CONTROL METHOD OF THE CONTENT OF A PRESENTATION DISPLAY CASE WITH PRODUCTS EQUIPPED WITH RFID TAGS

Description

Technical Field

[0001] The invention relates to the general field of product display cases. More specifically, the invention concerns display cases equipped with antennas for detecting RFID tags attached to products placed within the display cases.

[0002] The invention finds particular application in display cases for food products. Previous technique

[0003] The technology designated by the Anglo-Saxon acronym RFID (“Radio Frequency Identification”) allows for the automatic detection of labels, generally called “tags” by those in the know.

[0004] This technology traditionally uses waves in a band called ultra-high frequency (i.e., frequencies above approximately 100 megahertz and below a few gigahertz) to enable label detection at distances on the order of a meter. This technology can therefore be used to detect which products are present in a display case, for example, using one or more antennas placed within the case. This allows for the creation of inventories of the products displayed in the case.

[0005] As one might expect, if the display case has a glass door, an object removed from the display case but placed near the glass door can be detected as being inside the display case when it is not. False positives are therefore common in these systems.

[0006] To overcome this drawback, it has been proposed to apply transparent films containing transparent conductive mesh to the glass doors. Such films limit the propagation of waves emitted by RFID tags.

[0007] Document FR 2964217 describes the use of such films on a food product display case.

[0008] The inventors of the present invention have observed that these films do not sufficiently limit the propagation of waves emitted by RFID tags.

[0009] Indeed, if a user removes a product and remains in front of the shop window door, the product can still be detected.

[0010] The earlier patent document US 2006 / 022827 also describes a display case of this type using signal-filtering walls. The solution in this document is also unsatisfactory. Prior art documents CN 204 965 590 and US 2014 / 184391 are also known.

[0011] The invention aims in particular to overcome these drawbacks. Description of the invention

[0012] To this end, the invention proposes a method according to claim 1.

[0013] As an example, detection of products present in the enclosure may include unwanted detection of products located outside the enclosure.

[0014] The inventors of the present invention observed that to prevent the detection of a product removed from a display case, one can track the value of a detection characteristic of the product label between the first and second detections: these two values ​​can reflect movement of the product outside the display case. By using predefined parameters (for example, during a calibration step or a learning step), one can then implement a process to verify whether the product has been removed or not.

[0015] It can be noted that it is the use of these predefined parameters that allows for even better determination of whether or not a product will be released.

[0016] It can also be noted that this process can be implemented by a content management module for the display case, which may include a computer system. This computer system may include a processor, volatile memory, and non-volatile memory (in which at least one predefined parameter is stored). Alternatively, the process can be implemented by a remote server for the display case, which also includes a computer system, with the server and the display case communicating via a network such as the Internet.

[0017] As an example, the first detection is implemented before a shop window door is opened.

[0018] Also, the process may include the preparation of an inventory of the products remaining in the display case after the second detection.

[0019] The predefined detection threshold is the aforementioned predefined parameter. The inventors observed that the output of a product could result in deviations in the characteristic value that accurately reflect the product's output.

[0020] As an indication, if the detection characteristic is representative of the power of a signal received by the antenna and emitted by a tag (better known by the Anglo-Saxon acronym "RSSI: Received Signal Strength Indication"), then the output of a product can cause a significant drop in this power, and this drop is generally stable: this allows us to define a threshold which, when exceeded, illustrates the output of a product.

[0021] The use of a neural network allows for an even more precise determination, as it enables the implementation of learning.

[0022] Indeed, the learning and modification of the artificial neural network can be implemented by receiving commands from the user indicating whether a list of products taken out is correct, or by receiving commands from the user indicating whether an inventory established by the storefront is correct.

[0023] The use of an artificial neural network also makes it possible to adapt, through learning, the operation of each display case (which is advantageous if a fleet of display cases is deployed).

[0024] In addition, this artificial neural network can receive more inputs, and in particular: the number of products detected as present in the display case, the characteristics of the products and their packaging (weight, volume, proportion of water or fat, etc.), or even historical data (typically values ​​of the detection characteristic obtained by different antennas during different detections, or even the user's purchase history data).

