Radiation monitoring method and circuitry

Abstract

A method of estimating suppression of background radiation, by a target object, at an evaluation location, the method comprising: estimating a region of a field of view at the evaluation location which is occupied by the target object; and estimating a degree of suppression of background radiation at the evaluation location, by the target object, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object.

Description

[0001] P131108GB

[0002] RADIATION MONITORING METHOD AND CIRCUITRY

[0003] Field of the Disclosure

[0004] The present disclosure relates to radiation monitoring methods, circuitry, and software.

[0005] Description of the Related Art

[0006] The “background” description provided herein is for the purpose of generally presenting the context of the disclosure. Work of the presently named inventors, to the extent it is described in the background section, as well as aspects of the description which may not otherwise qualify as prior art at the time of filing, are neither expressly or impliedly admitted as prior art against the present disclosure.

[0007] Radiation portal monitors (RPMs) are a class of radiation monitoring apparatus in current usage. There is currently a widespread use of RPMs at points of entry (POEs) into countries. These include drive-through RPMs for both containerised and non-containerised cargo. Air-freight and rail-freight RPMs are also used at border crossing points in order to detect undeclared radioactive materials concealed in cargo. In particular, such RPMs serve to help detect the presence of illicit Special Nuclear Materials (SNM). RPMs typically comprise a number of radiation detectors (e.g. gamma-ray and I or neutron detectors) through which nuclear materials, and SNMs in particular, can be detected and classified based on the radiation emitted by such materials.

[0008] Mobile radioisotope identification devices are another class of radiation monitoring apparatus in current usage, comprising sub-classes such as (i) handheld, (ii) backpack, and (iii) vehiclemounted radiation monitoring apparatuses, configured for mobile deployment to areas of interest. Mobile radiation monitoring devices may be used to provide complementary functionality to fixed or semi-fixed radiation monitoring apparatuses such as RPMs. Thus, for example, personnel at a site where one or more RPMs are installed for vehicular inspection may use one or more mobile nuclear radiation monitoring apparatuses to conduct investigations of vehicles which have been identified as presenting a potential risk based on nuclear radiation monitoring conducted by an RPM.

[0009] Nuclear radiation monitoring apparatuses may also be used for monitoring of pedestrians and I or baggage in locations such as airports, and other points-of-entry between jurisdictions and I or at entrances into secure facilities.

[0010] Radiation data captured by one or more radiation detectors of a radiation monitoring apparatus (such as a nuclear radiation monitoring apparatus) in proximity to a target object to be assessed is typically understood to comprise at least two broad classes of components. These broad classes are (1) a measure of any radiation emitted from the materials comprised within the targetP131108GB

[0011] object, and (2) a measure of other sources of radiation or noise which do not directly result from the radiation emission characteristics of the target object. It will be appreciated that this second broad class of components may comprise one or more of a plurality of sub-components, such as a component associated with passive environmental emission (e.g. from the ground, the atmosphere, materials in the built environment, and other materials which may be referred to in the field as - NORM - ‘naturally occurring radioactive material’), a noise component, an active interference component, and so on. This second class of components, in other words, those not associated with radiation emissions from the target object, may generally be referred to as ‘background radiation’. The presence of background radiation requires consideration when classifying radiation data, particularly since variation in background radiation can make it challenging to assess whether on not observed changes in characteristics of radiation data are associated with a target object under assessment.

[0012] Approaches to mitigating the influence of background radiation in the use of radiation monitoring apparatuses are therefore of interest.

[0013] SUMMARY

[0014] According to a first aspect of the present disclosure, there is provided a method of estimating suppression of background radiation, by a target object, at an evaluation location, the method comprising: estimating a region of a field of view at the evaluation location which is occupied by the target object; and estimating a degree of suppression of background radiation at the evaluation location, by the target object, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object.

[0015] According to a second aspect of the present disclosure, there is provided circuitry implementing control logic for estimating suppression of background radiation, by a target object, at an evaluation location, wherein the control logic configured to: estimate a region of a field of view at the evaluation location which is occupied by the target object; and estimate a degree of suppression of background radiation at the evaluation location, by the target object, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object.

[0016] According to a third aspect of the present disclosure, there is provided a non-transitory computer program product configured to control a computer to perform the method of the first aspect. According to a fourth aspect of the present disclosure, there is provided a recording medium storing a non-transitory computer program product according to the third aspect.P131108GB

[0017] The described embodiments, together with further advantages, will be best understood by reference to the following detailed description taken in conjunction with the accompanying drawings.

[0018] BRIEF DESCRIPTION OF THE DRAWINGS

[0019] A more complete appreciation of the disclosure and many of the attendant advantages thereof will be readily obtained as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings, wherein:

[0020] Figure 1 schematically shows a Radiation Portal Monitor (RPM) according to embodiments of the present disclosure;

[0021] Figures 2a to 2e schematically show configurations of a target object and an RPM at sequential time points t1 to t5 during transit of the target object through the RPM according to embodiments of the present disclosure;

[0022] Figure 3 schematically shows an idealised plot of radiation detection rates for target and background radiation components as detected at an evaluation point over time for the time period represented in Figures 2a to 2e;

[0023] Figures 4a and 4b respectively show schematic plan and front views of an RPM according to embodiments of the present disclosure;

[0024] Figures 5a and 5b schematically show overview radiographs through the side and top respectively of a target object, obtained by first and second image acquisition systems respectively, according to embodiments of the present disclosure;

[0025] Figure 6 schematically shows data connections between image acquisition control logic, model generation control logic, and field of view estimation control logic, according to embodiments of the present disclosure;

[0026] Figure 7 schematically shows a model of a target object and a model of aspects of an RPM in a modelled scene, according to embodiments of the present disclosure;

[0027] Figure 8 schematically shows a sub-region of the modelled scene of Figure 7, showing a modelled target object transit path and a modelled evaluation point location, according to embodiments of the present disclosure;

[0028] Figures 9a and 9b schematically show views of model of a detector element and shielding element with a hemispheric field of view modelled from an evaluation point in the detector element, according to embodiments of the present disclosure;

[0029] Figure 10a shows a plot of a solid angle fraction not subtended by a target object in a field of view from an evaluation point, according to embodiments of the present disclosure;P131108GB

[0030] Figures 10b to 10g show polar plots of the intersections of a plurality of vectors with the surface of a unit hemisphere used to model a field of view at the evaluation location, according to embodiments of the present disclosure;

[0031] Figures 11a and 11b schematically illustrate variation in illuminated detector volume with different incident radiation angles, according to embodiments of the present disclosure;

[0032] Figures 12a and 12b are plots illustrating variation of a radiation detector fraction illuminated by incident radiation for different azimuthal and polar angles respectively, according to embodiments of the present disclosure;

[0033] Figures 13a to 13D are plots illustrating comparisons between an estimated degree of background suppression by a target object at an evaluation location with real-world suppression measured in radiation data collected at the same evaluation location;

[0034] Figure 14 schematically shows radiation monitoring control logic according to embodiments of the present disclosure;

[0035] Figure 15 is a flow chart schematically showing steps of a method according to embodiments of the present disclosure.

[0036] DESCRIPTION OF THE EMBODIMENTS

[0037] Referring now to the drawings, wherein like reference numerals designate identical or corresponding parts throughout the several views, embodiments of the present disclosure are described in detail.

[0038] The present technique relates to estimating suppression of background radiation, by a target object, at an evaluation location. For the purposes of describing approaches according to the present disclosure in the context of a concrete use case, the evaluation location will typically be described herein as being associated with a radiation monitoring apparatus comprising a radiation detector. A radiation portal monitor (RPM) is used herein as an illustrative example of a (nuclear) radiation monitoring apparatus, which may comprise circuitry and I or software configured to carry out the methods described herein. However, it will be appreciated that a RPM is only one embodiment of a radiation monitoring apparatus with which methods according to the present disclosure can be implemented, and circuitry and / or software implementing methods of the present disclosure are suitable for use with any known class of radiation monitoring apparatus (e.g. nuclear radiation monitoring apparatuses) comprising at least one radiation detectors (e.g. gamma and I or neutron detector).

[0039] Thus, in addition to implementation in RPMs, it will be appreciated the methods described herein are suitable for use in mobile radiation monitoring apparatus such as (i) handheld, (ii) backpack, and (iii) vehicle mounted devices. Accordingly, any reference herein to an RPM can be substituted for a mobile radiation monitoring apparatus, including specifically mobile nuclear radiationP131108GB

[0040] monitoring apparatuses, and vice versa. More generally, the skilled person will appreciate from the present disclosure that methods according to the present disclosure, and software and I or circuitry implementing said methods, can be used in any other contexts where it is of interest to the skilled person to estimate a degree of suppression of background radiation at an evaluation location, by a target object, whatever the source of background radiation is in a particular context. While such estimation of background suppression has been recognised by the inventors to have particular utility in the context of radiation monitoring apparatuses (and specifically nuclear radiation monitoring apparatuses), nothing in the present disclosure should be considered as excluding the applicability of circuitry, software, and I or methods, described herein from other use contexts. Indeed, the methods used herein can be used for estimating suppression of background radiation at any location, regardless of whether or not such a location is coincident with a particular apparatus or not.

[0041] Thus, the present disclosure details circuitry, software, and methods, for estimating suppression of background radiation, by a target object, at an evaluation location in a manner that is agnostic to any specific component or apparatus associated with the evaluation location. However, the present disclosure also details use cases where the evaluation location corresponds to a location in or on a radiation detector of a radiation monitoring apparatus (which can comprise respectively a nuclear radiation detector of a nuclear radiation monitoring apparatus). In the latter set of contexts, where the evaluation location corresponds to a location where radiation (such as nuclear radiation) is detected, the processing of the detected radiation by the radiation monitoring apparatus can take into account the estimate of suppression of background radiation at the evaluation location. As discussed further herein, this can reduce the likelihood that variations in background radiation suppression at the evaluation location (and thus at the detector used to obtain radiation data) will lead to misclassification of radiation data associated with a target object that monitored by the radiation monitoring apparatus.

[0042] Figure 1 shows an exemplary Radiation Portal Monitor (RPM) 1 with which methods, software, and I or circuitry may be implemented according to embodiments of the present disclosure. The RPM 1 comprises a plurality of radiation detector panels 100A to 100D (which form respective portions of an overall radiation detector system). It will be appreciated a single detector might be included in other contexts, and that the specific number of detector panels is not of particular significance to the approaches set out herein. In RPM 1, detector panels 100A and 100C form a right side (R) of the detector and panels 100B and 100D form a left side (L) of the detector. The R and L sides of the detector system in RPM 1 are placed apart on either side of a pathway (i.e. road, watercourse, or set of tracks) so as to allow a target object 102 (here a vehicle) passing through the RPM to travel between the R and L sides of the detector system. The panels of each of the R and L sides of the detector house one or more gamma ray detectors and / or one or moreP131108GB

[0043] neutron detectors, thereby allowing gamma radiation and I or neutron radiation to be detected at each of the R and L sides of the vehicle.

[0044] Gamma-ray and neutron detectors are instruments that respond, respectively, to gamma-ray and neutron emissions emanating from both a target object being monitored I assessed (e.g. a cargo comprised in or transported by a vehicle passing through the RPM) and the surrounding environment. A dual-sided (in particular, a R and L sided) detection system is presented as an example in Figure 1. However, the present approaches are applicable to other possible arrangements of detectors, including, but not limited to, single-sided deployments (e.g. R side or L side only) or multi-sided deployments (e.g. R and L side, as shown in Figure 1). In one example, the detector system of RPM 1 may comprise detector panels above and / or below the cargo in addition to or instead of or in addition to detectors panels at the L and R sides of the cargo. The cargo being scanned is not limited to containerised cargo associated with a road vehicle, but could equally comprise pedestrian traffic, air-freight, rail-freight, non-containerised cargo, automobiles, watercraft, etc. It will be appreciated that, depending on a particular class of target object 102 to be assessed, the example RPM of Figure 1 will be adjusted accordingly.

[0045] Each gamma-ray detector, where present, may be configured to record data indicative of an energy-loss spectrum of gamma radiation (gamma-ray spectrum) over a given time interval. This data can be referred to as gamma-ray data, and is a sub-set of radiation data as described herein. Each gamma-ray detector may be, for example, a plastic scintillator detector (in particular, a polyvinyl-toluene scintillator detector), a garnet ceramic scintillator detector (and I or other detectors known to the skilled person based on garnet and I or ceramic technology), a crystal scintillator detector, or any other gamma-ray detector known to the skilled-person. The output of the gamma-ray detector will typically comprise a rate of detection events (referred to herein as a ‘count rate’), typically segregated into ranges of energy, such that separate rates are discriminated for different ranges of energy across a detectable range, enabling computation of a spectrum. The detectable energy range and noise floor of count rate will be a function of the particular detector design, and differs across different detector types. The determination of count rates associated with different energy levels is typically achieved using multi-channel analyser (MCA) control logic, implemented in hardware and I or software, connected to radiation detector hardware, and configured to process voltage pulses generated by each of one or more gammaray detectors in response to excitation by incident radiation, converting them to a representation of count rate with respect to ranges of incident energy. MCA control logic may be integrated into a gamma-ray detector unit (e.g. a detector panel such as panels 100A to 100D shown in Figure 1), such that each gamma-ray detector (or panel) comprises an inbuilt MCA which analyses the raw data from each detector and returns radiation data comprising an indication of detection event count rate with respect to energy. Alternatively an MCA may be implemented as a standaloneP131108GB

[0046] module of control logic configured to receive raw radiation data generated by one or more radiation detectors of a radiation monitoring apparatus (such as the RPM 1 of Figure 1), and configured to pass processed radiation data to radiation monitoring circuitry 190 configured to perform further processing steps such as classifying detected radiation using the radiation data to identify at least one radioactive isotope comprised in a target object from which the detected radiation is received. In other embodiments, MCA functionality may be integrated into monitoring circuitry 190 of an RPM 1, as shown in Figure 14.

