System and method for loss prevention in physical retail

Abstract

Systems and methods for loss prevention in a physical retail space are provided. In an embodiment, a method includes: generating plurality of registries as a customer moves about in the physical retail, each registry comprising a template characterizing the customer and a list of detected items; matching a first one of the plurality of registries with a second one of the plurality of registries based at least in part on the template; comparing the lists of items in the matched registries; and causing at least one downstream loss prevention action to be taken based on a comparison between the lists of items. A corresponding system and non-transitory computer-readable medium are also provided.

Description

SYSTEM AN D METHOD FOR LOSS PREVENTION IN PHYSICAL RETAI LTECH NICAL FIELD

[0001] The present disclosure generally relates to loss prevention in physical retail, and more specifically to systems and methods for detecting potential tampering with theft deterrence devices in unmonitored spaces.BACKGROUND

[0002] Many Large-Scale Retailers utilize Radio Frequency Identification (RFID) tagging on items to allow for streamlined inventory management and facilitated self-checkout payment. The self-checkout systems use RFID scanning to automatically account for all items that are to be purchased by the customer. Often there is a bin in which the items are placed, and the machine automatically scans the items in the bin and lists them for the customer.

[0003] Changing room theft accounts for a large percentage of all apparel retail theft. There are no cameras or surveillance in the changing areas, so thieves can go there to tamper with RFID or other theft deterrence devices without being seen. Similarly, small items within the store may be quickly placed in shielded bags or garments anywhere in the physical store, effectively removing any possibility of tracking that item using standard RFID reader systems.

[0004] Retailers also have RFID sensors at the store entrances that will detect if unpaid items are being carried out the door. However, if the tags are removed from the item, or if RF shielding techniques are used to block scanner functionality, there is no way to detect and report the theft.

[0005] Moreover, when customers use the self-checkout system, only the items that still have tags on them will be accounted for. This allows thieves to pay for only certain items and not others, all the while executing a valid transaction at the point of sale. The thief can then exit the store without drawing any attention to themselves, and there is no way to know that items have been stolen.SUMMARY

[0006] According to an aspect, a method for loss prevention in a physical retail space is provided. The method includes: generating plurality of registries as a customer passes through at least one evaluation zone in the physical retail space, each registry comprising a template characterizing the customer and a list of detected items; matching a first one of the plurality of registries with a second one of the plurality of registries based at least in part on the template; comparing the lists of items in the matched registries and causing at least one downstream loss prevention action to be taken if there is a mismatch between the lists of items.

[0007] According to an aspect, a system for loss prevention in a physical retail space is provided. The system includes: at least one video camera monitoring at least one evaluation zone; at least one RFID reader positioned in the at least one evaluation zone; and a computing system in operative communication with the at least one video camera and at least one RFID reader. The computing system is configured to generate a plurality of registries as a customer passes through the at least one evaluation zone, each registry comprising: a template characterizing the customer, the template being generated using video from the at least one video camera; and a list of items associated with the customer, the list of items being generated using readings from the at least one RFID reader. The computing system is further configured to: match a first one of the plurality of registries with a second one of the plurality of registries based at least in part on the template; and compare the lists of items in the matched registries and cause at least one downstream loss prevention action to be taken if there is a mismatch between the lists of items.

[0008] According to an aspect, a non-transitory computer-readable medium is provided. The non-transitory computer-readable medium has instructions stored thereon which, when executed by at least one processor of a computing system, cause the computing system to: generate plurality of registries as a customer passes through at least one evaluation zone in the physical retail space, each registry comprising a template characterizing the customer and a list of detected items; match a first one of the plurality of registries with a second one of the plurality of registries based at least in part on thetemplate; and compare the lists of items in the matched registries and cause at least one downstream loss prevention action to be taken if there is a mismatch between the lists of items.

[0009] According to an aspect, a method for loss prevention in a physical retail space is provided. The method includes: regularly scanning the retail space using RFID scanners to detect RFID tags associated with items in the retail space; monitoring the retail space using cameras to detect customers interacting with one or more of the items in the retail space; determining that detection of an RFID tag has ceased during the regular scanning of the retail space; identifying a correlation between the ceased detection of the RFID tag and an interaction of a customer with one or more of the items in the retail space; and causing at least one loss prevent action to be taken in response to identifying the correlation.