[0025] The film that filters electrical and / or magnetic waves can be called an "anti-radiation filter" by those skilled in the art. This filter can be translucent or transparent. For example, a filter marketed by the German company YSHIELD under the reference RDF 72 can be used; it attenuates high-frequency radiation by 30 decibels (at 1 gigahertz) and has a light transmission of 72%.

[0026] According to a particular implementation method, the process includes a threshold adjustment step.

[0027] This particular method of implementation can be especially advantageous if errors are observed in determining whether or not a product is output.

[0028] For example, this threshold adjustment can be implemented automatically.

[0029] As an example, automatic threshold adjustment can be performed based on user error detection. For example, user error detection may include: a display, on a human-machine interface of the showcase, of a list of products considered to have been removed from the showcase, a receipt of a command from the user indicating whether the list is correct (for example if a product is missing from the list or if a product is in the list without having been removed), from the user's command, a modification of the threshold (for example an increase or decrease of the threshold by a predefined amount).

[0030] Alternatively, a user can manually perform an inventory, and the process then only involves receiving a command indicating whether the inventory performed by the system is correct. The threshold can then also be modified (for example, automatically by increasing or decreasing the threshold by a predefined quantity).

[0031] According to yet another alternative, the threshold value can be set by an artificial neural network, which receives as input, for example, the detection characteristics of the tags before the first detection and after the second detection (e.g., the RSSI of a signal emitted by a tag, an antenna identifier, a phase value, a Doppler shift value), as well as a list of products that have been considered out of service. This artificial neural network outputs a threshold value to be used. For training this neural network, one can use information such as the results of manually conducted inventories, or even user order information indicating whether an inventory performed by the display system is correct.

[0032] According to a particular implementation method, at least two types of products are placed in said enclosure, and a different threshold is used for each type of product.

[0033] As a guideline, products of the first type may have smaller dimensions than products of the second type. More than two types of products can be used, and threshold values ​​can be adjusted during calibration or learning phases.

[0034] According to a particular implementation method, the interface between the door and the inside of the display case is equipped with a seal that filters electric and / or magnetic waves, the seal being connected (electrically) to the electrical ground of the display case.

[0035] The inventors of the present invention observed that a label can still be detected if the product has been removed and is present in an area close to this interface (where the edges of the door rest against a surface of the display case, where the seal is located). A filtering seal may contain metal that will prevent the propagation of waves. By connecting this seal to the chassis of the display case, a Faraday cage is also created.

[0036] A seal consisting of foam surrounded by a metal mesh can be used. The French company JACQUES DUBOIS markets such seals under the name TISCAT.

[0037] According to a specific implementation method, the display case is equipped with a plurality of antennas configured to detect the RFID tags of the products present in said enclosure, the first and second detections comprising: a detection, by each antenna of the plurality of antennas, of the products present in said enclosure, for each product detected by several antennas of the plurality of antennas, a value of the characteristic to be retained is determined from all the values ​​of the characteristic obtained for the product by applying a predefined processing, a list of detected products is drawn up with for each detected product a value of the characteristic to be retained or, if the product was detected by a single antenna, the value of the characteristic obtained for that product.

[0038] Using multiple antennas facilitates the detection of products placed in different locations within the display case. In prior art solutions, detection by at least one antenna was sufficient for a label or product to be considered present in the display case. This is not the case here, where if a product is detected by an antenna, a value for the characteristic to be retained is also determined (typically the highest value obtained by one of the antennas).

[0039] Determining the value of the feature to retain may involve selecting the value that indicates the product was most likely not to have been ejected (for example, the highest value if using RSSI). Alternatively, an average of the values ​​for each antenna that detected the product can be calculated.

[0040] According to a particular method of implementation, the antennas are arranged inside recesses formed in shelves of the display case.

[0041] These recesses allow the antennas to be hidden, which are thus placed between two layers of the material forming the shelf.

[0042] It has been observed that contact with the material forming the shelf affects and shifts the operating frequency range of the antennas. Preferably, the antennas are tuned so that the desired operating frequency range (e.g., corresponding to the labels) is obtained after contact with the shelf material.