[0047] It will be appreciated that the functionality of MCA I analyser control logic, and indeed any of the control logic or ‘circuitry’ described herein, may be implemented as hardware control logic (e.g. one or more application specific integrated circuit (ASIC) or field programmable gate array (FPGA) modules), or may be implemented as one or more modules of software running on processor circuitry of a general purpose computing device, configured to process commands via one or more CPU and I or GPU modules. Thus the MCA I analyser control logic may be implemented as and referred to as either a hardware analyser, or a software analyser. In Figure 1, (nuclear) radiation monitoring control logic 190 is shown, which as discussed further herein can, in embodiments, integrate one or more of the functions which are herein associated with MCA I analyser control logic, image acquisition control logic, model generation control logic, and field of view detection control logic, and indeed any other control logic I circuitry described herein.

[0048] Similar to the gamma-ray detector described above, each neutron detector, where present, is configured to record data indicative of a flux of incident neutrons over a given time interval. This data may be referred to as neutron data. The raw sensor output of one or more neutron detectors may, in a similar manner to the sensor output of one or more gamma-ray detectors, be processed by one or more modules of MCA control logic to discriminate count rates by incident energy. Suitable gamma-ray and neutron detectors for (nuclear) radiation monitoring apparatuses are known in the art and are therefore not discussed further in detail here.

[0049] In the context of the RPM of Figure 1 , each of the radiation detectors comprised in detector panels 100A to 100D is in data communication with radiation monitoring control logic 190 so as to allow the generated radiation data to be provided to the radiation monitoring control logic 190. This data transmission between each radiation detector and the radiation monitoring control logic 190 is achieved using wired or wireless data communication approaches known to the person skilled in the art. A user-feedback apparatus 191 (e.g. monitor comprising a liquid crystal display (LCD) or similar for displaying images, and / or a speaker module for providing audible alarms) may be connected to the radiation monitoring control logic to enable the radiation monitoring control logic 190 to provide feedback (e.g. in the form of an alarm) to a user.P131108GB

[0050] Where the use context for the radiation monitoring control logic of the present disclosure is an RPM, the RPM will typically also comprise a vehicle or pedestrian detection apparatus for detecting when a vehicle or pedestrian has entered the RPM. As shown in Figure 1, a vehicle detection apparatus comprises one or more pairs of beam-break instruments, each pair comprising a beam emitter 103A which emits a light beam and a beam detector 103B which detects said light beam. Beam emitters and detectors which perform such functions are known in the art and are therefore not discussed in detail here. The light beam may, for example, be a laser beam. One pair of beam-break instruments is typically positioned at the entrance to the RPM 1 (i.e. upstream of the detector panels) and one pair of bream break instruments is typically positioned at the exit of the RPM (i.e. downstream of all the detector panels). The distance (d) between each pair of beam-break instruments is a known parameter stored in a memory of the radiation monitoring control logic 190 of the RPM 1. Figure 1 shows an entrance beam-break pair including beam emitter 103A and beam detector 103B. Although not shown, a similar arrangement (exit beam-break pair) is present at the exit to the RPM. The respective positions of the pairs of beam-break instruments need not necessarily be at the entrance and exit of the RPM. Rather, they may be at different positions of the RPM in the direction of travel of the vehicle, as long as the distance between different sets of pairs along a transit path for target objects through the RPM is known. It will be appreciated the use of break-beam instruments is described herein for the sake of providing a concrete example of a context in which approaches according to the disclosure may be implemented, and other ways of detecting the occupancy of parts of a target object 102 atone or more positions along the RPM 1 could also be used instead of pairs of beambreak instruments. In this respect any form of occupancy and I or velocity detection approach known to the skilled person may be used in place of or in addition to break-beam instruments 103A and 103B shown in Figure 1 and described herein. What is significant is that the RPM 1 comprises a mechanism for triggering measurements to be recorded from the radiation detector system (here comprising detector panels 100A to 100D); for radiation data processing steps to be implemented by radiation monitoring control logic 190, as described further herein, when occupancy of the RPM is detected); and for triggering recording of measurements from the radiation detector(s) to be stopped, when occupancy of the RPM 1 is no longer detected). Where the radiation monitoring apparatus in which the radiation monitoring control logic 190 of the present disclosure is implemented is a different type I class of apparatus to an RPM 1 , such as a hand-held, backpack, or mobile radiation monitoring apparatus, it will be appreciated the triggering of measurement and data processing routines will typically not be implemented using beam-break or other occupancy-detection instruments, but by alternative user input devices (for example, suitable mechanical controls, or a user interface provided on a display panel of a general purpose computing device).P131108GB

[0051] While Figure 1 schematically illustrates an RPM 1 which comprises a plurality of radiation detectors, it will be appreciated approaches according to the present disclosure may be applied in respect of radiation monitoring control logic configured to receive radiation data from a single radiation detector. Methods according to the present disclosure may be applied separately for each radiation detector of a radiation monitoring apparatus which comprises a plurality of radiation detectors, or may be applied in respect of a radiation monitoring apparatus comprising a single radiation detector. A module comprising radiation monitoring control logic 190 as described herein may therefore be provided on a per-detector basis, or on a per-radiation-monitoring-apparatus basis.

[0052] While not wishing to be constrained by any particular physical theory, radiation data captured by one or more radiation detectors of a radiation monitoring apparatus in proximity to a target object to be assessed is typically understood to comprise at least two broad classes of components. These broad classes are (1) a measure of any radiation emitted from the materials comprised within the target object, or from any other source within or on the target object, and (2) a measure of other sources of radiation or noise which do not directly result from radiation emission characteristics of the target object. It will be appreciated that this second broad class of components may comprise one or more of a plurality of sub-components, such as a component associated with passive environmental emission (e.g. from the ground, the atmosphere, materials in the built environment, and other naturally occurring radioactive material (NORM) sources), a noise component, an active interference component, and so on. This second class of radiation components - in other words, those not associated with radiation emissions from the target object - may generally be referred to herein as ‘background radiation’.

[0053] The classification of detected radiation using radiation data acquired by a radiation monitoring apparatus to detect, for example, the presence of a radiation source (e.g. a radioactive isotope) comprised in a target object, must therefore account for the presence of background radiation components, which form part of the radiation data but are not associated with emission from the target object. It is of interest to be able to more accurately and I or quickly determine the influence of background radiation at an evaluation location, such as but not limited to an evaluation location coincident with a radiation detector of a radiation monitoring apparatus such as an RPM.

[0054] A particular way in which background radiation can influence analysis of radiation at such an evaluation location is the effect of shielding I attenuation of background radiation by a target object located between the evaluation location and at least one background radiation source. For instance, in an RPM 1 as shown schematically in Figure 1 , if a uniform field of incident background radiation is assumed at an evaluation location 110 positioned in or on a detector panel 100A (where the location may be a point, area, or volume), the presence of a target object 102 (e.g. a vehicle) in a field of view at the evaluation location will shield the incident background signalP131108GB

[0055] approaching the evaluation location through the field of view, doing so over regions where the material of the target object 102 is interposed between the evaluation location 110 and the source(s) of background radiation emission. Other physical objects proximate the evaluation location 110 (i.e. parts of the RPM 1) may also contribute to such shielding, but unlike the target object, these may typically remain static over long periods of time. However, variation in factors such as the shape and radio-density distribution of the target object, and the spatial relationship of the target object to the evaluation location 110, can lead to variation in radiation count rates at the evaluation location 110 over time, as a consequence of the differential attenuation with respect to time of the background radiation component by the target object.

[0056] Figures 2a to 2e show schematically a scenario where a target object 102 (which here is a vehicle, but could in principle be any radiation-shielding object) moves past an evaluation point 110. For the purpose of providing a concrete example, the evaluation point is illustrated in the context of a radiation monitoring apparatus configured as an RPM 1. Thus, Figures 2a to 2e show schematically a top view of the RPM 1 shown in Figure 1, and elements not explicitly shown (e.g. detector panels 100C and 100D, radiation monitoring control logic 190, and display apparatus 191), can be assumed to be present in the manner described in relation to Figure 1. Elements labelled in Figure 2a can be assumed to be present in the same manner in each of Figures 2b to 2e.

[0057] Figure 2a shows detector panels 100A and 100B, configured as described in relation of Figure 1. Also shown is an entrance break beam pair comprising beam emitter 103A and beam detector 103B, and exit break beam pair comprising beam emitter 104A and beam detector 104B. Figure 2a shows a scenario at a time t1, where a target object 102 is approaching the RPM with a constant velocity given by vector v. The target object 102 is at this point outside the RPM 1, upstream of the entrance break-beam pair. The radiation monitoring control logic 190 consequently does not detect occupancy of the RPM 1, and has not triggered collection of radiation data from the radiation detector panels 100A and 100B. In Figure 2b, at time t2, the target object 102 has moved to cross the entrance break beam pair, with the front of the target object 102 now coincident with a vector normal from detector panel 100A at the evaluation location 110. The radiation monitoring control logic 190 has received a signal from the entrance break beam pair when the front of target object 102 crossed the beam axis, and triggered collection of radiation data from the radiation detector panels 100A and 100B, and thus by time t2, radiation data collection by the radiation monitoring control logic 190 is already in progress (having started between t1 and t2). In Figure 2c, at time t3, the front of target object 102 has moved further into the RPM 1 to cross the exit break beam pair, with a geometric centre of the target object 102 now coincident with a vector normal from detector panel 100A at the evaluation location 110. Radiation data collection by the radiation monitoring control logic 190 is still inP131108GB

[0058] progress at t3, since the exit break beam pair has not yet detected the target object 102 leaving the RPM 1. In Figure 2d, at time t4, the back of target object 102 has moved further through the RPM 1 to exit the entrance break beam pair, with the back of the target object 102 coincident with a vector normal from detector panel 100A at the evaluation location 110. The back of target object 102 has not yet exited the exit break beam pair, and radiation data collection by the radiation monitoring control logic 190 is still in progress at t4. In Figure 2e, at time t5, the back of target object 102 has moved out of the RPM 1, to exit the exit break beam pair, and the entire target object 102 is once again outside the region between the detector panels 100A and 100B. The radiation monitoring control logic 190 receives a signal from the exit break beam pair when the back of target object 102 exits the beam axis, and ceases the collection of radiation data from the radiation detector panels 100A and 100B. Thus by time t5, radiation data collection by the radiation monitoring control logic 190 has already ended (having ended between times t4 and t5). Figure 3 shows schematically an idealised profile of suppression of background radiation at the evaluation point 110, where this profile corresponds to that produced as the target object 102 progressed through the RPM 1 over the time series comprising t1, t2, t3, t4, and t5, and the respective spatial positions shown schematically over Figures 2a to 2e. For the purposes of explanation herein, an idealised scenario is considered in which the background radiation comprises a uniform incident radiation field around the evaluation location, which does not exhibit temporal variation.

[0059] At time t1 in Figure 3, the target object 102 is distant from the evaluation location 110 (as per Figure 2a), such that the shielding of portions of the background radiation field at the evaluation location 110 by the target object 102 is comparatively low. In other words, only a small proportion of the total incident background radiation travelling towards the evaluation location 110 intersects with portions of the target object 102, and thus the suppression of background radiation at the evaluation point 110 (i.e. in terms of a reduction in background count rate due to the interaction of the target object 102 with incident background radiation), is low. The background radiation count rate at the evaluation location has not dropped significantly relative to its steady-state value, as shown schematically at t1 in Figure 3. At time t2 in Figure 3, the target object 102 has moved significantly closer to the evaluation point 110 (as per Figure 2b), such that it shields a larger proportion of the incident background radiation field from reaching the evaluation location 110. This increased suppression of the background radiation component as compared to the state at t1 is shown in schematically at t2 in Figure 3. At time t3 in Figure 3, the target object 102 is maximally shielding the evaluation point 110 from background radiation, since the intersection of its volume with the incident background radiation travelling towards the evaluation point is at a maximum (as per Figure 2c). Thus as shown at t3 in Figure 3, the suppression of background radiation has reached its maximum magnitude. At time t4 in Figure 3, the degree of shielding ofP131108GB

[0060] the background radiation field at the evaluation location has reduced relative to that at t3 (as per Figure 2d), and as shown at t4 in Figure 3, the degree of suppression of the background count rate has thus decreased as compared to t3. Finally, at time t5 in Figure 3, the target object 102 is again distant from the evaluation point 110 (as per Figure 2e), such that the shielding of the background radiation field, and thus suppression of background count rate at the evaluation location 110, is again comparatively low. In other words, the background radiation count rate at the evaluation location has returned to be significantly similar to its steady-state value, as shown schematically at t5 in Figure 3.

[0061] Such temporal variation in the shielding of a background radiation component at an evaluation location, as a consequence of changing spatial relationship over time of a shielding target object to the evaluation location, can be problematic, such as, for example, when the evaluation location represents a location at which radiation data is being collected as part of a (nuclear) radiation monitoring protocol. If, for example, a metric such as radiation count rate at the evaluation location, is used as a parameter to trigger a determination that a target object is an emitter of radiation, then a reduction in count rate resulting from temporal variation in shielding of background radiation by that target object may serve to offset an increase in count rate resulting from radiation emission (e.g. nuclear radiation emission) from one or more sources within the target object itself. This could lead to the presence of a radiation source within the target object failing to be detected.

[0062] This scenario can be better understood with respect to the illustrative, idealised radiation count profiles of Figure 3. The times at which a target object 102 respectively enters the RPM 1 of Figure 2 (i.e. first crosses the entrance break beam pair) and exits the RPM 1 (i.e. moves beyond the exit break beam pair) are indicated in Figure 3 as tent and texrespectively. This corresponds to the period over which radiation data is collected by the radiation monitoring control logic 190 of the RPM 1 as described in relation to Figures 1 and 2. If it is assumed that the target object 102 contains a radiation emitter (for example, an illicit special nuclear material (SNM) source), it will be appreciated that over the time period t1 to t5, as the target object approaches the evaluation location 110, and moves away from it again, the count rate of the radiation component associated with emission from the target object will increase, reach a maximum, and then decrease. This is shown schematically via the solid radiation count profile in Figure 3. The dot-dash line shows the total radiation count profile, comprising the combined contribution at the evaluation location of target and background radiation components (i.e. indicated by the solid and dashed profiles respectively).