[0010] According to some aspects, a system and a non-transitory computer-readable medium are provided for implementing the method describe above.BRIEF DESCRI PTION OF DRAWI NGS

[0011] Figure 1 is a schematic of a system for loss prevention in a physical retail space, according to an embodiment.

[0012] Figure 2 is a schematic illustrating profile generation points in a physical retail space, according to an embodiment.

[0013] Figure 3 is a flowchart illustrating a method for loss prevention in a physical retail space using the system of Figure 1.DETAILED DESCRIPTION

[0014] It will be appreciated that, for simplicity and clarity of illustration, where considered appropriate, reference numerals may be repeated among the figures to indicate corresponding or analogous elements or steps. In addition, numerous specific details are set forth in order to provide a thorough understanding of the exemplary embodiments described herein. However, it will be understood by those of ordinary skill in the art, that the embodiments described herein may be practiced without thesespecific details. In other instances, well-known methods, procedures and components have not been described in detail so as not to obscure the embodiments described herein. Furthermore, this description is not to be considered as limiting the scope of the embodiments described herein in any way but rather as merely describing the implementation of the various embodiments described herein.

[0015] As should be appreciated, various embodiments described herein may also be implemented as methods, apparatus, systems, computing devices, computing entities, and / or the like. As such, embodiments may take the form of an apparatus, system, computing device, computing entity, and / or the like executing instructions stored on a computer-readable storage medium to perform certain steps or operations. However, embodiments of the present invention may also take the form of an entirely hardware embodiment performing certain steps or operations. Such devices can each comprise at least one processor, a data storage system (including volatile and non-volatile memory and / or storage elements). For example, and without limitation, the programmable computer may be a programmable logic unit, a mainframe computer, server, personal computer, cloud-based program or system, laptop, etc.

[0016] Embodiments are described below with reference to block diagrams and flowchart illustrations. Thus, it should be understood that each block of the block diagrams and flowchart illustrations, respectively, may be implemented in the form of a computer program product, an entirely hardware embodiment, a combination of hardware and computer program products, and / or apparatus, systems, computing devices, computing entities, and / or the like carrying out instructions on a computer- readable storage medium for execution. Such embodiments can produce specifically configured machines performing the steps or operations specified in the block diagrams and flowchart illustrations. Accordingly, the block diagrams and flowchart illustrations support various combinations of embodiments for performing the specified steps or operations.

[0017] With reference to Figure 1, a system 100 for loss prevention is shown according to an embodiment. In the illustrated embodiment, the system 100 is configured tooperate in a physical retail space 1, such as a clothing store. The retail space 1 includes a plurality of zones or regions, such as a shopping space, a changing room 3, an evaluation zone 5, a checkout area 7, and an exit 9. Although particular zones are discussed herein in connection with retail space 1, it is appreciated that other retail spaces can include different zones and / or different arrangements thereof.

[0018] At least some of the zones in the retail space 1 can be monitored. For example, in the present embodiment, the evaluation zone 5, checkout area 7, and exit 9 are monitored via video surveillance, such as via video cameras 103. More specifically, a first camera 103a or set of cameras is positioned to monitor evaluation zone 5, a second camera 103b or set of cameras is positioned to monitor checkout area 7, and a third camera 103c or set of cameras is positioned to monitor exit 9. The evaluation zone 5 and exit 9 are further monitored via RFID scanners 105. More specifically, a first RFID scanner 105a or set of scanners is positioned to monitor ingress and egress through the evaluation zone 5, and a second RFID scanner 105b or set of scanners is positioned to monitor ingress and egress through the exit 9. Although not specifically illustrated, it is appreciated that the retail space 1 can include other zones that are monitored. In some embodiments, video surveillance and / or RFID scanning can be implemented throughout the shopping space. Such video surveillance can allow tracking movements and / or identifying actions of customers throughout the shopping space and / or monitored areas therein, such as identifying when customers are proximate to or interact with retail items in the retail space. Such RFID scanning can allow implementing inventory monitoring whereby RFID tags associated with retail items in the retail space are regularly scanned to maintain an inventory of all retail items currently in the retail space, and detect when an RFID tag associated with a retail item is unexpectedly no longer detected, indicating that an RFID tag has potentially been destroyed or shielded. Moreover, it is appreciated that some zones can be monitored using a different combination or configuration of monitoring and / or theft deterrent devices. For example, in the present embodiment, the checkout area 7 is monitored via video surveillance (i.e., via camera 103b), but is not equipped with RFID scanners monitoring ingress and egress points through the checkout area 7. Instead,checkout area 7 utilizes RFID as part of a point-of-sale (POS) device as will be described in more detail hereinbelow. Finally, it is appreciated that while video surveillance and RFID are described herein, other embodiments of the loss prevention system can be implemented using different surveillance and / or theft deterrent devices and / or combinations thereof.