[0043] For example, antennas can be tuned so that the 865-868 MHz range is reached at -10 dB after contact with the shelf material (e.g., PVC). Tuning antennas may involve creating engravings. Note that a -10 dB tuning is considered sufficient for the antenna to convert 90% of the conducted energy into radiated energy.

[0044] Antennas can have a substantially rectangular shape (which helps to limit material waste during antenna manufacturing).

[0045] According to a particular implementation method, the said list of detected products is drawn up with an identifier of the antenna that obtained the value of the characteristic to be retained.

[0046] This particular implementation method can be used when the value of the characteristic to be retained has been selected from those of each detection.

[0047] The antenna identifier can be used, in particular, to determine the product's position in the display case. If the products are intended to remain in predefined locations within the display case, the detection of a change in antenna (associated with the value of the characteristic to be retained) can be an event that needs to be addressed (which may lead to the generation of an alert message, etc.).

[0048] According to a particular implementation method, the antenna(s) are used in a dual target search mode.

[0049] This search mode, or "Search Mode" in English, is more often referred to by its English name, "dual target". This search mode aims for a search in which tags are entered into the inventory whether their flag is in state A or state B (for example, both possible states for a tag flag), as defined in particular in the EPCglobal Gen2 standard in its version 2.0 and specifically in part 6.3.2.12.1.

[0050] It should also be noted that RFID readers from the American company Impinj compatible with this EPCglobal Gen2 standard can use this "dual target" search mode.

[0051] The inventors observed that the "dual target" search mode is particularly well-suited to detecting products in a display case using multiple antennas: this mode does not prevent the subsequent (short-term) detection of a label by another antenna. This ensures that all antennas are used in a way that yields the best detection characteristics.

[0052] In a particular implementation, the detection characteristic of a detected product label is chosen from the group comprising: the RSSI of a signal emitted by the label, a phase difference of the signal emitted by the label, a frequency variation of the signal emitted by the label, and the number of detections per antenna that detected the product. Alternatively, several detection characteristics are used, and these characteristics can be chosen from this group.

[0053] Regarding the number of detections per antenna that have detected the product, this implementation method is particularly advantageous for display cases equipped with multiple antennas. For these cases, the antennas can be used sequentially until each label has been detected (for example) twice. Therefore, for N antennas, there can be 2N detections of a label. The number of detections, then, between 1 and 2N, represents the probability that a product is present or removed. Also, if only one antenna is used, there will be zero, one, or two detections by that single antenna.

[0054] According to a particular implementation method, the first detection is implemented after a user has authenticated with the storefront and before the storefront has been opened, and the second detection is implemented after the first closing of the storefront which follows said opening of the storefront.

[0055] This specific implementation method ensures that it was indeed the authenticated user who handled and potentially removed products between the two detections. Therefore, if a payment system linked to user accounts is used, the user's account can be debited with the correct amount corresponding to the products actually removed.

[0056] According to a particular implementation method, the user can authenticate themselves using a wireless electronic device.

[0057] In particular, the user will be able to use a wireless electronic device such as a near field communication badge (“NFC: Near Field Communication”), or a smartphone equipped with an NFC function.

[0058] According to a particular implementation method, additional detections of products present in the enclosure are implemented using said antenna to detect product labels on a regular basis, each additional detection further including obtaining, for each detected product, an additional value of the detection characteristic of the detected product label.

[0059] For example, additional detections can be implemented every minute.

[0060] According to a particular implementation method, if an additional detection has been implemented in less than a predefined time period, then the products detected during this additional detection are considered to be the products detected during the first detection, and for each product detected during the additional detection, the additional value is the first value.

[0061] For example, if additional detections are implemented every minute and the predefined duration is 30 seconds, if a user authenticates 25 seconds after the last additional detection, then the last additional detection is considered the first detection.

[0062] This particular method of implementation makes it possible to speed up the process when a user needs to remove a product.

[0063] The invention also provides a showcase for displaying products according to claim 12.