[0063] In the idealised example of Figure 3, as a consequence of the geometry of the RPM 1 shown in Figures 1 and 2, and the corresponding emission rates by the target object and background radiation source(s), an increasing suppression of the background component with increasingP131108GB

[0064] proximity of target object and evaluation location offsets a corresponding increase in the target component, leading to the total radiation count rate exhibiting no change over the evaluation time (i.e. the period of RPM occupancy between times tent and tex). In this case, it would not be possible to determine from the total count rate during evaluation of the target object that a radiation emitter is present in the target object. While this is an idealised scenario for the purposes of explanation, it will be appreciated that the effect of shielding of background radiation at an evaluation location by a target object can be problematic in any scenario where radiation data collected at that evaluation location is being used to attempt to determine the presence of, and I or otherwise characterise, a radiation emission source in or on said target object.

[0065] The inventors have recognised a need for improved methods for characterising the suppression of background radiation by a target object, at one or more evaluation locations. As set out further in the present disclosure, the inventors have recognised this can be achieved by estimating a region of a field of view at each evaluation location which is occupied by the target object; and estimating a degree of suppression of background radiation at each evaluation location, by the target object, based on the region of the field of view at each evaluation location which is estimated to be occupied by the target object. It will be appreciated from the disclosure herein that such approaches have a particular set of applications in contexts where the evaluation location corresponds to a location in or on a radiation detector of a radiation monitoring apparatus (e.g. a nuclear radiation detector of a nuclear radiation monitoring apparatus, such as a RPM, or mobile apparatus). However, the present disclosure should not be considered as being limited to such contexts, and the evaluation location can in principle be any location in 3D space at which the skilled person may consider it advantageous to quantify a degree of suppression of background radiation by a target object, whether or not this is a location at which a radiation detector is located, and whether or not the degree of suppression of background radiation is used as an input in processing of radiation data obtained by a radiation detector.

[0066] It is common to each of the approaches exemplified herein that estimation of a degree of suppression of background radiation at an evaluation location, by a target object, is based on estimation of a region of a field of view at the evaluation location which is occupied by the target object. A useful concept in understanding the present disclosure is that of the solid angle subtended at the evaluation location by a target object. As a target object approaches an evaluation location (as schematically shown for target object 102 and evaluation location 110 in Figures 2a to 2c), the proportion of the viewing angle of the evaluation location subtended by the target object will typically increase. If a viewing angle at the evaluation location corresponds to a viewing angle over which incident radiation can be received at the evaluation location, then the inventors have recognised that the fraction of this viewing angle subtended by a target object in the field of view will be proportional to the degree of suppression of a uniform radiation fieldP131108GB

[0067] incident on the evaluation location. It will be appreciated that the viewing angle at an evaluation location that is subtended by a target object will be a function of at least the shape and position of the target object relative to the evaluation position. Thus, approaches described herein provide means to characterise the spatial extent and I or position of a target object relative to an evaluation position, so that a region of a field of view at the evaluation location (e.g. a solid angle subtended in a field of view at the evaluation location) can be estimated from this information. In other detailed approaches described herein, the region of the field of view at the evaluation location (e.g. a solid angle subtended in the field of view at the evaluation location) can be directly estimated without requiring a separate step of quantifying the shape of the target object itself, and I or its position relative to the evaluation location.

[0068] As will be evident to the skilled person from the present disclosure, there are a number of specific approaches which may be used to determine, calculate, derive, or otherwise quantify the region of a field of view at an evaluation location which is occupied by a target object. While the estimation of the region of the field of view occupied by the target object may be based on imaging and I or other sensor data obtained in the vicinity of the evaluation position, it may also be based on imaging and I or other sensor data obtained distant to the evaluation position. The skilled person will recognise that information about a spatial extent of a target object can be spatially and I or temporally interpolated in a suitable reference frame, to register it to a specific location (i.e. a position having a known spatial relationship with an evaluation position).

[0069] Approaches according to the present disclosure will now be described in more detail.

[0070] According to some embodiments of the present disclosure, estimating the region of the field of view at the evaluation location which is occupied by the target object comprises: obtaining a model characterising a spatial extent of the target object with respect to the evaluation location; and estimating a region of a field of view at the evaluation location which is occupied by the target object based on a geometric relationship between the model and the evaluation location.

[0071] In a first set of examples, a model of a target object is generated based on image data representing the target object. What may be considered significant in these examples is that the image data allows spatial information about the target object to be derived, such that a model characterising a spatial extent of the target object can be generated. The skilled person will appreciate that there are numerous ways in which a geometric model of a target object can be constructed based on images of the target object. In some instances, a single image of a target object from a single imaging location may be sufficient to derive a model characterising a spatial extent of the target object, particularly if certain assumptions are made as to the geometry of the target object (i.e. if it has a known, predefined spatial extent in one or more dimensions). In other instances, a model may be derived based on a plurality of images of the target object obtainedP131108GB

[0072] by at least one image sensor, where the images are obtained with respect to a plurality of nonparallel imaging orientations. Generating a model of a target object using a plurality of images of the target object obtained along non-parallel imaging orientations may be referred to herein as ‘reconstructing’ a model based on said plurality of images.

[0073] It will be appreciated herein that the target object can be imaged ex situ from the evaluation location at which background suppression by the target object is to be estimated, and the influence on background suppression of the target object being in a specific spatial position relative to the evaluation location can be estimated by registering an image-based model to said spatial position in a virtual scene. In other words, the effect of a real-world target object on suppression of background radiation at a real-world evaluation location, based on the spatial relationship of the target object to the evaluation location, can be simulated to estimate the degree of suppression, by (1) modelling the target object, (2) modelling an evaluation location having the same specific spatial relationship of interest to the model of the target object as the real world evaluation location has to the physical target object, and (3) estimating using a computational approach the region of a simulated field of view at the modelled evaluation location which is estimated to be occupied by the model of the target object.

[0074] Where the evaluation location is coincident with a radiation detection apparatus (such as an RPM), one or more image acquisition systems for obtaining image data used to derive a model of the target object may be integrated into the radiation detection apparatus (e.g. an RPM), as described in more detail herein. However, image data of a target object may alternatively be acquired at one or more imaging locations distant from the evaluation location. The skilled person will appreciate that provided the image acquisition system(s) used to obtain images for deriving a model of the target object are calibrated to allow determination of the spatial extent of the target object, it is not essential that the image acquisition system(s) be located near the evaluation location at which suppression of background radiation is to be estimated.

[0075] Herein, an RPM (of the general kind shown schematically in Figures 1 and 2) will be used as an exemplary context for the sake of providing a concrete use case for understanding embodiments of the present disclosure. However, it will be appreciated that the use of a model characterising a spatial extent of a target object with respect to an evaluation location is an approach that can be applied in estimating suppression of background radiation by the target object at the evaluation location in other contexts where the evaluation location is not associated with a specific type of radiation monitoring apparatus.

[0076] Figure 4A schematically shows an RPM 2 which is similar to and will be understood from the RPM 1 of Figures 1 and 2, but which further comprises at least one image acquisition system (235, 255). In this example, a first image acquisition system 235 comprises an X-ray source 220 andP131108GB

[0077] an X-ray imaging detector 225, positioned along the x direction (i.e. a vector normal to a z direction along which the target object 202 passes along a transit path through the RPM 2). The X-ray source 220 is configured to emit a beam of X-rays at a side of a target object 202 within the RPM 2, with the X-ray imaging detector 225 collecting one or more plane radiographs of the resulting X-ray signal as the target object 202 passes along its transit path. The X-ray image detector 225 may, for example, comprise one or more flat-panel detectors as known to the skilled person (e.g. a 2D photodiode array, coupled to a scintillation screen), with the imaging area of the detector being configured to allow the entire vertical extent of the target object 202 to be imaged. This may be achieved either by configuring the X-ray imaging detector 225 as a single detector panel or as a plurality of tiled detector panels which extend from the ground to a position higher than the tallest expected target object, or by mounting one or more detector panels on a manipulator allowing at least vertical translation of the detector to acquire images at different vertical positions. The images can be stitched together computationally using information about the translation distance of the detector between each image acquisition, to form an overview image in which the entire vertical extent of the target object is captured.

[0078] Image acquisition control logic 230 is connected to the X-ray source 220 and X-ray imaging detector 225 using conventional wired or wireless data transfer means, to control the acquisition of X-ray images by synchronising the turning on of the X-ray source 220 and the acquisition of images by the X-ray imaging detector 225, according to approaches known to the skilled person. For example, in embodiments of the disclosure, the image acquisition control logic 230 receives information from the entrance break beam pair comprising beam emitter 203A and beam detector 203B, and exit break beam pair comprising beam emitter 204A and beam detector 204B, to allow an X-ray image acquisition procedure implemented by the image acquisition control logic 230 to be triggered when the entrance break beam pair detects the entry of the target object 202 into the RPM, and to be ended when the exit break beam pair detects the exit of the target object 202 from the RPM. Thus, in one example, when the image acquisition control logic 230 receives a signal from the entrance break beam pair that a target object 202 has entered the RPM 2, the image acquisition control logic 230 triggers the X-ray source 220 to turn on and triggers the X-ray image detector 225 to begin acquiring radiographs. In this example, a continuous acquisition of radiographs is triggered, with the X-ray source 220 remaining on while the X-ray image detector 225 collects sequential radiographs with a suitable exposure time and other imaging parameters (as determined by the skilled person) as the target object 202 moves through the RPM 2 along the transit path v, parallel to the z direction. This acquisition procedure continues until the image acquisition control logic 230 receives a signal from the exit break beam pair that the back edge of the target object 202 has exited the RPM 2, at which point the image acquisition control logic triggers the X-ray source 220 to turn off, and ends the acquisition of X-ray images (i.e. radiographs) by the X-ray image detector 225. The sequence of images is stored in a suitableP131108GB

[0079] memory element either integrated in the image acquisition control logic 230, or hosted, for example, in cloud storage. It will be appreciated the image acquisition control logic 230 may be implemented in any way known to the skilled person, for example, as a general purpose computing device running one or more software packages controlling the functionality attributed to the image acquisition control logic 230 herein. It may also be integrated into radiation monitoring control logic 290 of a radiation monitoring apparatus 2, as described in relation to Figure 14.

[0080] In this example, the image acquisition control logic 230 of the first image acquisition system 235 is configured to reconstruct a 2D transmission radiograph of the target object 202 by stitching together the plurality of stored radiographs of the target object, obtained by the X-ray image detector 225 during transit of the target object 202 through the RPM 2. Information on the speed of transit of the target object 202 past the X-ray imaging detector 225 is used in this example to calibrate the stitching together of the radiographs, by dividing the speed of transit (i.e. in m / s) of the target object 202 along a transit path Pznormal to the imaging direction x, by the frequency of image acquisition (i.e. in Hz), to obtain, for each pair of images in the sequence, an offset value which represents the distance the target object 202 moved in the horizontal direction in the image plane, between the acquisition times of the first and second images in the pair. Based on calibration of the imaging system (e.g. using reference objects of known dimensions, according to approaches known to the skilled person), this distance can be converted to a translation expressed in a reference frame for the image itself (i.e. in pixels), to be applied between each sequential pair of images in the sequence, in order to reconstruct from the sequence of images a single 2D radiograph in which the entire outline of the target object 202 is visible (in side view in this example).

[0081] The speed of transit of the target object 202 through the RPM 2 along transit path Pzcan be determined in a number of different ways. For example, where the speed can be assumed to be constant, this can be derived based on dividing the elapsed time between the target object passing entrance and exit break beam pairs (i.e. the time tex-tent) by the predefined, known distance d between the break beam pairs along the direction of travel of the target object (i.e. along the transit path Pzthrough the RPM). The speed may also be derived using one or more ultrasonic sensors, by tracking the target object 20 using one or more cameras, or using GPS tracking, as non-limiting examples. In other examples, the target object 202 may be towed, pushed, or otherwise conveyed through the RPM 2 via a mechanical conveying system such as a winch or continuous belt system, configured to translate the target object past the image acquisition system at constant speed. The skilled person will appreciate that a number of different approaches are available for monitoring or otherwise deriving the speed of a target object 202 relative to an image acquisition system location, and any of these can be used.P131108GB

[0082] Stitching of the images obtained by the first image acquisition system 235 results in an overview image 2021 as shown in Figure 5a, which is an exemplary 2D overview radiograph showing a side view of the entirety of the imaged target object (which is this example is a vehicle).

[0083] Optionally, imaging may also be conducted along a different imaging direction to that of the first image acquisition system, such as for example, a vector normal to both the transit path Pzof the target object 202 through the RPM 2 (this path being parallel to the z axis), and to the imaging axis (i.e. the beam direction) of the first image acquisition (this axis being parallel to the x axis). Thus, for example, as shown in Figure 4b, the RPM 2 may comprise a second image acquisition system 255. In this example, the second image acquisition system 255 is configured to acquire radiographs based on transmission of X-rays from a second X-ray source 240 mounted on a gantry above a roadway along which target objects 202 transit through the RPM 2, towards an X-ray image sensor 245 mounted on or below a roadway over which they transit. The second image acquisition system 255 in this example thereby generates transmission radiographs acquired along an imaging direction orthogonal to those acquired by the first image acquisition system 235. The acquisition of 2D radiographs, and stitching of these into a single overview radiograph, using the second image acquisition system 255, can be carried out by image acquisition control logic 250 of the second image acquisition system 255 according to the same process described in respect of the first image acquisition system 235. However, it will be appreciated where there are plural image acquisition systems (e.g. 235 and 255), these may share one instance of image acquisition control logic (e.g. 230) responsible for controlling X-ray imaging sources, X-ray detectors, image acquisition, storage, and image stitching, for all image acquisition systems (i.e. the image acquisition control logic may parallelise these processes for each of a plurality of image acquisition systems).