[0019] At least one of the zones in the retail space 1 can be unmonitored. By unmonitored, it is understood that such zones are deprived of surveillance systems to respect privacy rules or regulations affecting private spaces. For example, in the present embodiment the changing room 3 is unmonitored in that there is no video surveillance in that zone. Although changing room 3 is described in the present embodiment, it is appreciated that in other embodiments the retail space 1 can include one or more other zones that are unmonitored. Moreover, in some embodiments, for example where privacy rules or regulations permit, the unmonitored zones can be equipped with theft deterrent devices, such as RFID scanners.

[0020] As can be appreciated, individuals or customers can roam the shopping space to browse and select retail items on display, such as clothing items. Such items can each be equipped with an RFID tag that allows identifying the type of the item (e.g., via a "type" identifier or code that can be shared among all items of the same type) and / or uniquely identifying the item itself (i.e., via a globally unique identifier or code that is specific to each individual item). The customer can carry selected items into the changing room 3 to try them on. Physical restrictions can be implemented (such as via walls or barriers) such that customers are forced to pass through an evaluation zone 5 when entering or exiting the changing room 3.

[0021] Customers who are ready to purchase items can enter the checkout area 7 where they can scan and pay for those items. In the present embodiment, the checkout area 7 is configured for self-checkout and as such includes a self-checkout POS system 107. More specifically, the POS can include a receptacle (such as a basket) configured with an RFID scanner to scan the RFID tags of items placed therein. Those items can then be tallied, and presented to the customer for payment, for example via one or more humaninterface devices (HID) provided as part of the POS system 107, such as a touchscreen, payment keypad, etc. After paying, customers can exit the retail space 1 via exit 9. As can be appreciated, physical restrictions can be implemented (such as via walls or barriers) such that customers are forced to pass through the exit zone 9 when exiting the retail space 1. It can be further appreciated that exit zone 9 can also serve as the entrance to retail space 1. Although in the present embodiment a self-checkout POS system 107 is described, it is appreciated that in other embodiments there system can include other POS systems.

[0022] Due to the lack of surveillance systems in unmonitored zones (e.g., in changing room 3), such zones can correspond to areas where a potential thief is likely to tamper with RFID tags affixed to items, and / or take other measures to attempt to circumvent theft-deterrent systems. Moreover, it is possible that RFID tag tampering can occur within monitored areas of the retail space. A computing system 101 is thus provided to implement a loss prevention method that utilizes video surveillance and RFID scanning to identify potential RFID tag tampering implement corresponding interventions. In the present embodiment, the computing system 101 comprises one or more servers operatively connected to cameras 103, RFID scanners 105 and POS system 107. The computing system 101 comprises at least one processor, and memory having instructions stored thereon which, when executed by the processor, cause the computing system to implement the loss prevention method as will be described in more detail hereinbelow. The computing system 101 can further comprise storage, such as a database, for storing registries generated and modified as part of the method. In the present embodiment, the computing system 101 is located on the premises of the retail space 1. It is appreciated, however, that other configurations are possible. For example, in other embodiments, at least some functions of the loss prevention system can be carried out remotely, such as on one or more remote or cloud-based servers.

[0023] Broadly described, the computing system 101 is configured to generate and store registries as customers move about within the retail space 1. A plurality of registries can be generated as a customer passes through different evaluation zones in the retail space 1at different times, as will be described in more detail hereinbelow. Each registry includes a template characterizing the customer, and a list of items detected on that customer as they pass through the corresponding evaluation zone. The registry can further include other relevant metadata, such as one or more flags that can be applied to the registry, for example to cause one or more downstream actions to be carried out as will be described in more detail hereinbelow.

