIG. 6 wasseemingly misaligned due to the influence of the (undetected) third buried object and forcing thepattern resulting from two buried objects into the pattern of a single buried object. This is anillustration of the lower false negative rate (missing buried objects) when using methods including multiple-object inversion, as described herein, and increased accuracy.

[0098] FIG. 8 shows the results of a comparison between the accuracy of techniques usingonly single-object inversion, shown in histogram 810, and methods using multiple-objectinversion, as shown in histogram 820. The difference between the detected location of a buriedobject using each method and the actual location of the buried object (e.g., verified by inspection)is provided on the x-axis. The number of buried objects that where a given distance from the actual location is provided in the y-axis. By comparing the histograms 810, 820, the method usingmultiple-object inversion can be seen to have a greater accuracy.Example variants

[0099] The example method 400, as explained above, provides one way of implementing thegeneral method 300. However, it will be appreciated that there are many variants of this method that are different ways of implementing the general method 300, some specific, non-exhaustive, examples of which are provided below.

[0100] Additionally, or alternatively to the example method, in some examples the clusteringuses a different clustering algorithm (i.e., other than DBSCAN), or different parameters for the DBSCAN algorithm. Alternatively, the clustering may be omitted and the selecting 306 may include selecting any two or more buried objects based on the estimated distance between them being below a threshold separation distance or may include selecting a first buried object and itsnearest neighbour. In some examples, some buried objects may be used in more than one multiple-object inversion and the results combined into a final value of the property for that buried object.

[0101] Additionally. or alternatively, in some examples an inversion method other than Gauss-Newton is used, according to any suitable inversion method.

[0102] Additionally, or alternatively, in some examples the preliminary inversion using asingle-object inversion 408 to produce starting values for the multiple-object inversion is not used.Instead, default or user-defined starting values for the at least one property of the buried objectsmay be used for the multiple-object inversion. In some examples, the preliminary inversion(s) usea single starting value for one or more of the at least one property, or a single combination ofstarting values for all of the at least one property. Additionally, or alternatively, in some examplesdifferent cluster parameter values are used as multiple starting values. Processing system

[0103] FIG. 9 shows a block diagram of one implementation of a processing system 900 in theform of a computing device within which a set of instructions for causing the computing device toperform any one or more of the methods discussed herein, may be executed. In alternativeimplementations, the computing device may be connected (e.g., networked) to other machines in a Local Area Network (LAN), an intranet, an extranet, or the Internet. The computing device may operate in the capacity of a server or a client machine in a client-server network environment, or as a peer machine in a peer-to-peer (or distributed) network environment. The computing device may be a personal computer (PC), a tablet computer, a set-top box (STB), a Personal Digital Assistant (PDA), a cellular telephone, a web appliance, a server, a network router, switch or bridge, or any machine capable of executing a set of instructions (sequential or otherwise) that specify actions to be taken by that machine. Further, while only a single computing device is illustrated, the term “computing device” shall also be taken to include any collection of machines (e.g.,computers) that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the methodologies discussed herein.

[0104] The example processing system 900 includes a processor 902, a main memory 904(e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM) or Rambus DRAM (RDRAM), etc.), a static memory 906 (e.g., flash memory, static random access memory (SRAM), etc.), and a secondary memory (e.g., a data storage device 918), which communicate with each other via a bus 930.

[0105] Processor 902 represents one or more general-purpose processors such as amicroprocessor, central processing unit, or the like. More particularly, the processor 902 may be a complex instruction set computing (CISC) microprocessor, reduced instruction set computing (RISC) microprocessor, very long instruction word (VLIW) microprocessor, processor implementing other instruction sets, or processors implementing a combination of instruction sets. Processor 902 may also be one or more special-purpose processors such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), network processor, or the like. Processor 902 is configured to execute the processing logic (instructions 922) for performing the operations and steps discussed herein.

[0106] The processing system 900 may further include a network interface device 908, e.g., acommunication module. The processing system 900 also may include a video display unit 910 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)), an alphanumeric input device 912 (e.g., a keyboard or touchscreen), a cursor control device 914 (e.g., a mouse or touchscreen), and an audio device 916 (e.g., a speaker).