[0064] This showcase can be configured for the implementation of all the implementation modes of the process as defined above. Brief description of the drawings

[0065] Other features and advantages of the present invention will become apparent from the description below, with reference to the accompanying drawings, which illustrate an example of an embodiment without being limiting in any way. In the figures: [ Fig. 1 ] There figure 1 illustrates schematically a shop window according to an example. Fig. 2 ] There figure 2 is a flowchart of the steps in a process, based on an example. Fig. 3 ] There figure 3 illustrates the detection results. Fig. 4 ] There figure 4 illustrates the steps of a treatment using a threshold. Fig. 5 ] There figure 5 illustrates the use of an artificial neural network. Fig. 6 ] There figure 6 illustrates the receipt of feedback from the user. Description of the implementation methods

[0066] We will now describe a display case and a method for managing the content of this display case according to examples of implementation and realization methods of the invention.

[0067] On the figure 1 A display case 100, designed to hold food products within an enclosure, is shown. This display case 100 can be refrigerated and can be installed, for example, in a dining area such as a canteen, cafeteria, or restaurant. In such environments, it may be necessary to closely monitor the contents of the display case, particularly if removing a product requires payment from the user who retrieved it.

[0068] To this end, the invention proposes the use of RFID technology. The following will describe the elements of the display case that enable the use of RFID technology while allowing for the precise detection of products contained within the display case, without detecting products that have been removed but are close to the case as being contained.

[0069] It can be noted that, preferentially, in the present description, the detection of an RFID tag is implemented according to the mode called "dual target".

[0070] Here, the display case 100 has a door 101 that can be opened to remove products placed within the display case. The door 101 is made of glass, while the other parts of the display case 100 are metallic, or at least they form a Faraday cage that prevents the passage of magnetic and / or electrical waves.

[0071] It can be noted that the door can be locked, for example by means of an electromechanical strike plate.

[0072] To limit the passage of these waves through the display case, a 102 anti-radiation filter is applied to the translucent or transparent door, this 102 filter preferably being at least partially translucent or transparent. As a guide, the filter marketed by the German company YSHIELD under the reference RDF 72 can be used; it attenuates high-frequency radiation by 30 decibels (at 1 gigahertz) and has a light transmission of 72%.

[0073] Filter 102 filters in particular the waves emitted or which could be received by antennas which are arranged to the enclosure of the display case 100. Here, these three antennas 103A, 103B, and 103C are arranged so as to be respectively arranged above a shelf 104A, 104B, and 104C.

[0074] As an example, the antennas can be such as those marketed by the Chinese company ECLE Communication Co., Ltd under the reference DP866G12. For example, an antenna operating in the frequency band between 865 and 868 megahertz can be used for RFID applications.

[0075] Other antenna arrangements are possible, and in particular, it is possible to use more antennas.

[0076] Products 105A, 105B, and 105C were placed on shelves 104A, 104B, and 104C, respectively, and all these products are equipped with RFID tags, each containing a unique identifier. These RFID tags can therefore receive signals from antennas 103A through 103C and emit signals that these antennas will receive.

[0077] As can be seen in the figure, the 105A products are quite small (compared to the other products) and are of a first type, the 105B products are of an intermediate size and are of a second type, and the 105C products are the largest and are of a third type.

[0078] As can be understood, the second and third type products 105B and 105C can more easily form an obstacle to the waves between the 106 tags and the 103A to 103C antennas, these different types of products can be taken into account in subsequent detection.

[0079] To implement the process of managing the content of the showcase, a showcase content management module 110 is used, for example a computer system.

[0080] Module 110 includes a processor 111, volatile memory 112, and non-volatile memory 113. In non-volatile memory 113, computer program instructions 114 are stored.

[0081] These instructions 114, when executed by processor 111, cause the following steps to be executed:

[0082] - an initial detection of the products present in the enclosure using antennas 103A to 103C to detect the product labels, the initial detection also including obtaining, for each detected product, an initial value of a characteristic of the detection of the detected product label, subsequent to the first detection, a second detection of the products present in the enclosure using antennas 103A, to 103C to detect the product labels, the second detection further comprising obtaining, for each detected product, a second value of the characteristic of the detection of the label of the detected product, and for each product detected during the first detection and during the second detection, a processing of the first value and the second value of the characteristic, the processing using one or more predefined parameters to determine if the product has been removed from said enclosure between the first detection and the second detection.