[0024] In the present embodiment, the template is an Artificial Intelligence (Al)-based template that is generated using Al video analytics. More specifically, the template is generated by analyzing surveillance footage from cameras 103 using one or more Al models to create a unique profile of a customer that can be used for tracking purposes across different zones of the retail space 1. The template is preferably anonymous in that it does not include personally identifiable information. For example, the template can be generated such that it does not include biometric information, or any other uniquely identifiable personal information. In some embodiments, the template can be generated by analyzing anonymous body regions extracted from surveillance video, as described in PCT / CA2023 / 050706 filed May 23, 2023, and entitled "SYSTEM AND METHOD FOR REALTIME ANALYSIS OF ANONYMOUS VIDEO IMAGES", the entirety of which is incorporated herein by reference. In some embodiments, the template can be generated by analyzing anonymized surveillance video, as described in PCT / CA2021 / 050131 filed February 5, 2023, and entitled "SYSTEM AND METHOD FOR PRIVACY-AWARE ANALYSIS OF VIDEO STREAMS AND OPERATION AND LEARNING OF COMPUTER-IMPLEMENTED CLASSIFICATION MODULE", the entirety of which is incorporated herein by reference. It is appreciated that templates generated in these manners can allow distinguishing one customer from other customers in the retail space 1, e.g., to allow locating the customer in the retail space 1 and / or tracking the customer's movements across one or more cameras in the retail space without compromising the customer's privacy. Although Al- based templating is described herein, it is appreciated that other techniques can be used for generating templates for tracking customers within the retail space 1 without relying on biometric information.

[0025] In the present embodiment, the list of items is generated using RFID scanners 105 configured to scan information from RFID tags affixed to the items. The items can be denominated in a list using as much information as possible. For example, such information can include the item type, inventory status, item price, and payment status and / or history (e.g., whether the item was paid for, and / or at what time the item was paid for). Such information can further include a unique identifier, such as a unique item serial number, that allows to uniquely identify an item from other similar items. It is appreciated that in other embodiments, other scanning techniques can be used to detect items where possible.

[0026] In more detail now, and with further reference to Figure 2, the computing system 101 is configured to generate registries at specific evaluation zones within the retail space 1. In the present embodiment an inbound registry 13a is generated when a customer 11 passes through evaluation zone 5 into changing room 3 at a first time. The inbound registry 13a includes a unique profile 15 associated with the customer 11 generated using surveillance footage from camera 103a. The inbound registry 13a further includes a list of items 17a comprising details relating to each item in the customer's possession as detected by RFID scanners 105a while the customer enters the changing room 3. An outbound registry 13b is generated when the customer 11 passes through evaluation zone 5 when exiting changing room 3 at a second time. The outbound registry 13b includes a unique profile 15 associated with the customer 11 generated using surveillance footage from camera 103a. The outbound registry 13b further includes a list of items 17b comprising details relating to each item in the customer's possession as detected by RFID scanners 105a while the customer exits the changing room 3. A checkout registry 13c is generated when the customer 11 attempts to pay for items using the POS system 107 in the self-checkout area 7. The checkout registry 13c includes a unique profile 15 associated with the customer 11 generated using surveillance footage from camera 103b. The checkout registry 13c further includes a list of items 17c comprising details relating to each item the customer attempted to pay for using the POS system 107, for example as scanned using an RFID scanner provided as part of POS system 107. Finally,an exit registry 13d is generated when the customer attempts to leave the retail space 1 through the exit 9. The exit registry 13d includes a unique profile 15 associated with the customer 11 generated using surveillance footage from camera 103c. The exit registry 13d further includes a list of items 17d comprising details relating to each item in the customer's possession as detected by RFID scanners 105b while the customer exits through the exit 9.

[0027] While some registries have been described above, it is appreciated that in other embodiments, additional registries can be created in other areas as needed. For example, in some embodiments, a tampering registry can be generated and / or updated when it is detected that a customer 11 has potentially tempered with RFID tags in a monitored area of the retail space. The tampering registry includes a unique profile 15 associated with the customer 11 generated using surveillance footage from cameras in the retail space, and includes a list of items that the customer 11 potentially tampered with. More specifically, and as described above, RFID tags associated with retail items in the retail space can be regularly scanned to maintain an inventory of all retail items currently in the retail space. If an RFID tag ceases to be detected unexpectedly, e.g. when RFID readers in the retail space do not detect an RFID tag during regular scans for a predetermined period of time, and the RFID tag is not expected to have left a monitored area of the retail space (e.g., the item associated with the RFID tag is not marked as "paid", and the RFID tag is not detected to have left the retail space 1 through the exit 9, or through the evaluation zone 5 to the changing room 3, and / or the item associated with the RFID tag is otherwise still expected to be part of the product inventory in the retail space), and it is determined that the ceasing of the RFID tag detection is correlated with a customer 11 interacting with a retail item in the retail space (e.g., analysis of the video surveillance in the retail space reveals that customer 11 interacted with and / or was in proximity to a retail item at substantially the same time or within a defined window of time when the RFID tag ceased to be detected), the list of items in the customer's tampering registry can be updated to include the item associated with the RFID tag that ceased to be detected, and / or the registry or the customer's profile 15 can be flagged for downstream loss prevention actionor intervention. In some embodiments, an item can be removed from a tampering registry if the RFID tag associated with the item starts being detected again.