[0107] It will be apparent that some features of the processing system 900 shown in Figure 9may be absent. For example, the processing system 900 may have no need for display device 910 (or any associated adapters). This may be the case, for example, for particular server-side computer apparatuses which are used only for their processing capabilities and do not need to display information to users. Similarly, user input device 912 may not be required. In its simplest form, processing system 900 comprises processor 902 and main memory 904.

[0108] The data storage device 918 may include one or more machine-readable storage media(or more specifically one or more non-transitory computer-readable storage media) 928 on which is stored one or more sets of instructions 922 embodying any one or more of the methods or functions described herein. The instructions 922 may also reside, completely or at least partially, within the main memory 904 and / or within the processor 902 during execution thereof by the processing system 900, the main memory 904 and the processor 902 also constituting computer- readable storage media 928.

[0109] The various methods described above may be implemented by a computer program.The computer program may include computer code arranged to instruct a computer to perform the functions of one or more of the various methods described above. The computer program and / or the code for performing such methods may be provided to an apparatus, such as a computer, on one or more computer readable media or, more generally, a computer program product. The computer readable media may be transitory or non-transitory. The one or more computer readable media could be, for example, an electronic, magnetic, optical, electromagnetic, infrared, orsemiconductor system, or a propagation medium for data transmission, for example fordownloading the code over the Internet. Alternatively, the one or more computer readable media could take the form of one or more physical computer readable media such as semiconductor or solid-state memory, magnetic tape, a removable computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disc, and an optical disk, such as a CD- ROM, CD-R / W or DVD.

[0110] The computer program is executable by the processor 902 to perform functions of thesystems and methods described herein.

[0111] In an implementation, the modules, components, and other features described hereincan be implement-ed as discrete components or integrated in the functionality of hardware components such as ASICS, FPGAs, DSPs, or similar devices.

[0112] A “hardware component” is a tangible (e.g., non-transitory) physical component (e.g.,a set of one or more processors) capable of performing certain operations and may be configured or arranged in a certain physical manner. A hardware component may include dedicated circuitry or logic that is permanently configured to perform certain operations. A hardware component may be or include a special-purpose processor, such as a field programmable gate array (FPGA) or an ASIC. A hardware component may also include programmable logic or circuitry that is temporarily configured by software to perform certain operations.

[0113] Accordingly, the phrase “hardware component” should be understood to encompass atangible entity that may be physically constructed, permanently configured (e.g., hardwired), ortemporarily configured (e.g., programmed) to operate in a certain manner or to perform certainoperations described herein.

[0114] In addition, the modules and components can be implemented as firmware or functionalcircuitry within hardware devices. Further, the modules and components can be implemented in any combination of hardware devices and software components, or only in software (e.g., code stored or otherwise embodied in a machine-readable medium or in a transmission medium).

[0115] Unless specifically stated otherwise, as apparent from the following discussion, it isappreciated that throughout the description, discussions utilizing terms such as "receiving”,“determining”, “selecting”, “comparing”, “enabling”, “calculating”, “identifying”, “analysing”, “estimating”, “performing”, “providing” or the like, refer to the actions and processes of a computer system, or similar electronic computing device, that manipulates and transforms data represented as physical (electronic) quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system memories or registers or other such information storage, transmission or display devices.

[0116] It is to be understood that the above description is intended to be illustrative, and notrestrictive. Many other implementations will be apparent to those of skill in the art upon reading and understanding the above description. Although the present disclosure has been described with reference to specific example implementations, it will be recognized that the disclosure is not limited to the implementations described but can be practiced with modification and alteration within the spirit and scope of the appended claims. Accordingly, the specification and drawings are to be regarded in an illustrative sense rather than a restrictive sense. The scope of the disclosure should, therefore, be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

[0117] While at least one exemplary embodiment has been presented in the foregoing detaileddescription, it should be appreciated that a vast number of variations exist, only some of which have been mentioned above. It should also be appreciated that the exemplary embodiment or exemplary embodiments are only examples, and are not intended to limit the scope, applicability, or configuration of the disclosure in any way. Rather, the foregoing detailed description will provide those skilled in the art with a convenient road map for implementing the exemplary embodiment or exemplary embodiments. It should be understood that various changes can be made in the function and arrangement of elements without departing from the scope of the appended claims and the legal equivalents thereof.