[0083] The showcase 101 also includes a human-machine interface 115, for example a touch screen on which the showcase can display lists of products taken out, inventories, or buttons so that the user can indicate whether the list of products taken out is correct or whether the inventory is correct.

[0084] Furthermore, to unlock door 101 of the display case, a user can authenticate themselves using an authentication module 116, which is capable of communicating with wireless authentication devices (typically RFID badges, contactless cards, or mobile phones). This authentication can trigger the initial detection, followed by the unlocking of the electromechanical strike plate. Finally, the authentication allows a user-associated account to be debited with an amount corresponding to the products removed.

[0085] On the figure 2 The steps of a process have been represented in more detail using an example. This process can be implemented using showcase 100 described with reference to the figure 1 In particular, the following steps can be implemented in the form of computer program instructions.

[0086] In the first step, DET_SUPP, a detection process called supplementary detection is implemented for products present within the display case using the display case's antenna(s). The DET_SUPP step is executed regularly, for example, every minute, and it provides, for each detected product, an additional value for a characteristic of the product label detection, for example, an RSSI value of a signal emitted by the label.

[0087] The DET_SUPP step is implemented continuously, unless user authentication (AUTH) is implemented. At that point, the user performs an E01 step to present their unique badge.

[0088] Next, in an EXP_DEL step, we check if an additional detection has been implemented within a predefined timeframe (typically 30 seconds). If so, we consider the products detected during this additional detection to be the same as those detected during the initial detection, and for each product detected during the additional detection, the additional value is considered the first value.

[0089] If the predefined time has expired since the last additional detection, then step DET1 is implemented, in which an initial detection of the products present in the enclosure is performed using the antenna(s) of the display case to detect the product labels. This initial detection also includes obtaining, for each detected product, a first value for the label detection characteristic of the detected product.

[0090] Subsequently, a display case door can be opened by unlocking it, and it can then be opened in the OPEN step. The user then carries out step E02, in which they open the (unlocked) door and remove products.

[0091] The DET_FER step is then implemented, in which the display case checks if the door has been closed. As soon as it is closed, the DET2 step is implemented, in which a second detection of the products present in the enclosure is performed using the display case's antenna(s) to detect the product labels. This second detection also includes obtaining, for each detected product, a second value for the label detection characteristic of the detected product.

[0092] Furthermore, after the DET_FER door closure is detected, the door can be locked. For example, the door can be locked automatically after opening or after a predefined delay (typically a few seconds) if the door has not been opened.

[0093] The TT_SOR step is then implemented in which for each product detected during the first detection and during the second detection, a processing of the first value and the second value of the characteristic, the processing using one or more predefined parameters to determine if the product has been removed from said enclosure between the first detection and the second detection.

[0094] This provides a list of the products that were taken out when the user opened the display case door.

[0095] On the figure 3 , we have represented a table T1 which illustrates the detections of products in a display case equipped with two antennas: antenna 1 and antenna 2.

[0096] During the first detection DET1, antenna 1 detects three products with identifiers P1, P2 and P3 with characteristic values ​​which are here RSSIs of a signal emitted by the label of each product (the characteristic and the RSSI of a signal emitted by the label).

[0097] Antenna 2 detects only product P2, and product P2 is the only product detected by multiple antennas. For this product, we will determine a value for the characteristic to retain based on the values ​​obtained by antennas 1 and 2. Here, this process consists of determining the highest RSSI. The value to retain is therefore the one obtained by antenna 2 for product P2.

[0098] This leads to the development of an L1 list in which are present the products detected with the antenna that detected them.

[0099] During the second detection, only products P1 and P2 were detected, and their RSSIs were different. Specifically, product P2 was detected by antennas 1 and 2 with RSSIs of -55dB and -70dB, respectively.

[0100] The value to retain is -55dB, resulting in list L2 in the figure. P3 no longer appears in the list of detected products.

[0101] Between the DET1 and DET2 detections, we note that for product P2, the RSSI to be retained has gone from -30dB to -55dB, which corresponds to a difference of 25dB.