[0028] As can be appreciated, the various registries can be matched and compared to determine whether intervention is required to prevent a potential loss. For example, as customers enter the changing room 3 through the evaluation zone 5, an inbound registry 13a can be created for each customer. A plurality of inbound registries 13a created in this fashion can be maintained in a pool of inbound registries. As customers exit the changing room 3 through the evaluation zone 5, an outbound registry 13b can be created for each customer. Once created, each outbound registry 13b can be matched with one of the inbound registries 13a. Outbound registries can be matched with an inbound registry using various criteria. For example, registries can be matched by using an Al-based reconnect, for example by determining that their profiles 15 match the same customer / individual and / or substantially match the same customer / individual (e.g., determining that both profiles correspond to the same customer / individual with a predetermined confidence threshold, such as above 90%, 95%, or 99%). In some embodiments, the matching can be augmented using information from the lists of items 17a and 17b from both registries, for example by determining that both registries contain at least a subset of similar items and / orthat both registries include the same unique item. Once the registries are matched, the lists of items 17a and 17b can be compared. If there is a discrepancy between the lists, it may indicate that the customer has tampered with the RFID tags while in the changing room 3 (e.g. by removing, destroying, or swapping tags, and / or otherwise using a shielded bag or other technique to interfere with the RF scanning process), or that the customer left one or more items in the changing room 3. Therefore, in the case of such a discrepancy, actions can be taken in real time to deter a theft, or to simply confirm that an item was left in the changing room. It is appreciated that other potential losses can be detected, and corresponding actions can be triggered by leveraging registries created at different areas within the retail space 1.

[0029] With further reference to Figure 3, a loss prevention method 150 implemented by computing system 101 will now be described according to a possible embodiment. Whena customer enters the changing room 200, an inbound registry 13a can be created at step 201. When the customer exits the changing room 300, an outbound registry 13b can be created at step 301. A subsequent step 303 involves comparing the inbound registry 13a with the outbound registry 13b. This can involve matching the outbound registry 13b with an inbound registry from a pool of a plurality of inbound registries, and comparing the lists of items 17a and 17b, as discussed above. If there is a match between the lists of items (i.e., the customer walked out of the changing room 3 with the same items than they brought in) no action needs to be taken. If there is a discrepancy (i.e., the customer walked out of the changing room 3 with fewer detected items than they brought in), a changing room sweep 305 can be triggered.

[0030] The changing room sweep 305 can involve a manual intervention whereby an employee in the retail space 1 (such as an employee assigned to the changing room 3) is directed to manually inspect the changing room 3 for any abandoned or missing items. The employee can be directed to conduct the sweep, for example, by receiving instructions or a notification via an employee device such as a smartphone, tablet, desktop computer, radio device, etc. that is part of, or in operative communication with, computing system 101. The instructions can include an indication of the specific item that is missing, such as the item type and / or the item's unique ID. Preferably, the changing room sweep 305 is performed rapidly before the customer is able to check out or leave the retail space 1. In step 307, the employee can provide the results of the changing room sweep via their employee device, e.g., by scanning a found item using an RFID scanner that is provided as part of the employee device, or by providing an input indicating that the item was not found. If the item was found, the inbound registry 13a can be updated in step 309 such that the item that was left in the changing room is removed from the list of items 17a. If the item was not found, or if the RFID tag was found removed from the item, the inbound registry 13a can be flagged in step 311. Alternatively, the inbound registry 13a can be flagged automatically following a determination that the inbound list 17a and outbound list 17b do not match at step 303, and the flag can be cleared only once it has been confirmed that the item was found by the employee during the changing roomsweep 305. As can be appreciated, flagging the registry 13a can allow the retailer to take any action deemed necessary downstream in the customer journey, including restricting the checkout process or signaling a potential theft at the exit, as will be described in more detail hereinbelow.