[0084] Thus, in this example, a target object 202 (e.g. vehicle) passing through the RPM 2 is imaged with respect to two orthogonal imaging orientations (respectively parallel to the x and y axes, and both normal to the transit path of target objects). The result of the process of image acquisition from the first and second image acquisition systems (235, 255) is respectively the generation of overview images showing side and top views of the entire outline of the target object 202. Figures 5a and 5b show respectively exemplary 2D radiographs 2021 and 2022 of a target object comprising a truck carrying a container of cargo which has been imaged respectively by first and second image acquisition systems (235, 255) according to the processes described herein. Radiation monitoring control logic 290, having functionality as described in respect of radiation monitoring control logic 190 of Figures 1 and 2, is connected to the detector panels and break beam pairs of the RPM 2, to trigger acquisition of radiation data from the detector panels based on occupancy of the RPM being indicated by the break beam pairs, and to store and process said radiation data as described further herein. Image acquisition control logic (230, 250) can beP131108GB

[0085] integrated into the radiation monitoring control logic 290 as described further herein (e.g. in relation to Figure 14).

[0086] Once image data has been obtained using one or more of first and second image acquisition systems as described herein, where the image data contains a representation of an outline of a target object, approaches for estimating suppression of background radiation at an evaluation location by the target object may comprise determining an indication of a spatial extent of the target object based on the image data; and estimating from the indication of the spatial extent a region of the field of view at the evaluation location which is occupied by the target object. As discussed herein, this may comprise obtaining via model generation control logic a model characterising a spatial extent of the target object with respect to the evaluation location, where the model is obtained based on analysis of the image data to determine an indication of a spatial extent of the target object based on the image data. The process of estimating from the indication of the spatial extent a region of the field of view at the evaluation location which is occupied by the target object may comprise field of view determination control logic estimating a region of a simulated field of view at a modelled evaluation location which is estimated to be occupied by the image-based model of the target object.

[0087] For the purposes of explanation herein, the control logic (e.g. circuitry or software) used to generate a model of the target object may be referred to as model generation control logic. This may be implemented as a standalone module, such as a general purpose computer configured to run control logic implementing approaches for model generation described herein. However, it will be appreciated the functions herein associated with the model generation control logic may in embodiments be carried out by the image acquisition control logic described herein. It will be further appreciated that different modules of control logic can be responsible for (1) acquisition and storage of image data of a target object acquired by one or more image acquisition systems, and (2) obtaining of a model of the target object from this image data, and (3) estimation of a region of a field of view at an evaluation location occupied by the model, that that these modules may be spatially separated from each other, and I or that the processes of acquisition, obtaining, and estimation, may be temporally separated. Figure 6 shows schematically model generation control logic 260 (e.g. circuitry or software) configured to receive data from image acquisition control logic 230 (e.g. circuitry or software) of the first image acquisition system 235, and image acquisition control logic 250 (e.g. circuitry or software) of the first image acquisition system 255. Typically, a model of a target object is derived by the model generation control logic 260 from image data obtained by at least one image acquisition system either by: (1) analysing each of a plurality of images acquired from non-parallel imaging orientations to distinguish between image regions associated with the target object and image regions associated with the background (i.e. not associated with the target object), and using this classification of image regions to reconstructP131108GB

[0088] a 3D model based on the known offset between the imaging orientations; or (2) by reconstructing a 3D volume (i.e. a volumetric image) from a plurality of images acquired from non-parallel imaging orientations, and then analysing the 3D volume to distinguish between image regions (e.g. voxels) associated with the target object and image regions associated with the background. The first approach may be considered to use 2D image-analysis I computer vision approaches to extract spatial information from 2D images, with this extracted information used to reconstruct a 3D model. The second approach may be considered to use tomographic reconstruction to obtain a 3D image I volume from which a 3D model can be directly extracted via 3D image-analysis I computer vision approaches. Such tomographic reconstruction approaches (e.g. filtered back-projection) are known to the skilled person, and can be implemented by the model generation control logic 260 using commercial or open source software.

[0089] Figures 5a and 5b show schematically aspects of a procedure applied by model obtaining control logic to obtain a 3D model of a target object from a plurality of images. Figure 5a shows the results of an edge-detection routine applied to a 2D overview radiograph 2021 showing a side view of a target object 202 imaged by a first image acquisition system 235 of RPM 2, and Figure 5b shows the results of an edge-detection routine applied to a 2D overview radiograph 2022 showing a top view of the same target object imaged by a second image acquisition system 255 of RPM 2. The edge detection approach in this example, applied to the 2D overview radiographs 2021 and 2022 by the model generation control logic 260, uses simple thresholding of the greyscale spectrum for each 2D radiograph to discriminate between pixels belonging to a ‘target object’ class, and pixels belonging to a ‘background’ (non-target-object) class. Points are then fitted to the boundary between these two classes (which in this example is achieved by searching down vertical lines of pixels in each image to find points at which a transition from one class to the other is found). A plurality of exemplary edge points 2031 are shown in Figure 5a, and a plurality of exemplary edge points 2032 are shown in Figure 5b. The edge points for each image provide a characterisation -in the image reference frame - of an estimated spatial extent of the target object in each image of the image data (e.g. by connecting edge points using lines). However, the skilled person will appreciate that classifying an image between target and background image regions can be carried out by model obtaining control logic according to any of a plurality of different computer vision approaches known to the skilled person. Some such approaches allow direct detection of edges between regions associated respectively with the target and the background. Other such approaches classify each of a plurality of image regions between target and background classes, with edge detection between said regions of the image belonging to these respective classes being implemented as a secondary computational step once the image has been so classified. In the present example, a 3D model is reconstructed by the model obtaining control logic using spatial extent data of target objects extracted from each of the plurality of 2D images, usingP131108GB

[0090] information about the offset(s) in imaging orientations used to acquire each of these images. The skilled person will appreciate there are different ways in which this step can be carried out, but what is significant is that non-parallel imaging orientations allow interpolation of information extracted from 2D images into 3D space. Thus, for example, a 2D ‘target object’ region classified as lying within the object boundary in Figure 5a can be computationally projected along a direction perpendicular to the plane of this image, with the extent in said direction of each portion (e.g. pixel) of the target object region being varied with respect to the z-direction in dependence on the target object width as defined by the 2D ‘target object’ region characterised as lying within the object boundary in Figure 5b. However, it will be appreciated a 3D model could be reconstructed from a single one of the 2D overview images by assuming the extent of the target object in the direction orthogonal to the overview image plane. For example, where the target object belongs to a class having a predefined width (e.g. a class of vehicle, such as a train or truck) a 2D targetobject region derived from a 2D side view may be projected by a distance in the direction normal to the image plane which is set based on the predefined width, to obtain a 3D model of the target object.

[0091] The result of the model reconstruction step is the characterisation of the spatial extent of the target object in the form of a simulated representation, which can be used to estimate (e.g. by simulation) a region of a field of view at an evaluation location which is occupied by the target object. Figure 7 shows schematically a 3D model 202’ of a target object 202, reconstructed by the model obtaining control logic 260 based on the classification of target object regions in each of the 2D radiographs shown in Figures 5a and 5b, according to a reconstruction approach as described herein. The model 202’ in this example comprises a set of cuboids defined with respect to the coordinate scheme of the system, where each cuboid is labelled as either associated with the target object, or associated with the background. As described further herein, and as also shown in Figure 7, the model 202’ can be placed by the model obtaining control logic (or field of view evaluation control logic) in a scene in which an evaluation location 210’ is also modelled. Figure 7 shows schematically a scene in which the model 202’ has been positioned in a specific spatial relationship to an evaluation location 210’, which here is positioned on a detector element of a model of a detector panel 200A’. In Figure 7, an RPM 2’ has been modelled in the scene, comprising models of detector panels 200A’ and 200B’, which are scaled to the model 202’ of the target object 202. The x, y, and z, axes show the scale of the scene in metres. The pathway between modelled detector panels 200A’ and 200B’ in this example is ~5 m wide in the x axis, and the length of the model in the z axis (i.e. the transit direction through the RPM 2’) is ~14 m. The dimensions and locations of elements of the model RPM 2’ can be defined based on any measurement technique known to the skilled person, and it is appreciated this will be different for different RPMs. Where the method herein is applied in respect of an evaluation location which isP131108GB

[0092] not associated with and RPM, it will be appreciated different detector geometry may be placed into the modelled scene that is appropriate to the specific context.

[0093] Once the model obtaining control logic has generated or otherwise obtained a model of the target object, whether based on image data generated by one or more image acquisition systems, or based on another approach as described further herein (e.g. obtaining a predefined model from a database), a region of the field of view at an evaluation location estimated to be occupied by the model is then estimated by field of view determination control logic, based on a spatial relationship between the model and the evaluation location.

[0094] Figure 6 shows schematically field of view determination control logic 270 configured to analyse a model of a target object generated or otherwise obtained by the model generation control logic 260 (e.g. based on image data from image acquisition control logic 230 of the first image acquisition system 235, and I or image acquisition control logic 250 of the first image acquisition system 255, shown in Figures 4a and 4b). It will be appreciated that while field of view determination control logic 270, model generation control logic 260, and image acquisition control logic 2301250 are shown as separate modules, they may be implemented as part of the same circuitry and I or software package. It will further be appreciated that the functions ascribed to any control logic herein may be implemented in hardware control logic, and I or by one or more software modules implemented on a processor (e.g. CPU and or GPU) of a suitable computing device, such as a general purpose computer.

[0095] Figure 8 will be recognised from Figure 7, and shows a modelled representation of two detector panels of an RPM 2 (as shown in Figures 4a and 4b) in a scene modelled by the model generation control logic 260. The detector panel models 200A’ and 200B’ are models of detector panels 200A and 200B as shown in Figures 4a and 4b, and these model detector panels 200A’ and 200B’ will be recognised as part of the RPM model 2’ in Figure 7. In the view of Figure 8, the detector panel model 200A’ is seen to comprise scale models of a detector element 205A’ and a shielding element 206A’ comprised within the modelled detector panel. The shielding element 206A’ extends around the back of the detector element 205A’ and simulates the presence of a shielding material (e.g. lead) which controls the field of view which is open to the detector element 200A’. The detector panel model 200B’ comprises a detector element 205B’ and a shielding element 206B’ configured in the same manner. The transit path Pz, parallel to the z axis, simulates the path along which target objects transit through the model RPM 2’ between the detector panel models 200A’ and 200B’. A model evaluation point 210’ is shown coincident with detector element 205A’, modelling the evaluation point 210 on detector panel 200A of the RPM 2 shown in Figures 4a and 4b.P131108GB

[0096] In embodiments of the disclosure, a region of a field of view at an evaluation location which is occupied by a target object is estimated by the field of view determination control logic 270 by determining whether each of a plurality of non-coincident paths intersecting a model of the evaluation location intersects with a model of the target object in the same scene. In embodiments of the present disclosure, this estimation can be considered to correspond to a method of simulating or otherwise computing a solid angle fraction at a simulated evaluation location which is subtended by a model of the target object positioned at a defined location relative to the simulated evaluation location.

[0097] Figures 9a and 9b schematically show 2D and 3D representations of a hemispheric field of view 250’ defined at an evaluation point 210’, which here is coincident with the detector element 205A’ of the detector panel model 200A’, as described and shown in relation to Figures 7 and 8. Figure 9a shows a cross section through the xz plane (i.e. transverse to a vertical I long axis of the detector element 205A’) at the position of the evaluation location 210’. In Figure 9b, the hemispheric field of view 250’ has been placed at an evaluation location (not shown), such that a full sphere comprising the hemisphere would be centred at said evaluation location. The hemispheric field of view 250’ is oriented towards the transit path Pzalong which target objects transit through the RPM. Where radiation shielding 206A’ is simulated, to approximate the effect of shielding material behind and to the side of a detector element, as typically used in RPM contexts, a hemisphere (i.e. covering 2TT steradians) as represented in Figures 9a and 9b may be considered a particularly suitable approximation of the field of view over which incident radiation may be received at an evaluation location 210’. However, it will be appreciated that other extents may be defined for the field of view, such as for example a full sphere (i.e. covering 4TT steradians), depending on the context of the evaluation location in question.

[0098] It will be appreciated that to estimate suppression of background radiation at a real-world evaluation location by a real-world target object having a specific spatial relationship to the evaluation location (in terms of proximity and orientation around its rotational degrees of freedom), the model of the target object and its position relative to a modelled evaluation location are scaled to the real-world scenario of interest. Thus the modelled scene represents, to scale, the spatial extent of a target object relative to an evaluation location in the real world.

[0099] Returning to Figure 7, this shows a model of a target object 202’, comprising an image-derived model of a truck carrying a cargo, positioned in a specific location relative to a simulated evaluation location 210’ defined on a detector panel model 200A’. The geometry of RPM 2 as shown in Figures 4a and 4b has been modelled, in that the evaluation location 210’ is simulated as the origin of a hemispheric field of view oriented perpendicular to a transit path along which a centroid of a model of the target object 202’ is translated during the simulation. In other words, a field of view 150’ can be considered to be defined at the evaluation location 210’ in Figure 7, asP131108GB

[0100] shown in Figures 9a and 9b. The field of view is defined as 2TT steradians. One or more elements of the RPM model 2’ (and particularly the evaluation location 210’) and the target object model 202’ are scaled to their real-world counterparts, and located so as to approximate a specific spatial relationship of interest between elements of a real-world RPM 2 and a real-world target object 202 of interest. With the target object model 202’ defined in a specific position in the scene (i.e. in terms of its rotational position within the coordinate system of the scene, and separation distance from the model centroid to the evaluation location 210’), the region of the field of view 205’ at the evaluation location 210’ which is occupied by the model 202’ can be estimated using a numerical method.

[0101] In embodiments, this numerical method may comprise projecting a plurality of vectors or rays out from the simulated evaluation location, each vector or ray passing through a different respective point on a surface of the hemisphere (or other shape) defining the field of view, and then interrogating whether or not each vector or ray, when extended to infinite length, intersects with the model of the target object in the simulated scene. In examples where the model comprises a plurality of cuboids individually labelled as being associated with one of a target class or a background class, this interrogation may comprise computing whether each vector intersects with each one of the plurality of the cuboids using a 3D Delaunay triangulation. Once each of a plurality of vectors passing through the evaluation location into the field of view (i.e. intersecting the surface of a unit hemisphere centred on the evaluation location) has been labelled as either intersecting at least one cuboid of the model or not intersecting any cuboids of the model, the angle of the field of view subtended by the model can be estimated from this information.