[0102] If a threshold of 20dB is used, this threshold is exceeded, and product P2 is considered to have been emitted between the two detections. Product P3, which was not detected by any antenna during the second detection, is also considered to have been emitted.

[0103] It can be noted that lists L1 and L2 indicate the antennas that best detected each product, which indicates the position of the products in the display case.

[0104] On the figure 4 We have represented a process that can be implemented during the TT_SOR step described with reference to the figure 2 .

[0105] In this process, a DET_EC step is first implemented to determine the difference between two characteristic values, here RSSI values ​​denoted RSSI_1 and RSSI_2. Then, a comparison is made between this difference and a threshold (CMP_S step). If the threshold is exceeded, the product is considered to have been delivered; if it is not, the product is considered not delivered.

[0106] It should be noted that for different types of products such as products 105A to 105C of the figure 1 Different thresholds can be used. Calibration steps can be implemented to find the appropriate thresholds for these different products.

[0107] On the figure 5 An alternative to the implementation method described in reference to the figure 4 .

[0108] Here, we use an artificial neural network (ANN) which receives as input the RSSI values ​​noted RSSI_1 and RSSI_2 and also possibly one or more PAR parameters, for example the number of labels detected during the first or second detection.

[0109] The artificial neural network RNA delivers an output indicating whether the product has been dispensed or not.

[0110] It should be noted that the artificial neural network RNA can be of the type recurrent neural network (“Recurrent Neural Network” in English).

[0111] The artificial neural network RNA of the figure 5 can be trained by implementing at least the steps that will be described with reference to the figure 6 .

[0112] On the figure 6 , we described a first step AFF_P or AFF_INV in which either we display a list of products taken out on a human-machine interface of the showcase (case AFF_P), or we display an inventory of the products detected in the showcase (case AFF_INV).

[0113] This display can be read by a user who either knows which products have actually been sold or has manually performed an inventory. The user can indicate whether they agree with the display using a button on the human-machine interface. If the display receives information from the user indicating that the displayed list is incorrect, a MOD_S or MOD_RNA step can be implemented.

[0114] The MOD_S step includes modifying the value of a threshold, if a threshold is used as illustrated in the figure 4 .

[0115] This adjustment can be automatic and can consist of increasing or decreasing a predefined quantity. For example, if it is determined that too many products were incorrectly marked as out (information that can be entered by the user), the threshold value can be increased. Conversely, if it is determined that too few products were incorrectly marked as out (information that can be entered by the user), the threshold value can be decreased.

[0116] Alternatively, an artificial neural network can be used which delivers the threshold value and whose training can be done with a user who indicates whether too many or not enough products have been taken out in error.

[0117] Finally, the MOD_RNA step can also involve modifying, or training, an artificial neural network. This training can also be based on user input indicating whether too many or too few products were mistakenly dispensed.

[0118] The implementation and execution methods described above allow for precise management of the content of a display case. This enables the establishment of a more accurate billing system for items removed from the display case.

[0119] It should be noted that if a detection error is detected after the implementation of a second detection, during an additional detection or the next first detection, this makes it possible to debit or credit the account of the last user so that the amount actually debited reflects the products issued.

[0120] In fact, if a change in inventory occurs between two additional detections, the account of the latter user can also be debited or credited so that the amount actually debited accurately reflects the products issued.