[0031] When a customer proceeds through the self-checkout 400 to pay for items, a checkout registry 13c can be created at step 401. A subsequent step 403 can involve comparing the checkout registry 13c with the inbound registry 13a. This can involve matching the checkout registry 13c with an inbound registry from a pool of a plurality of inbound registries, and comparing the lists of items 17a and 17c, as discussed above in connection with inbound and outbound registries. If there is a match between the list of items 17a and 17c, or if the checkout registry 13c cannot be matched with an inbound registry (i.e., the customer never entered the changing room), the customer's basket can be marked as "paid" in step 409 once the checkout transaction completes. This can include, for example, marking each item in the list of items 17c as being "paid".

[0032] If there is a mismatch between the list of items 17a and 17c, a subsequent step 405 can involve verifying if inbound registry 13a is flagged. In some embodiments, if the registry 13a is not flagged, additional checks can be carried out in step 406 to determine whether an action should be taken. For example, a subset verification can be carried out where it can be determined whether the list of items 17c in the checkout registry comprises a subset of the list of items 17a in the inbound registry. If the list of items 17c comprises a subset of the list of items 17a, the relative price of the items in lists 17a and 17c can be compared. For example, the total price of the items in the list of items 17c (and / or the price of the subset of items comprised by the list of items 17c) can be compared against the total price of the items in the list of items 17a. If there is a large discrepancy between the total prices of items in lists 17a and 17c, such as if the difference between the total price of lists 17a and 17c is above a predetermined threshold, an action can be taken in step 407 as will be described in more detail below. As can be appreciated, the threshold can be an absolute threshold (e.g., such as a fixed predetermined dollar amount), or a relative threshold (e.g., such as a predetermined percentage of the totalprice of list 17a or 17c). As another example, if there is at least one high-value item missing from list 17c compared to list 17a, such as at least one item having a price that is above a predetermined threshold, an action can also be taken in step 407 as will be described in more detail below. Again, it can be appreciated that the threshold can be an absolute threshold (e.g., such as the item being worth at least a fixed predetermined dollar amount), or a relative threshold (e.g., such as an item representing at least a predetermined percentage of the total price of list 17a or 17c). As can be appreciated, this additional verification can allow triggering an action to be taken to prevent potential theft in scenarios where a customer may have walked into changing room 3 with at least one high-value item and one low-value item, and swapped the RFID tags in the changing room 3 before leaving the changing room with both items (e.g., in an attempt to fool the checkout by paying a low price for the high-value item).

[0033] If the registry 13a is not flagged and if any additional verifications have passed, the customer is not suspected of tampering with RFID tags, and the transaction can proceed as normal (e.g., the customer may have simply added or put back items before going to the checkout). The basket can thus be marked as "paid" in step 409. If the registry is flagged and / or if any additional verifications have failed, an action can be taken. More specifically, the transaction can be paused, and an alert can be generated in step 407. For example, this can involve pausing the transaction at the POS system 107 and providing a notification on the POS system 107 (e.g., via lights and / or sounds) and / or on an employee device, and requiring a store employee to intervene (e.g., by manually verifying the customer's basket of items to make sure there are no tampered or removed RFID tags and / or to confirm that the customer is paying for all items in their possession) before the transaction can proceed. If the employee clears the transaction, the basket can then be marked as "paid" in step 409. Although in the present embodiment the transaction is paused, it is appreciated that the retailer can configure different actions to be taken depending on their requirements. For example, in some configurations, instead of pausing the transaction, a record can be created, for example to assist the retailer inquantifying and / ortracking potential thefts without actively interfering with an individual customer's experience.

[0034] In some embodiments, when a customer proceeds through the checkout 400 and a checkout registry 401 is created, an additional step can involve matching the checkout registry with a tampering registry (e.g., by matching the profile 15 of the checkout registry with a tampering registry having the same profile). If the checkout registry can be matched with a tampering registry that is flagged, a tampering registry that contains a non-null list of items, or a tampering registry that includes items that are not in the checkout registry, a loss prevention action can be taken. This can involve pausing the transaction at the POS system 107 and providing a notification on the POS system 107 (e.g., via lights and / or sounds) and / or on an employee device, and requiring a store employee to intervene (e.g., by manually verifying the customer's basket of items to make sure there are no tampered or removed RFID tags and / or to confirm that the customer is paying for all items in their possession) before the transaction can proceed. If the employee clears the transaction, the basket can then be marked as "paid", and the tampering registry can be cleared.