[0102] In a first example, a Monte Carlo method is implemented by the field of view estimation control logic, in which a random set of / V vectors is defined, passing from the evaluation location through the surface of a unit hemisphere (simulating the field of view) defined at the evaluation location, wherein the angle of each vector within the field of view is randomised. The set of / V random vectors is classified into a subset Ni which are determined to intersect the model of the target object, and a subset No which are determined not to intersect the model of the target object. The region of the field of view at the evaluation location which is occupied by the target object can then be characterised as the proportion of the total set of N vectors (where N=Ni+N0) which intersects the model of the target object, given in this example by Ni / N. It will be appreciated that this approach is in effect a numerical estimation of the solid angle fraction of a hemispherical field of view which is subtended by the model at the evaluation location.

[0103] The approach described provides an estimate of the region of the field of view at an evaluation location which is estimated to be occupied by a target object, based on modelling the target object and the evaluation location, and numerically computing the occupancy of the field of view at the modelled evaluation location by the modelled target object. This provides an estimate of theP131108GB

[0104] proportion of the field of view at the evaluation location which is occupied by the target object, for one position of the target object relative to the evaluation location. This metric is used as a measure of suppression of background radiation by the target object at the evaluation location. It will also be appreciated that an image produced by a lens (e.g. ‘fish eye’ lens), or lens simulation, providing or simulating an equidistant, stereographic, equisolid angle, or orthographic mapping of a scene onto an imaging sensor, can be used to estimate a region of the field of view at an evaluation location corresponding to the location of the imaging sensor (i.e. the imaging sensor location), which is occupied by the target object, using a variation of the Monte-Carlo approach described herein. It will be appreciated the distance of a given feature in such an image from the centre of the image is proportional to an angle between the imaging direction (i.e. the central vector of the field of view of the imaging system) and the direction from the imaging location to the given feature in the field of view. Thus, a Monte-Carlo approach can be applied to images derived produced using a lens (e.g. ‘fish eye’ lens), or lens simulation, providing or simulating an equidistant, stereographic, equisolid angle, or orthographic mapping of field of view onto an imaging sensor (i.e. an image plane), by randomly selecting a set of / V points in the image, classifying them into a subset Ni which are determined to be coincident with a representation of the target object in the image, and a subset No which are determined not to be coincident with a representation of the target object in the image, and estimating the region of the field of view at the imaging location which is occupied by the target object based on Ni / N.

[0105] In some use cases, estimation of suppression of background radiation by a target object at an evaluation location can assume a static spatial relationship between target object and the evaluation location. However, as described further herein, it is inherent to certain contexts of interest that there is temporal change in the relative spatial positions of evaluation location and target object. This is the case in the example of an RPM described herein, where a target object comprising a vehicle transits over time past one or more evaluation locations coincident with at least one radiation detector panel of the RPM. Thus over a plurality of sequential time points, there exist different spatial relationships between the target object and each of one or more evaluation locations, and background suppression may be estimated for each of these spatial relationships, at each evaluation location, based on the different region of the field of view at each one or more evaluation locations which is occupied by the target object in each instance. This scenario may also arise in other use cases, such as use of mobile (e.g. hand-held or vehicle mounted) radiation monitoring apparatuses, where an evaluation location is coincident with a radiation detector of the apparatus, and movement of the apparatus relative to one or more target objects under investigation leads to dynamic variation in spatial relationship between evaluation location and target object(s) over time.P131108GB

[0106] Accordingly, according to embodiments of the present disclosure, a degree of suppression of background radiation at an evaluation location, by a target object, based on a region of the field of view at the evaluation location which is estimated to be occupied by the target object, may be estimated for each of a plurality of time points, wherein the spatial relationship between the target object and evaluation location is different at each of the plurality of time points. It will be appreciated that these time points, in relation to a moving target object, may also be characterised as different spatial locations of the target object relative to the evaluation location.

[0107] Figure 10a shows a plot of solid angle fraction computed using a Monte Carlo approach as set out herein, wherein the proportion I fraction of a hemispherical viewing angle (i.e. of a field of view) from the evaluation point 210’ of Figure 7 which is not subtended by the target object model 202’ is plotted with respect to linear position of the geometric centre of model 202’ as it is incremented to different positions along the z axis (i.e. to different positions along the path Pzshown in Figure 8). The plot in this example shows the proportion of a 2TT steradian field of view at the evaluation location 210’ which is not occupied by the model, for different linear positions of the centroid of the model of the target object along the target object path Pz(i.e. different modelled transit positions past the evaluation location). To generate such a plot, for each of a plurality of values of the linear position of the model centroid along the path Pz(i.e. simulating different snapshots in the transit of a vehicle 202 past the real-world evaluation location on detector panel 200A of Figures 4a and 4b), a Monte Carlo simulation as described herein is carried out by the field of view determination control logic 270, and the region of the field of view at the evaluation location which is estimated to be occupied by the target object is calculated (i.e. based on determining intersection of each of a plurality of vectors through the field of view with a model of the target object). It will be appreciated that where a velocity v of the model is known, each linear position of the model, indicated on the horizontal axis of the plot of Figure 10a, can be converted into a corresponding time, based on the elapsed time between positions. In other words, a spatial to temporal conversion, or vice versa, can be carried out. Each of Figures 10b to 10g shows a polar plot of intersections of each of a plurality of vectors, generated using the Monte Carlo method, with the surface of a hemisphere used to model the field of view at the evaluation location 210’ (e.g. a hemisphere 250’ as shown in Figure 9b), where each point plotted is associated a vector determined by the field of view evaluation control logic to intersect the target object model 202’. Each of the sequential polar plots of Figures 10b to 10g respectively corresponds to a sequential one of the target object distances 1 to 6 along the transit path Pz, as shown in the plot of Figure 10a. It is seen that where a greater number of vectors intersect the model 202’ (as seen in the polar plots of Figures 10e and 10f) the region of the field of view not subtended by the model 202’ is lower (as seen for corresponding positions 4 and 5 in Figure 10a). Where a lower number of vectors intersect the model 202’ (as seen in the polar plots of Figures 10b and 10g) the region of the field of view not subtended by the model 202’ is higher (as seen for correspondingP131108GB

[0108] positions 1 and 6 in Figure 10a). In an example where a Monte Carlo approach is applied to images derived produced by a lens (e.g. ‘fish eye’ lens), or lens simulation providing or simulating an equidistant, stereographic, equisolid angle, or orthographic mapping of field of view onto an imaging sensor (i.e. an image plane), as described herein, the points plotted in each of Figures 10b to 10g could be considered to represent positions of points in subset / y which are determined to be coincident with a representation of the target object in the image. This approach recognises such an image, being a transform of a 3D field of view onto a 2D plane, is an analogue of a polar plot (as shown, for example, in Figures 10b to 10g) which also represents a 3D field of view (e.g. a hemispherical field of view) on a 2D plane.

[0109] Thus, where it is of interest to assess time-variation in suppression of background radiation by a target object at an evaluation location, as the target object moves relative to the evaluation location, approaches according to the present disclosure may comprise: estimating a region of a field of view at the evaluation location that is occupied by the target object for each of a plurality of relative positions of evaluation location and target object; where for each relative position, the relative spatial locations of the evaluation point and the target object are modelled, and the region of a field of view at the modelled evaluation location which is occupied by the model of the target object is estimated by field of view determination control logic. As further described herein, this computation may be carried out by determining whether each of a plurality of non-coincident paths intersecting the modelled evaluation location intersects with the model of the target object.

[0110] It will be appreciated that in some use contexts, as described in examples herein, each of one or more evaluation locations corresponds to a location in or on a radiation detector of a radiation monitoring apparatus, and the approaches herein can be implemented to obtain an estimate of suppression of background radiation suppression for a first time or time period, which can be used in processing radiation data obtained from the radiation detector at that same time or time period. In such contexts, for a real-world scenario in which radiation data is being obtained at an evaluation location of a detector, and a given spatial relationship between the evaluation location and the detector, it is of significance to be able to register a model of the target object 202’ to a model of the evaluation point 210’, so that this spatial relationship corresponds to that between the real target object 202 and real evaluation point 210. This allows an estimated degree of suppression of background radiation at a modelled evaluation location, by a modelled target object, based on the region of the field of view at the modelled evaluation location which is estimated to be occupied by the modelled target object, to be applied as an estimate of real-world suppression of background radiation for a particular spatial configuration between real-world target object and real-world evaluation location, and for this estimate of real-world suppression of background radiation to be applied in processing radiation data obtained at the real-world evaluation location for this same configuration.

[0111] 1P131108GB

[0112] Thus the model obtaining control logic or field of view evaluation control logic may be configured to register a model of a target object 202’ to a model of one or more evaluation locations 210’, so that the modelled scene corresponds, in the relevant model scale, to a real-world spatial relationship between a real world target object on which the model 202’ is based and at least one real-world evaluation location. In principle, any sensing method which allows position of a target object relative to one or more evaluation locations to be derived can be used, and this approach may make use of certain assumptions. For example, in the exemplary context of RPM 2 in Figures 4a and 4b, the transit path Pzof the target object 202 is constrained to be parallel to a single degree of freedom, parallel to the z axis. The distance of the evaluation location 210 from the closest point on the transit path of the target object 202 is fixed and predefined by the RPM 2 geometry, as is the distance of the evaluation location 202 from a pathway (i.e. the ground) along which target objects transit through the RPM 2. Thus, either of the first and second image acquisition systems (235, 255) or break beam pairs of a radiation monitoring apparatus (e.g. an RPM) can be used to provide information about the position of the target object along the transit path at a specific time t. Alternatively, an ultrasonic or laser proximity detection system can be oriented to align the beam axis substantially parallel to the transit path direction (z), to detect a position of the target object along the transit path at a specific time t. Whether or not assumptions can be made about possible degrees of freedom in which target objects can move relative to the evaluation point, the skilled person will appreciate different approaches are known which can allow this information to be resolved. For example, a plurality of cameras or other sensors positioned in different locations relative to the evaluation point can be used to triangulate the position of target objects in 3D space based on detection of reference points or regions (e.g. edges) of the target object in the resulting images. The specific approach that may be used to characterise the position of a target object relative to one or more evaluation points is not of particular importance. The skilled person will appreciate what is relevant is the use of the resulting positional information for a certain time point, in registering a model of the target object to a model of the evaluation point for estimation of the region of the field of view at the modelled evaluation location which is occupied by the modelled target object at that time point.

[0113] There are a number of variations of aspects of the methods set out herein which are within the scope of the general approach described for estimation of suppression of background radiation. For example, in the RPM example given above, first and second image acquisition systems are described which each use an X-ray imaging source and an X-ray imaging detector positioned on different sides of a transit path along which target objects travel through the RPM. However, it will be appreciated that other imaging modalities can be used which provide information about at least an external shape of a target object. In particular, it will be appreciated that not all imaging modalities require an active source (such as an X-ray source), and not all imaging modalitiesP131108GB

[0114] operate on the basis of transmission of radiation through a target object. Thus, in other embodiments, either of the first and optional second image acquisition system may comprise an optical imaging system, for example. With reference to Figures 4a and 4b, a first image acquisition system 235 may be modified to replace X-ray imaging detector 225 with an optical imaging apparatus (e.g. a digital camera with a suitable lens arrangement), and replace the X-ray source 230 with a light source positioned the same side of the target object as the optical imaging apparatus. It will be appreciated the image acquisition control logic (230, 250) in such an embodiment will be modified to process optical images, but otherwise the approach taken to extract overview images and process these to determine spatial extent of the target object can be the same as described herein in relation to Figures 5a and 5b.

[0115] Alternatively, each of the first and optional second image acquisition system may be implemented as an array of break-beam sensors which provide information about the spatial extent of a target object. For example, an entrance break beam pair comprising beam emitter 203A and beam detector 203B may be duplicated at regular intervals along the transit path direction (i.e. the z axis in Figure 4a) and at regular vertical intervals spanning from the ground to a position above the maximum target object height (i.e. along the y axis in Figure 4b). For a given position of the target object 202 in the RPM 2, the resulting 2D array of beam detectors can be used to infer the spatial extent of the target object at a point in time relative to an evaluation location, based on the relative positions of a subset of the beam detectors which sense occupancy at that time (i.e. sense occluded beam). The resolution of this system can be improved by providing larger numbers of break beam pairs with closer spacing in the zy plane. This alternative embodiment of a first image acquisition system can provide a side view of the target object, and a second image acquisition system rotated through 90 degrees around the transit path through the RPM can provide a top-down view of the target object. It will be appreciated the image acquisition control logic (230, 250) in such an embodiment will be modified to process a 2D grid of points labelled as either associated with the target object or associated with the background, with the spacing of the beam detectors used to scale the resolution of the grid (i.e. image), but otherwise the approach used by the image acquisition control logic to generate overview images (e.g. for use in obtaining a model of the target object) can be the same as described herein in relation to Figures 5a and 5b.

[0116] It will further be appreciated that while first and second image acquisition systems are described as being provided to acquire images with respect to two different imaging orientations, it will be appreciated that n further image acquisition systems can be introduced, providing n further, different imaging orientations. As the skilled person is aware, providing a greater number of imaging orientations for a given target object typically increases the resolution with which the shape of the target object can be reconstructed from the resulting image data. In embodiments, the number of imaging orientations can be increased for a given number of image acquisitionP131108GB

[0117] systems by moving an imaging orientation of a single image acquisition system. Thus, for example, a first image acquisition system 235 as shown in Figures 4a and 4b may be configured to rotate with respect to the transit path of target objects through the RPM 2, such that the image acquisition angle around the transit path is incremented between image acquisitions.