Claims

CLAIMS 1. A computer-implemented method for evaluating at least one property of a plurality of buried objects, wherein the plurality of buried objects each comprise a metallic material, the method comprising: receiving magnetic field data, wherein the magnetic field data comprises a plurality of magnetic field measurements, wherein each magnetic field measurement comprises a magnetic field value and three-dimensional measurement location information; determining a target set of buried objects, wherein the target set comprises estimated two- dimensional location information for each buried object in the target set; selecting the plurality of buried objects from the target set of buried objects based on the estimated two-dimensional location information; and performing a multiple-object inversion of a model of a magnetic field pattern affected bythe plurality of buried objects, wherein the multiple-object inversion provides a respective valueof the at least one property for each buried object of the selected plurality of buried objects thatfits the received magnetic field data.

2. The computer-implemented method according to claim 1, wherein the at least one property of the buried objects includes at least one of: three-dimensional location; three-dimensional magnetic moment; magnetic moment magnitude; and buried depth.

3. The computer-implemented method according to claim 1 or 2, wherein the methodcomprises: performing a plurality of preliminary inversions, wherein each preliminary inversion is performed for one of buried objects of the selected plurality of buried objects, wherein a result of each preliminary inversion is used as a starting value for the at least one property for the multiple-object inversion.

4. The computer-implemented method according to claim 3, wherein each preliminaryinversion of the plurality of preliminary inversions uses a respective plurality of starting values of the at least one property.

5. The computer-implemented method according to any preceding claim, wherein themultiple-object inversion uses a plurality of starting values of the at least one property.

6. The computer-implemented method according to any preceding claim, wherein theselecting a plurality of buried objects comprises clustering the target set of buried objects into at least one cluster, wherein locations of the buried objects in each cluster are separated by no more than a threshold separation distance from other buried objects in that cluster.

7. The computer-implemented method according to claim 6, wherein the method comprisesperforming the multiple-object inversion for each cluster.

8. The computer-implemented method according to any preceding claim, wherein theplurality of respective values of the at least one property for each buried object of the selected plurality of buried objects fits a local region of the received magnetic field data, wherein the local region is defined by a threshold local distance from locations of the selected plurality of buried objects.

9. The computer-implemented method according to any preceding claim, wherein thedetermining the target set of buried objects comprises receiving the target set from a user input.

10. The computer-implemented method according to any preceding claim, wherein the targetset of buried objects includes estimated locations of known buried objects.

11. The computer-implemented method according to any preceding claim, the method comprising: displaying, at a user display device, the respective values of the at least one property for each buried object of the selected plurality of buried objects; and / or saving, to a memory, the respective values of the at least one property for each buried object of the selected plurality of buried objects; and / or sending, to a server or a user device, the respective values of the at least one property for each buried object of the selected plurality of buried objects.

12. The computer-implemented method according to any preceding claim, wherein thereceived magnetic field data includes a measurement time and measurement geographic coordinates.

13. The computer-implemented method according to any preceding claim, wherein themagnetic field measurements comprise magnetometer readings and the magnetic field values are scalar magnetic field values.

14. The method of any preceding claim, comprising measuring the plurality of magnetic fieldmeasurements using at least one magnetometer.

15. The method of any preceding claim, wherein the plurality of buried objects includeunexploded ordinance (UXO).

16. The method of any preceding claim, wherein the magnetic field pattern affected by theplurality of buried target objects is due to an external magnetic field applied to the plurality of buried objects.

17. The method of claim 16, wherein the external magnetic field is Earth’s magnetic field.

18. A system comprising:at least one processor; and at least one memory having stored thereon computer readable instructions configured to cause the at least one processor to perform the method of any preceding claim.

19. A computer-readable medium comprising instructions, that, when executed by at least oneprocessor, cause the at least one processor to perform operations comprising the method of any of claims 1 to 17.

Images