[0035] When a customer attempts to exit the retail space 500, an exit registry 13d can be created at step 501. A subsequent step 503 can involve determining if the registry is flagged. This can involve matching the exit registry 13d with an inbound registry from a pool of a plurality of inbound registries, as discussed above in connection with inbound and outbound registries. If the exit registry 13d can be matched with an inbound registry 13a, and the inbound registry 13a is flagged, it can be understood that a possible theft is occurring, and a corresponding action can be taken. More specifically, an alarm can be sounded, and an alert can be generated in step 507. This can comprise, for example, activating anti-theft alarms present in the retail space so that employees and / or security can intervene according to the store's policy. This can further comprise logging a potential theft event in the database of the computing system 101. The theft event can include, for example, registry information (including inbound registry 13a, exit registry 13d, and / or any other matched registries and their corresponding lists of items and / or alist of items that were possibly stolen), and video images of the offender, for example as captured by any of cameras 103a, 103b, 103c. The theft event can be used to create a record of the theft which can be shared with the authorities, and / or which can be used to alert employees in case the thief returns to the retail space in the future.

[0036] If the inbound registry 13a is not flagged, or if the exit registry 13d cannot be matched with an inbound registry, a subsequent step 505 can involve determining whether the basket was paid. This can involve matching the exit registry 13d with a checkout registry from a pool of a plurality of checkout registries, and comparing the lists of items 17c and 17d, as discussed above in connection with inbound and outbound registries. If the exit registry 13d can be matched with a checkout registry 13c, and all the items in the list of items 17d pf the exits registry 13d correspond to items marked as paid in the list of items 17c of the checkout registry 13c, it is understood that the customer paid for all items, and no action needs to be taken. If, however, the exit registry 13d cannot be matched with a checkout registry 13c, or if the list of items 17d in the exit registry 13d includes items that are not part of the list of items 17c in the checkout registry 13c, and / or items that are not marked as "paid" in the list of items 17c in the checkout registry 13, it can be understood that the customer did not pay for all their items and / or that a possible theft is occurring, and a corresponding action can be taken. More specifically, an alarm can be sounded, and an alert can be generated in step 507, as discussed above.

[0037] In some embodiments, when a customer proceeds through the exit 500 and an exit registry 501 is created, an additional step can involve matching the exit registry with a tampering registry (e.g., by matching the profile 15 of the exit registry with a tampering registry having the same profile). If the exit registry can be matched with a tampering registry that is flagged, a tampering registry that contains a non-null list of items, or if the tampering registry contains items that are not marked as "paid" in the exit registry, a loss prevention action can be taken. This can involve activating anti-theft alarms present in the retail space so that employees and / or security can intervene according to the store's policy, and / or logging a potential theft event as described above.

[0038] While the above description provides examples of the embodiments, it will be appreciated that some features and / or functions of the described embodiments are susceptible to modification without departing from the spirit and principles of operation of the described embodiments. Accordingly, what has been described above has been intended to be illustrative and non-limiting and it will be understood by persons skilled in the art that other variants and modifications may be made without departing from the scope of the invention.

Claims

Claims1. A method for loss prevention in a physical retail space, the method comprising: generating plurality of registries as a customer moves about in the physical retail space, each registry comprising a template characterizing the customer and a list of detected items; matching a first one of the plurality of registries with a second one of the plurality of registries based at least in part on the template; comparing the lists of items in the matched registries; and causing at least one downstream loss prevention action to be taken based on a comparison between the lists of items.

2. The method according to claim 1, wherein the template is an Al-based template generated using Al video analytics of surveillance footage of the at least one evaluation zone.

3. The method according to claim 1 or 2, wherein the template does not include biometric information of the customer or any other uniquely identifiable information of the customer.

4. The method according to any one of claims 1 to 3, wherein each item in the retail space has an RFID tag affixed thereto, wherein the list of items is generated by reading the RFID tag of each detectable item using at least one RFID reader positioned in the retail space or an evaluation zone in the retail space.