[0118] It will be further appreciated that the Monte Carlo method described herein is just one possible approach for estimating a region of a field of view at the evaluation location which is occupied by a target object based on a geometric relationship between a model of the target object and an evaluation location. Other analogous approaches may be used which take into account existing information about the geometry of the scene, such as ray tracing approaches known to the skilled person, in which a field of view from a viewing position, and representations of modelled objects within said field of view, are simulated using model generation control logic typically comprising one or more GPUs.

[0119] It will also be appreciated that a 2D representation of a target object may be directly used as a model of the target object, based on which a region of a field of view at an evaluation location which is occupied by the target object is estimated. For example, in the example of the RPM 2 shown in Figures 4a and 4b, a 2D overview image of the side of the target object, for example as shown in Figure 5a, may be used to derive a 2D model of the target object comprising the regions in the 2D overview image determined to lie within the outer edges of the representation of the target object in the overview image. This 2D model may be used in approaches described herein whereby estimating a region of a field of view at the evaluation location which is occupied by the target object comprises determining, by field of view determination control logic, whether each of a plurality of non-coincident paths intersecting the evaluation location intersects with the model. In all of the approaches described herein, it will be appreciated that the fraction of the field of view at the evaluation location estimated to be occupied by a target object can be used to derive an estimate of background suppression, due to the proportionality between this fraction and the suppression of background radiation at the evaluation location by the target object. In some examples, the fraction of the field of view at the evaluation location estimated to be occupied by a target object can be directly used as an approximation of the degree of background suppression. In other examples, this relationship may be modified by using weightings, as explained further herein.

[0120] While it has been described herein that a model of the target object may be generated from one or more images of the target object, it will be appreciated such a model may be derived by model generation control logic in alternative ways. For example, a database of models may be maintained based on available data regarding the spatial extent of different potential target objects. For example, where an evaluation location is associated with an RPM, and the targetP131108GB

[0121] objects comprise vehicles of a particular class (e.g. heavy goods vehicles), information from manufacturers of such vehicles may be used to generate 3D models which are stored in a database accessible to the model generation control logic, or the information itself may stored in such a database and used by the model generation control logic to generate a model of a given target object based on this information. Each of a plurality of vehicles can be catalogued in a database in this manner, with an identifier of the vehicle type (e.g. model number) being stored in a predefined association with a model of the vehicle, or information from which such a model can be generated. In this example, obtaining the model of the target object by the model generation control logic comprises identifying the target object, and obtaining from a data storage element a predefined model having a predefined relationship with the target object, based on identifying the target object. The target object can be identified on approach to the evaluation location (e.g. an RPM), for example, using an identifier encoded in a form such as an RFID tag, or barcode, or according to any other approach known to the skilled person, and the relevant model of the target object obtained from the data storage element based on querying a database of models stored on the data storage element using the identifier, to return a model having a predefined relationship with said identifier. It will be appreciated the registration of the model to a specific location relative to the evaluation location, and estimation of background radiation suppression based on the model can then be carried out according to the approaches described herein in respect of image-derived models.

[0122] In some approaches according to embodiments of the present disclosure, estimating the region of the field of view at the evaluation location which is occupied by the target object comprises; obtaining, from at least one imaging sensor, image data from a region of space containing the target object; establishing an indication of a spatial extent of the target object based on the image data; and estimating directly from the indication of the spatial extent a region of the field of view at the evaluation location which is occupied by the target object.

[0123] In accordance with embodiments of the present disclosure in which an imaging sensor is used, the imaging sensor is configured to acquire image data according to a field of view defined by characteristics of the imaging sensor, and optionally defined by characteristics of an image acquisition system in which the imaging sensor is comprised. For example, an imaging sensor may be coupled with one or more optical elements (such as lenses, gratings, and I or collimating elements) to apply a transformation to an incident field of electromagnetic or nuclear radiation. The skilled person is aware there are many ways in which optical elements can be used to map a field of view onto an imaging sensor. For example, a lens arrangement can be coupled to an imaging sensor, the lens arrangement being configured to provide an equidistant, stereographic, equisolid angle, or orthographic mapping onto the at least one imaging sensor of a view of the region of space containing the target object. Such a lens may be referred to as a ‘fish eye’ lens.P131108GB

[0124] Thus incident radiation from a defined field of view can be projected onto the imaging sensor via a mapping function predefined by the optical characteristics of the optical element(s). Additionally, or as an alternative, an algorithm (sometimes referred to as a lens simulation) can be applied to images acquired by an imaging sensor which transforms the images (by translating pixels or other image elements) to approximate the effect of a lens providing, for example, an equidistant, stereographic, equisolid angle, or orthographic mapping onto the at least one imaging sensor of a view of the region of space containing the target object. Thus where a first image acquisition system 230 as shown in Figure 4a is provided (e.g. an X-ray imaging system), producing plane radiographs, these may be transformed by the image acquisition control logic 230 according to a lens simulation to warp the image in a manner which approximates an equidistant, stereographic, equisolid angle, or orthographic projection of the field of view onto the image plane.

[0125] Thus, a first image acquisition system 235 as shown schematically in Figure 4a may be modified to replace X-ray imaging detector 225 with a different type of imaging sensor, such as an optical imaging sensor, and replace the X-ray source 230 with a light source positioned the same side of the target object as the optical imaging apparatus, or remove it entirely. In this example, an imaging sensor can comprise a charge-coupled device (CCD), complementary metal-oxide semiconductor (CMOS), or other photosensitive detector panel, the imaging sensor being coupled to a lens providing an equisolid angle projection onto the imaging sensor of a field of view of substantially 2TT steradians. The purpose of the equisolid angle projection is to map the field of view onto the imaging sensor such that each unit area of the imaging sensor (e.g. one detector pixel), when projected through the lens into the field of view, occupies the same solid angle magnitude. It will be appreciated that other projections, such as an equidistant, stereographic, or orthographic, projection, can provide similar results. Thus, where such mappings are provided (e.g. by a suitable lens, or lens simulation), the proportion of an acquired image in which a representation of a target object is present can be considered to correlate to the proportion of the field of view which is occupied by the target object, and analysis of one or more images acquired in this way can be used to estimate a region of a field of view at the location of the imaging sensor which is occupied by the target object, at the time of image acquisition. The region of the field of view occupied by the target object may be quantified as a solid angle subtended by the target object.

[0126] Typically, an image acquisition system configured in this way is positioned relative to a transit path for target objects such that an overview image of an entire target object can be obtained in a single exposure (i.e. without stitching together of images obtained at sequential time points). With reference to the RPM geometry of Figures 4a and 4b, it is also typically advantageous if the imaging sensor (and lens arrangement if included) are located at the same side of the transit path of target objects as an evaluation location 210, at the same height from the pathway (i.e. in the yP131108GB

[0127] axis), and the imaging orientation is centred on a vector normal to the surface of a radiation detector panel on or in which the evaluation location is defined (i.e. a vector perpendicular to the transit path for target objects). The image acquisition system can be advantageously positioned at a predefined location upstream of the evaluation location 210 along the transit path for target objects, such that a target object passing along the transit path can be imaged by the image acquisition system prior to passing the evaluation location 210.

[0128] Thus, in the exemplary RPM 2 of Figure 4a, the first image acquisition system 235 can be modified as described herein for direct determination from images of a solid angle subtended by a target object 202, as the target object 202 passes through the RPM 2. The target object 202 in such an example is first imaged by the image acquisition system 235 to obtain, by the image acquisition control logic 230, an image of the target object 202. Either as a result of a mapping provided by a lens, or a lens simulation applied to image data, as described herein, the resulting image view is a direct representation of a 2TT steradian field of view at the image sensor location, in which each unit area of the image corresponds to the same solid angle magnitude (in steradians) in the field of view. This allows the proportion of the field of view subtended by the target object to be directly estimated on the basis of classifying the image pixels into target object and background classes (e.g. using thresholding, edge-detection, and I or other computer vision approaches known to the skilled person), and dividing the number of target-object pixels by the total number of pixels. This value can be used directly as an estimate of the degree of suppression of background radiation by the target object for an evaluation point which corresponds to the imaging location. Where the imaging location is different to the evaluation location, the estimate of background radiation suppression derived from the imaging location at a time ti can be assumed to correspond to the estimate of background radiation suppression at the evaluation location at a different time t2, where the translation of the target object between times ti and t2 corresponds to the vector between the imaging location and the evaluation location. The translation of the target object between times ti and t2can be determined using approaches set out herein for tracking of target objects.

[0129] Thus, the method may comprise estimating a degree of suppression of background radiation at an imaging location coincident with an image sensor, by the target object, based on a region of the field of view at the imaging location which is estimated to be occupied by the target object, and then deriving an estimated degree of suppression of background radiation at an evaluation location spatially separated from the imaging location, based on the estimated degree of suppression of background radiation at the imaging location. However, it will also be appreciated the imaging location may correspond to an evaluation location in or on a radiation detector. For example, in embodiments corresponding to Figure 4a, the imaging sensor 225 (which may be, for example, a camera with a ‘fish eye’ lens or other lens providing an equidistant, stereographic,P131108GB

[0130] equisolid angle, or orthographic mapping), may be positioned in front of or within a radiation detector (e.g. in or on detector panel 200B) such that a region of the field of view at the imaging location which is estimated to be occupied by the target object based on an image acquired by the imaging sensor 225 at a time t can be used to estimate a degree of suppression of background radiation at the evaluation location which can be used in processing radiation data obtained by the radiation detector at t or associated time period. This may be advantageous in reducing or eliminating the need to perform spatial and / or temporal registration of the estimate of background suppression to be valid for the evaluation location and / or a radiation data acquisition time respectively.

[0131] A number of optional refinements to the background radiation suppression estimation approaches set out herein are possible, as will now be discussed. In particular, these refinements relate to the use of weighting parameters to separately weight the contribution of different sub-regions of a field of view in estimating the degree of suppression of background radiation at the evaluation point based on a region of the field of view subtended by a target object.

[0132] Thus in any of the embodiments herein, estimating the degree of suppression of background radiation at the evaluation location may comprise: defining a plurality of sub-regions of the field of view at the evaluation location, and applying at least one weighting parameter to weight the contribution of at least one of the plurality of sub-regions when determining of a proportion of the field of view at the evaluation location corresponding to the region occupied by the target object. In a first example, at least one weighting parameter for each of at least one of the plurality of subregions of the field of view is determined based on a location of the sub-region within the field of view. This can comprise determining at least one weighting parameter for each of at least one of the plurality of sub-regions based on an angle between a reference path intersecting the evaluation location, and a second path intersecting both the sub-region and the evaluation location.

[0133] Figures 11a and 11b will be recognised from Figure 9a, and show schematically a cross section through a model of a detector element 205A’. A shielding element 206A’ is modelled positioned behind, and around the sides, of the detector element model 205A’. Figure 11a shows schematically rays of incident radiation illuminating the detector element 205A’ along a direction angled at 01 to the front face of the detector element model 205A’. The cross-sectional area of the detector element model 205A’ which is illuminated directly by the incident radiation is indicated by the hatched region. Figure 11b shows the same arrangement, but with a different incident radiation angle 02. The cross-sectional area of the detector element model 205A’ which is illuminated directly by the incident radiation is again indicated by the hatched region, and is observed to be substantially greater for this specific incident angle. The count rate recorded in aP131108GB

[0134] detector (such as a gamma ray detector described herein) is typically proportional to the fraction of the volume of the detector that is illuminated by the radiation incident upon it. Assuming the detector to have uniform response along its vertical length, and treating it as a rectangular cuboid, the fraction of the volume illuminated at an azimuthal angle 0 (i.e. the angle between the incident radiation direction and the z axis, as measured around the y axis, in the reference scheme of Figure 9a) is illustrated by the hatched area shown in Figures 11a and 11b. Based on a shielded detector geometry as shown in Figure 11a, with a detector element width x and depth b, and a shield extending a distance of a each side of the detector element, and projecting out to align with the front face of the detector element in the depth direction, this fraction can be evaluated as:

[0135] < < <

[0136]

[0137] < >

[0138] Since the proportion of the detector element that is illuminated by incident radiation is proportional to the detector response, it will be appreciated that the illuminated fraction of the detector can be used to weight the response of the detector to radiation incident to the detector for a given azimuthal angle Q.

[0139] further weighting parameter can be derived to account for the variation in detector response based on the polar angle (i.e. the angle between the incident radiation direction and the x axis, as measured around the z axis, in the reference scheme of Figure 9a). This variation is present because of the aspect ratio of the detector (extending further in the vertical (y) direction than either of the depth (x) or width (z) directions indicated in Figures 9a and 9b. The inventors have recognised an assumption of cosine dependence can be used to evaluate the detector response as a function of the polar angle cp, as follows:

[0140] g(<p) = cos <p

[0141] Figure 12a shows a plot of the variation of the weighting parameter fas a function of the azimuthal angle 0, and Figure 12b shows a plot of the variation of the weighting parameter g as a function of the polar angle cp, where the weighting parameters are expressed in terms of a fraction of the radiation detector volume illuminated by radiation incident along the respective angle.

[0142] In certain contexts, the inventors have also recognised that objects within the field of view evaluated from the evaluation location can be emitters of background radiation, and use of a further weighting parameter can be used to account for the non-uniformity of the background radiation field at the evaluation location, resulting from differing distances from the evaluationP131108GB

[0143] location to such objects, along different directions from the evaluation point out into the field of view. For example, in an RPM context, the evaluation location may be in a tunnel comprising concrete which acts as the principal background radiation source (i.e. concrete is an example of a NORM radiation emitter). Thus, in embodiments of the present disclosure, at least one weighting parameter for each one of a plurality of sub-regions in a field of view from an evaluation location is determined based on a distance from the evaluation location to a point of intersection with a non-target object (i.e. a background object), along a path intersecting the sub-region and the evaluation location. The inventors have recognised that each one of a plurality of different sub-regions of a field of view at an evaluation location can be weighted in this way based on an inverse square law, as follows:

[0144] h(r) = 1 / r2,

[0145] where r is the distance from the evaluation location to a point of intersection with a non-target object (i.e. a background object), along a path intersecting both the sub-region and the evaluation location. This weighting parameter may be applied only for distances where r>rmin, where in an RPM context, rm / nmay be set as the distance between L and R detector panels of the RPM as shown schematically in Figure 1 (i.e. the width of the pathway between the opposing detector panels of the RPM system). In the context of the Monte Carlo method described herein, in which a random set of N vectors is defined, each vector passing from an evaluation location through the surface of a unit hemisphere (simulating the field of view) defined at the evaluation location, wherein the angle of each vector within the field of view is randomised; and in which the set of N random vectors is classified into a subset Ni which are determined to intersect the model of the target object, and a subset / Vo which are determined not to intersect the model of the target object; a weighting parameter based on distance to a background object along each given vector in subset No (for example based on 1 / r2) can be used to weight said vector. Thus when a region of the field of view at the evaluation location which is occupied by the target object is, as in an example herein, characterised as the proportion of the total set of N vectors which intersects the model of the target object (i.e. Ni / N), the summation of N (based on N=Ni+ No, where Ni is the number of vectors intersecting the target object, and No is the number of vectors not intersecting the target object) can be computed by weighting each vector in subset / Vo using a vector-specific weighting parameter derived based on the distance along that vector between the evaluation location and the nearest background object.