5. The method according to any one of claims 1 to 4, wherein the physical space comprises a monitored area at the entrance / exit of an unmonitored area, further wherein the first one of the plurality of registries is generated when the customer enters the unmonitored area through the monitored area, and the second one of the plurality of registries is generated when the customer exits the unmonitored area through the monitored area.

6. The method according to claim 5, wherein the unmonitored area is a changing room.

7. The method according to claim 5 or 6, wherein the at least one downstream loss prevention action comprises carrying out a manual sweep of the unmonitored area to locate a missing item.

8. The method according to any one of claims 1 to 4, wherein the physical space comprises a checkout area and an exit of the physical retail space, further wherein the first one of the plurality of registries is generated when the customer performs a checkout process in the checkout area, and the second one of the plurality of registries is generated when the customer exits the physical retail space through the exit.

9. The method according to claim 8, wherein the checkout area is a self-checkout area comprising a customer-operated self-checkout POS system.

10. The method according to claim 8 or 9, wherein the at least one downstream loss prevention action comprises pausing the checkout process to allow manual verification by an employee.

11. The method according to any one of claims 1 to 10, wherein the at least one downstream loss prevention action comprises sounding an anti-theft alarm and / or generating a record of a possible theft comprising video images of the customer and a list of possibly stolen items based on the comparing the lists of items in the matched registries.

12. The method according to any one of claims 1 to 11, wherein comparing the lists of items comprises determining whether the list of items in the second one of the plurality of registries comprises a subset of the list of items in the first one of the plurality of registries, identifying at least one item in the list of items in the first one of the plurality of registries that is missing from the list of items in the second one of the plurality of registries, and causing the at least one downstream loss prevention action to be taken if the at least one identified item has a value that is above a predetermined threshold.

13. The method of any one of claim 1 to 12, wherein the plurality of registries comprise a tampering registry generated when an RFID tag ceases to be detected in the retail space following regular scanning by RFID readers positioned in the retail space, and the ceasing of the detection is correlated with the customer interacting with an item in the retail space.

14. A system for loss prevention in a physical retail space, the system comprising: at least one video camera monitoring the retail space; at least one RFID reader positioned in the retail space; and a computing system in operative communication with the at least one video camera and at least one RFID reader, the computing system being configured to:o generate a plurality of registries as a customer moves about in the retail space, each registry comprising:■ a template characterizing the customer, the template being generated using video from the at least one video camera; and■ a list of items associated with the customer, the list of items being generated using readings from the at least one RFID reader; o match a first one of the plurality of registries with a second one of the plurality of registries based at least in part on the template; o compare the lists of items in the matched registries, and cause at least one downstream loss prevention action to be taken based on the comparison.

15. The system according to claim 14, comprising a self-checkout POS system in operative communication with the computing system, further wherein the at least one downstream loss prevention action is taken at the self-checkout POS system.

16. The system according to claim 15, wherein the at least one downstream loss prevention action comprises pausing a self-checkout process for manual employee intervention.

17. The system according to any one of claims 14 to 16, comprising an anti-theft alarm in operative communication with the computing system, further wherein the at least one downstream action comprises sounding the anti-theft alarm.

18. The system according to any one of claims 14 to 17, wherein the at least one downstream action comprises generating and storing a record of a possible theft, said record comprising video images of the customer captured using the at least one video camera and a list of possibly stolen items detected by the at least one RFID reader in at least one of the plurality of registries.

19. A non-transitory computer-readable medium having instructions stored thereon which, when executed by at least one processor of a computing system, cause the computing system to: generate plurality of registries as a customer passes moves about in the physical retail space, each registry comprising a template characterizing the customer and a list of detected items; match a first one of the plurality of registries with a second one of the plurality of registries based at least in part on the template; and compare the lists of items in the matched registries and cause at least one downstream loss prevention action to be taken based on the comparison.

0. A method for loss prevention in a physical retail space, the method comprising: regularly scanning the retail space using RFID scanners to detect RFID tags associated with items in the retail space; monitoring the retail space using cameras to detect customers interacting with one or more of the items in the retail space; determining that detection of an RFID tag has ceased during the regular scanning of the retail space; identifying a correlation between the ceased detection of the RFID tag and an interaction of a customer with one or more of the items in the retail space; and causing at least one loss prevent action to be taken in response to identifying the correlation.

Images