[0146] It will also be appreciated that a distance to background from an evaluation location along different directions into a field of view could be determined by using a light detection and ranging (LiDAR) imaging system comprised in each of one or more image acquisition systems (e.g. a first image acquisition system 235 and / or second image acquisition system 255 shown schematically inP131108GB

[0147] Figures 4a and 4b). For example, a LiDAR system (e.g. a commercially available system such as the Realsense™ range of LiDAR position sensing systems manufactured by Intel™) may provide overview images (as shown schematically in Figures 5a and 5b) corresponding to the field of view of interest, optionally including one or more target objects, where the pixel intensity (i.e. greyscale value) of each pixel in each overview image is proportional to linear distance from the imaging location to an object at a location in the scene corresponding to the pixel location. Thus in a model generated from such overview images, a spatial model of background surfaces relative to the evaluation location can be derived from the images, using LiDAR data analysis approaches known to the skilled person.

[0148] In embodiments comprising direct determination (i.e. without 3D model reconstruction) of a region of a field of view at an evaluation location which is estimated to be occupied by a target object, one or more LiDAR images (comprising image data), obtained from an imaging location as described herein, may be generated using a lens (e.g. ‘fish eye’ lens), or lens simulation providing or simulating an equidistant, stereographic, equisolid angle, or orthographic mapping of the imaged scene onto an imaging sensor, and such images can be used to quantify a distance to background for each of one or more regions in each image which do not correspond to a representation of a target object in the image.

[0149] Win embodiments where an imaging sensor is coupled to a lens arrangement configured to provide an equidistant, stereographic, equisolid angle, or orthographic mapping onto an imaging sensor of a view of a region of space containing the target object to obtain an image; or where at least one image obtained by an image sensor is processed by applying an equidistant, stereographic, equisolid angle, or orthographic transformation to the at least one image; the image sensor may therefore be a LiDAR image sensor, and the images generated by the image sensor comprise LiDAR images. Approaches set out herein can then be applied to determine a spatial extent of a representation of a target object in one or more such images, based for example on detecting the outer edge of a representation of a target object in each image and quantifying the spatial extent of the enclosed region. The degree of suppression of background radiation at the evaluation location can then be estimated based on a proportion of the field of view at the evaluation location that corresponds to the region occupied by the target object, based for example on dividing an area of the region corresponding to the target object representation in the image (Ao), by an area of the image (AT), where AT = AO+AB, where AB is the area of a background (i.e. non-target-object) region of the image. The contribution of each sub-unit of area (e.g. each pixel) of the background region, to the summation of AB in this calculation, can be individually weighted based on a distance associated with said sub-unit of area, quantified based on the LiDAR image (i.e. using the intensity of the pixel, or mean intensity of a unit area comprising a plurality of pixels).P131108GB

[0150] In embodiments of the present disclosure, the radiation monitoring control logic 290 may also be configured to determine at least one weighting parameter ( / ) used to weight the contribution of at least one of the plurality of sub-regions of the target object in determining a proportion of the field of view at the evaluation location that corresponds to a region occupied by the target object, where determining the weighting parameter comprises estimating a radio-density associated with a portion of the target object corresponding to at least one of the plurality of sub-regions. A radiodensity of a portion of the target object may be derived, for example, based on radiographs of the target object, such as X-ray radiograph overview images. Where a model of the target object is, for example, reconstructed based on X-ray radiographs obtained from different imaging orientations (e.g. via tomographic reconstruction), the skilled person will appreciate sub-regions of such a model can be parameterised with a corresponding radio-density parameter (e.g. a reconstructed volume may comprise voxels, each having a greyscale value which is proportional to the radio-density of the imaged object at the corresponding location). This information can be used when estimating a region of a field of view at an evaluation location which is occupied by a target object by determining whether each of a plurality of non-coincident paths intersecting the evaluation location intersects with such a model, by weighting the contribution of each intersection path to the overall estimation of the field of view occupied by the target object, based on a summing radio-density values through the target object along the intersection path direction. Alternatively, where a model is obtained from a database, as described herein, such a model may contain radio-density information based on experimentally determined or otherwise estimated radio-density values for different regions of the model.

[0151] Thus a plurality of exemplary weighting parameters have been described which can optionally be used in certain contexts (e.g. an RPM context) to weight the contribution of different sub-regions of a field of view in estimating the degree of suppression of background radiation at the evaluation point based on a region of the field of view subtended by a target object. These can be applied by the field of view determination control logic when estimating the field of view at an evaluation location which is estimated to be occupied by a target object, for example, using the Monte Carlo approach, ray tracing, or direct determination approaches set out herein. In model-based approaches, for each of a plurality of sub-regions of the field of view at evaluation point (e.g. the 2TT steradian field of view 250’ at an evaluation point 210’ shown schematically in Figures 9a and 9b), one or more of the weighting parameters f, g, h, and / , set out above can be computed for each of the plurality of vectors extended into a modelled scene. The parameters Q, cp, r, a, b, and x, are obtained from the geometry of the model (where r, for instance, is determined by modelling background radiation emission sources such as concrete walls, and extending vectors or rays into the field of view to determine the distance between the evaluation location and the intersection of each vector or ray with a closest surface of an intersecting model of a background radiation emissions source). Thus for each vector, determined either to intersect the model of theP131108GB

[0152] target object or not intersect the model of the target object, one or more of the weighting parameters f, g, h, and i, can be associated with the vector. Where, as described herein, the ratio of the number of model-intersecting vectors ( / y) to total vectors ( / V) is used as an estimate of background radiation suppression, this ratio can be computed by weighting each vector during summation to derive each of Ni and N.

[0153] Figures 13a to 13d show the outcome of sequentially applying the described weighting parameters f, g, and h, when modelling background suppression by a target object at an evaluation point, using the Monte Carlo approach described herein. In each of Figures 13a to 13d, the dotted line shows an experimentally derived intensity profile of background radiation as measured at a gamma radiation detector, with respect to different distances of a vehicle comprising a truck carrying a container, past the gamma radiation detector (i.e. over one ‘transit’ of a vehicle through an RPM containing the gamma radiation detector). Accordingly, each dotted line shows a radiation intensity measured at a gamma radiation detector positioned as per the radiation detector 100A shown in Figures 2a to 2e, as a vehicle moves an increasing distance along a transit path past the radiation detector 100A (as shown schematically for object 102 in Figures 2a to 2e), where the transit path is oriented perpendicular to the principal radiation detection orientation of the detector. The dotted lines show a profile of increasing and then decreasing suppression of background radiation by a vehicle, at a gamma radiation detector location, as the vehicle moves an increasing distance through an RPM in which the gamma radiation detector is located.

[0154] The solid line in each of Figures 13a to 13d shows an estimated background suppression, determined based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object, using a model-based solid-angle estimation approach set out herein. Figure 13a shows a profile of estimated background radiation suppression without any weighting parameters applied (i.e. using only the region of the field of view at the evaluation location which is estimated to be occupied by the target object as a direct proxy for background suppression). Figure 13b shows a profile of estimated background radiation suppression where weighting parameters f and g have been applied to account for the dependency of the gamma radiation detector response on the azimuthal and polar angles of incident radiation. Figure 13c shows a profile of estimated background radiation suppression where weighting parameter h has been applied to account for the dependency of the background radiation intensity on distance to background object(s). Figure 13d shows a profile of estimated background radiation suppression where weighting parameters f, g, and h, have all been applied. It is thus observed that the estimate of background suppression can be improved by adding weighting parameters as described herein, with additional weighting parameters leading to improved fit between the estimated suppression profile and the actual background suppression profile.P131108GB

[0155] Figure 14 schematically shows radiation monitoring control logic 290 which can be used to perform aspects of methods according to embodiments of the present disclosure. The radiation monitoring control logic 290 shown in Figure 14 is described here in the context of a configuration for use with an RPM 2 as shown in Figures 4a and 4b, but it will be appreciated it can be used with appropriate modifications in the context of other radiation monitoring apparatuses as described further herein. The radiation monitoring control logic 290 comprises a communication interface 2901 , an MCA I analyser 2902, a classifier 2903, image acquisition module 2904, model generation module 2905, field of view determination module 2906, output data generator module 2907, user output driver module 2908, controller 2909, and a storage medium 2910..

[0156] The communication interface 2901 is configured to send signals to and I or receive signals from detector(s) of a (nuclear) radiation monitoring apparatus with which the radiation monitoring control logic 290 is configured for use (and in particular, for receiving radiation data from the radiation detector(s). The communication interface 2901 is also configured to receive signals output by one or more beam-break pairs, or alternative user input device for controlling the triggering and halting of detection by each radiation detector. Where the radiation monitoring control logic 290 comprises MCA I analyser control logic 2902, the communication interface 2901 is typically configured to receive radiation data in the form of raw sensor output data (e.g. voltage pulses) from the radiation detector(s). Where the MCA I analyser control logic is integrated into a detector of a monitoring apparatus (e.g. as part of detector-integrated front-end signal processing circuitry), or implemented as one or more standalone modules (e.g. standalone front-end signal processing control logic), disposed between a radiation detector and radiation monitoring control logic 290, the communication interface 2901 is configured to receive part-processed radiation data, for example in the form of detection event count rates for different absolute or relative ranges of energy.

[0157] In embodiments where they are included, the optional output data generator 2907 and user output driver module 2908 are configured to operate cooperatively to cause information be transmitted to a user output device 291. A storage medium 2910 (e.g. in the form of a hard disk drive, solid state drive, tape drive or the like) is for storage of any data received and I or generated by submodules of the radiation monitoring control logic 290. It will be appreciated that, rather than the radiation monitoring control logic 290 comprising the storage medium 2910, the storage medium 2910 may be located in a separate apparatus (e.g. cloud server or local server) accessible to the radiation monitoring control logic 290 over a network or the like (e.g. via the communication interface 2901).

[0158] A classifier module (control logic) 2903 can be provided, being configured to receive radiation data processed by the MCA / analyser module 2902, and identify, using spectral analysisP131108GB

[0159] approaches known in the art, one or more radiation sources based on the characteristics of the radiation data.

[0160] An image acquisition module (control logic) 2904, where included, is configured to control one or more image acquisition systems to carry out imaging of a target object as described herein (e.g. as described in respect of image acquisition control logic 2301250 herein). The image acquisition module 2904 can transmit and receive data over the communication interface 2901 to trigger image acquisition, and receive images, which can be stored in the storage medium 2910.

[0161] A model generation module (control logic) 2905 is configured to obtain a model of a target object as described herein (e.g. as described in respect of model generation control logic 260 herein). The model generation module 2905 can transmit and receive data over the communication interface 2901 to obtain a model from an external database, and I or can obtain a model based on images generated by one or more image acquisition systems and stored in the storage medium 2910 by the image acquisition module 2904.

[0162] The field of view determination module (control logic) 2906 is configured to estimate a degree of suppression of background radiation at an evaluation location, by a target object, based on a region of a field of view at the evaluation location which is estimated to be occupied by the target object (e.g. as described in respect of field of view determination control logic 270 herein). The field of view determination module 2906 can perform this estimation based on one or more models of a target object (and optionally other objects, such as elements of an RPM or other nuclear radiation monitoring apparatus) generated by the model generation module 2905, and / or directly based on images generated by one or more image acquisition systems and stored in the storage medium 2910 by the image acquisition module 2904, using approaches as described further herein.

[0163] The controller 2909 is configured to control the operation of each of the other modules of the radiation monitoring control logic 290. The controller 2909 may also control, via the communication interface 2901, the operation of one or more radiation detectors, one or more break beam pairs, and one or more image acquisition systems, as described herein. Each of the sub-modules of the radiation monitoring control logic 290 may be implemented s hardware control logic (such as by one or more ASIC or FPGA modules), and I or as software control logic implemented on one or more general purpose computing devices. It will also be appreciated that sub-sets of the sub-modules 2901 to 2910 shown in Figure 14 may be distributed between different computing devices implementing their functionality via hardware and I or software control logic. Any or all of the sub-modules of the radiation monitoring control logic 290 may be implemented via cloud-computing resources.P131108GB

[0164] In the context of Figures 4a and 4b, in which the nuclear radiation monitoring apparatus with which the radiation monitoring control logic 290 is associated is an RPM 2, the collection of radiation data may be carried out as followed. When a vehicle, or other target object 202, enters the RPM 2, the beam detector 103B provides a beam-break signal to the controller 2909 of the radiation monitoring control logic 290. In response, the controller 2909 controls each of radiation detector panels 100A, 100B, 100C, and 100D, to start associating radiation data recorded by the radiation detector with the target object in the RPM. Radiation data is recorded by each radiation detector during each of a plurality of successive time intervals (referred to herein as capture intervals) as the target object 202 travels through the RPM 2, and is provided to the radiation monitoring control logic 290 via communication interface 2901. In non-RPM contexts, such as where the radiation monitoring control logic 290 is implemented with a mobile (nuclear) radiation monitoring apparatus, a suitable user interface is used to provide the triggering signals to the controller 2909 to cause the controller to control the detector of the mobile radiation monitoring apparatus to start associating radiation data recorded by one or more radiation detectors, and provide radiation data recorded by the detector(s) during each of a plurality of successive capture intervals to the radiation monitoring control logic 290 via communication interface 2901.

[0165] The analyser module 2902 receives the radiation data associated with a target object (e.g. a vehicle and I or cargo) captured by one or more radiation detectors during each of a plurality of capture intervals. In an embodiment, the radiation data from a plurality of adjacent time intervals is aggregated to improve the spectral classification performance. In one example, the radiation data of successive sets of plural adjacent capture intervals is aggregated. For neutron data captured by any neutron detectors during each capture interval, the analyser module 2902 may likewise aggregate the neutron count detected by each neutron detector in order to determine a total neutron count for that capture interval.

[0166] For gamma-ray data captured by any gamma-ray detectors during each capture interval, the analyser module 2902 may combine the gamma-ray spectra detected from each gamma-ray detector in order to generate a higher intensity, combined gamma-ray spectrum (the data representing the combined gamma-ray spectrum is referred to as combined gamma-ray data). This allows the captured gamma radiation to be analysed even for a low gamma-ray source intensity. In order to maintain a combined gamma-ray spectrum of good quality for use in the subsequent signal processing stages, each gamma-ray detector is identical and is continuously stabilised and calibrated in order to avoid the impact of changing environmental conditions (e.g. temperature fluctuations). The combining of gamma-ray data from multiple detectors is known in the art (see e.g. [1] and [2]) and is therefore not described in detail here.

[0167] The classifier module 2903 is configured to process data received from the MCA I analyser module 2902 to identify isotopes of interest in a target object under investigation by a nuclearP131108GB

[0168] radiation monitoring apparatus, using approaches known to the skilled person, such as those described in [4] or [5], Estimates of background radiation suppression determined according to the present disclosure can be used by the classifier module 2903 in processing of data obtained from one or more radiation detector(s) corresponding to one or more evaluation locations for which background radiation suppression has been estimated.

[0169] As described further herein, though an evaluation location at which suppression of background radiation is estimated may correspond to any real-world location, it may be particularly advantageous to implement the background radiation suppression estimation methods herein in contexts where the evaluation location corresponds to a location in or on a radiation detector of a radiation monitoring apparatus. This is because, as described in relation to Figure 3, variations in background suppression by a target object can be a particular issue when radiation data is being obtained relating to said target object. In particular, where an RPM or other radiation monitoring apparatus is configured to provide an alarm I alert to a user if an elevated radiation count rate is detected (e.g. by an MCA I analyser module 2902, or by a classifier module 2903, based on radiation data processed by such a module of radiation monitoring control logic 290), a measure of background radiation suppression as estimated herein can be used to adjust a threshold used to determine if a radiation count rate value should trigger an alarm I alert. For example, such a threshold can be reduced where the estimated degree of background radiation suppression is higher, to compensate for the drop in total radiation count caused by the greater degree of radiation shielding by the target object. Conversely, such a threshold can be increased where the estimated degree of background radiation suppression is lower, to compensate for the increase in total radiation count caused by the lower degree of radiation shielding by the target object. It will be appreciated the skilled person can define the specific relationship between the degree of estimated background radiation suppression and the alert threshold, based on a particular use context, via modelling and I or experimentation.

[0170] In any method as described herein, each of a plurality of detector elements can be associated with a single evaluation location (e.g. assumed as the centroid of the detector element surface or volume), such that where a radiation monitoring apparatus comprises a plurality of detector elements, the radiation monitoring control logic can compute for each of a plurality of corresponding evaluation points an estimate of background suppression using approaches set out herein. Thus detector output (i.e. radiation data) from each detector element can be associated with a specific estimate of background suppression at an evaluation point associated with the detector element position, and processing of the radiation data from each one of a plurality of detector elements can take into account this specific estimate of background suppression. Thus, for example, where a plurality of detector elements comprise detector pixels on a flat-panel detector, each detector pixel can be associated with one of a plurality of evaluationP131108GB

[0171] points, or where a plurality of detector elements comprise volumetric detectors (e.g. a plurality of plastic scintillator detectors), each volumetric detector can be associated with one of a plurality of evaluation points. It will be appreciated that increasing the number of evaluation points allows more accurate characterisation of estimated background suppression with respect to different locations in a system (e.g. a radiation detection apparatus). The approaches described herein may have particular utility where the evaluation location corresponds to a location in or on a radiation detector with a viewing angle (i.e. an angle over which radiation is received) which is oriented towards the ground (i.e. towards a surface over which objects such as vehicles travel through the RPM). For example, such a radiation detector may be mounted above a transit path along which target objects pass for analysis by a radiation monitoring apparatus in which the radiation detector is comprised, the radiation detector being positioned to detect radiation emitted upwards (i.e. in a direction away from the ground) from target objects passing along the transit path under the radiation detector location.

[0172] Thus, any method in accordance with embodiments of the present disclosure may comprise obtaining radiation data from a (nuclear) radiation detector coincident with an evaluation location, and processing radiation data obtained over a first period of time based on an estimated degree of suppression of background radiation at the evaluation location, wherein an occupation of the region of the field of view at the evaluation location by the target object, used to estimate the degree of suppression of background radiation, is the estimated occupation of the region of the field of view at the evaluation location by the target object during the same first period of time. Optionally, the method comprises adjusting, based on the estimated degree of suppression of background radiation, a parameter used in estimating, based on the radiation data, a likelihood a radiation source is present in the target object.

[0173] In so far as embodiments of the disclosure have been described as being implemented, at least in part, by software-controlled data processing apparatus, it will be appreciated that a non-transitory machine-readable medium carrying such software, such as an optical disk, a magnetic disk, semiconductor memory or the like, is also considered to represent an embodiment of the present disclosure.

[0174] It will be appreciated that the above description for clarity has described embodiments with reference to different functional units, circuitry and / or processors. However, it will be apparent that any suitable distribution of functionality between different functional units, circuitry and / or processors may be used without detracting from the embodiments.

[0175] Described embodiments may be implemented in any suitable form including hardware, software, firmware or any combination of these. Described embodiments may optionally be implemented at least partly as computer software running on one or more data processors and / or digital signalP131108GB

[0176] processors. The elements and components of any embodiment may be physically, functionally and logically implemented in any suitable way. Indeed, the functionality may be implemented in a single unit, in a plurality of units or as part of other functional units. As such, the disclosed embodiments may be implemented in a single unit or may be physically and functionally distributed between different units, circuitry and / or processors.

[0177] Although the present disclosure has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Additionally, although a feature may appear to be described in connection with particular embodiments, one skilled in the art would recognize that various features of the described embodiments may be combined in any manner suitable to implement the technique.

[0178] Figure 15 is flow chart schematically showing steps of a method according to the present disclosure, to be implemented by control logic of a radiation monitoring apparatus, wherein the method comprises a step S1 of estimating a region of a field of view at an evaluation location which is occupied by a target object, and a step S2 of estimating a degree of suppression of background radiation at the evaluation location, by the target object, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object.

[0179] REFERENCES

[0180] [1] GB patent GB 2463707

[0181] [2] GB patent GB 2504771

[0182] [3] EP patent EP3715914

[0183] [4] EP patent EP3637150

[0184] [5] GB patent GB 2445578

Claims

P131108GBCLAIMS1. A method of estimating suppression of background radiation, by a target object, at an evaluation location, the method comprising:estimating a region of a field of view at the evaluation location which is occupied by the target object; andestimating a degree of suppression of background radiation at the evaluation location, by the target object, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object.

2. The method of claim 1 , wherein the degree of suppression of background radiation at the evaluation location is estimated based on a proportion of the field of view at the evaluation location that corresponds to the region occupied by the target object.

3. The method of any preceding claim, wherein estimating the region of the field of view at the evaluation location which is occupied by the target object comprises;obtaining, from at least one sensor, data indicative of one or more regions of space in which the target object is present;establishing an indication of a spatial extent of the target object based on the data; and estimating from the indication of the spatial extent a region of the field of view at the evaluation location which is occupied by the target object.

4. The method of claim 3, wherein the location of each at least one sensor is distinct from the evaluation location.

5. The method of any of claims 3 to 4, wherein the at least one sensor is at least one image sensor, the data is image data, and establishing the indication of the spatial extent of the target object comprises analysing at least one image comprised in the image data to establish a spatial extent of a representation of the target object in the at least one image.

6. The method of claim 5, wherein the spatial extent of the representation of the target object in the at least one image is detected based on classifying a plurality of discrete image elements of the at least one image based on whether or not they correspond to the representation of the target object in the at least one image.

7. The method of any of claims 5 to 6, wherein the spatial extent of the representation of the target object in the at least one image is detected based on detecting the outer edge of the representation of the target object in the image, and quantifying the spatial extent of the enclosed region.

8. The method of any of claims 5 to 7, wherein the at least one image sensor is coupled to a lens arrangement configured to provide an equidistant, stereographic, equisolid angle, or orthographic mapping onto the at least one sensor of a view of a region of space containing the target object; or wherein the at least one image is processed by applying an equidistant, stereographic, equisolid angle, or orthographic transformation to the at least one image.P131108GB9. The method of any preceding claim, wherein estimating the region of the field of view at the evaluation location which is occupied by the target object comprises determining an angle subtended at the evaluation location by the target object.

10. The method of claim 9, wherein the subtended angle comprises a solid angle.

11. The method of any preceding claim, wherein estimating the region of the field of view at the evaluation location which is occupied by the target object comprises:obtaining a model characterising a spatial extent of the target object with respect to the evaluation location; andestimating a region of a field of view at the evaluation location which is occupied by the target object based on at least one geometric relationship between the model and the evaluation location.

12. The method of claim 11, wherein the estimating the region of a field of view at the evaluation location which is occupied by the target object comprises determining whether each of a plurality of non-coincident paths intersecting the evaluation location intersects with the model.

13. The method of any of claims 11 to 12, wherein the spatial extent is three dimensional.

14. The method of any of claims 11 to 13, wherein obtaining the model comprises identifying the target object, and obtaining from a data storage element a predefined model having a predefined relationship with the target object, based on identifying the target object.

15. The method of claim any of claims 11 to 14, wherein the model is generated based on image data representing the target object.

16. The method of claim 15, wherein the at least one image comprises a plurality of images of the target object obtained by at least one imaging sensor along non-parallel imaging orientations.

17. The method of claim 16, wherein the model is generated from the plurality of images via tomographic reconstruction.

18. The method of any preceding claim, wherein estimating the degree of suppression of background radiation at the evaluation location comprises:defining a plurality of sub-regions of the field of view at the evaluation location, and applying at least one weighting parameter to weight the contribution of at least one of the plurality of sub-regions in determining a proportion of the field of view at the evaluation location that corresponds to a region occupied by the target object.

19. The method of claim 18, wherein at least one weighting parameter for each of at least one of the plurality of sub-regions is determined by estimating a radio-density associated with a portion of the target object corresponding to the at least one sub-region.

20. The method of any of claims 18 to 19, wherein at least one weighting parameter for each of at least one of the plurality of sub-regions is determined based on a location of the sub-region within the field of view.P131108GB21. The method of claim 20, wherein at least one weighting parameter for each of at least one of the plurality of sub-regions is determined based on a viewing angle, in the field of view at the evaluation location, of the sub-region.

22. The method of any of claims 20 to 21 , wherein at least one weighting parameter for each of at least one of the plurality of sub-regions is determined based on a distance from the evaluation location to a point of intersection with a non-target object along a path intersecting the evaluation location, and the sub-region location in the field of view at the evaluation location.

23. The method of claim 22, wherein the distance from the evaluation location to a point of intersection with a non-target object, along the path intersecting the evaluation location and the sub-region location in the field of view at the evaluation location, is determined by obtaining one or more light depth and ranging, LiDAR, images from a LiDAR imaging system configured to obtain LiDAR images of the non-target object; and deriving from the one or more LiDAR images the distance to the point of intersection with the non-target object.

24. The method of any preceding claim, the method comprising:determining at least one further evaluation location;estimating a region of a field of view at each further evaluation location which is occupied by the target object; andestimating a degree of suppression of background radiation at each further evaluation location, by the target object, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object.

25. The method of any preceding claim, the method comprising;estimating a second, different region of a field of view at the evaluation location which is occupied by the target object, andestimating a second degree of suppression of background radiation at the evaluation location, by the target object, based on the second, different region of the field of view at the evaluation location which is estimated to be occupied by the target object.

26. The method of any preceding claim, wherein each of one or more evaluation locations corresponds to a location in or on a radiation detector of a radiation monitoring apparatus.

27. The method of claim 26, further comprising obtaining radiation data from the radiation detector, and processing radiation data obtained over a first period of time based on the estimated degree of suppression of background radiation, wherein the occupation of the region of the field of view at the evaluation location by the target object used to estimate the degree of suppression of background radiation is the estimated occupation of the region of the field of view at the evaluation location by the target object during the same first period of time.

28. The method of claim 27, wherein the processing comprises adjusting, based on the estimated degree of suppression of background radiation, a parameter used in estimating based on the radiation data a likelihood a radiation source is present in the target object.

29. Circuitry implementing control logic for estimating suppression of background radiation, by a target object, at an evaluation location, wherein the control logic is configured to:P131108GBestimate a region of a field of view at the evaluation location which is occupied by the target object; andestimate a degree of suppression of background radiation at the evaluation location, by the target object, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object.

30. Circuitry comprising control logic configured to implement the method of any of claims 1 to 28.

31. A non-transitory computer program product configured to control a computer to perform the method of any of claims 1 to 28.

32. A recording medium storing a non-transitory computer program product according to claim 32.

33. Circuitry means implementing control logic means for estimating suppression of background radiation, by target object means, at an evaluation location, wherein the control logic is configured to:estimate a region of a field of view at the evaluation location which is occupied by the target object means; andestimate a degree of suppression of background radiation at the evaluation location, by the target object means, based on the region of the field of view at the evaluation location which is estimated to be occupied by the target object means.

Images