Claims

1. A method for managing the content of a display case (100) for presenting products (105A, ..., 105C) equipped with RFID tags (106) placed in an enclosure of the case, the case comprising at least one antenna (103A, ..., 103C) configured to detect the RFID tags of products present in said enclosure and a door (101) at least partially transparent or translucent and covered with a film (102) filtering electrical and / or magnetic waves, the method comprising: - before the door is opened, a first detection (DET1) of products present in the enclosure using said antenna to detect the tags of the products, the first detection also comprising obtaining, for each detected product, a first value of a characteristic of the tag detection of the detected product, - after the first detection and after the door has been closed, a second detection (DET2) of products present in the enclosure using said antenna to detect the tags of the products, the second detection also comprising obtaining, for each detected product, a second value of the characteristic of the tag detection of the detected product, and - for each product detected during the first detection and during the second detection, a processing (TT_SOR) of the first value and the second value of the characteristic, the processing using one or more predefined parameters to determine if the product has been removed from said enclosure between the first detection and the second detection, wherein the processing comprises: - a determination of the difference (DET_EC) between the first value and the second value of the characteristic, and a comparison (CMP_S) of the difference to a predefined threshold, the product being considered to have been removed from said enclosure between the first detection and the second detection if the difference is greater than the predefined threshold, or - the use of an artificial neural network (ANN) having as inputs at least the first and second value of the characteristic or the difference between the first and second value of the characteristic, and as output an indicator whose value indicates if the product has been removed from said enclosure between the first detection and the second detection; wherein if a product is not detected during the second detection, it is considered removed.

2. The method according to claim 1, comprising a step of threshold adjustment (MOD_S).

3. The method according to claim 1 or 2, wherein at least two types of products are placed in said enclosure, and for each type of product, a different threshold is used.

4. The method according to claim 1, wherein the interface between the door and the inside of the case is equipped with a joint filtering electrical and / or magnetic waves, the joint being connected to the electrical ground of the case.

5. The method according to any one of claims 1 to 4, wherein the case is equipped with a plurality of antennas (103A, ..., 103C) configured to detect the RFID tags of products present in said enclosure, the first detection and the second detection comprising: - a detection, by each antenna of the plurality of antennas, of products present in said enclosure, - for each product detected by several antennas of the plurality of antennas, a value of the characteristic to be retained is determined from all the values of the characteristic obtained for the product by applying a predefined processing, - a preparation of a list of detected products (L1, L2) with a value of the characteristic to be retained for each detected product or, if the product has been detected by a single antenna, the value of the characteristic obtained for this product.

6. The method according to claim 5, wherein the antennas are arranged inside recesses formed in the shelves of the case.

7. The method according to claim 5 or 6, wherein the list of detected products is prepared with an identifier of the antenna having obtained the value of the characteristic to be retained.

8. The method according to any one of claims 1 to 7, wherein the characteristic of the tag detection of the detected product is chosen from the group comprising: the RSSI of a signal emitted by the tag, a phase difference of the signal emitted by the tag, a frequency variation of the signal emitted by the tag, a number of detections by the antenna having detected the product, and a number of antennas having detected the product.

9. The method according to any one of claims 1 to 8, wherein the first detection is implemented after user authentication (AUTH) with the case and before opening the case, and the second detection is implemented after the first closure of the case following said opening of the case.

10. The method according to any one of claims 1 to 9, wherein additional detections (DET_SUPP) of products present in the enclosure are implemented using said antenna to detect the tags of the products regularly, each additional detection also comprising obtaining, for each detected product, an additional value of the tag detection characteristic of the detected product.

11. The method according to claim 10, wherein if an additional detection has been implemented within a time less than a predefined duration, then the products detected during this additional detection are considered the products detected during the first detection, and for each product detected during the additional detection, the additional value is the first value.

12. A product display case (105A, ..., 105C) equipped with RFID tags (106), the case comprising an enclosure wherein the products are placed, at least one antenna (103A, ..., 103C) configured to detect the RFID tags of products present in said enclosure, a door (101) at least partially transparent or translucent and covered with a film (102) filtering electrical and / or magnetic waves, and a content management module (110) of the case configured to implement the following steps: - a first detection, before the door is opened, of products present in the enclosure using said antenna to detect the tags of the products, the first detection also comprising obtaining, for each detected product, a first value of a characteristic of the tag detection of the detected product, - after the first detection and after the door has been closed, a second detection of products present in the enclosure using said antenna to detect the tags of the products, the second detection also comprising obtaining, for each detected product, a second value of the characteristic of the tag detection of the detected product, and - for each product detected during the first detection and during the second detection, a processing of the first value and the second value of the characteristic, the processing using one or more predefined parameters to determine if the product has been removed from said enclosure between the first detection and the second detection, wherein the processing comprises: - a determination of the difference (DET_EC) between the first value and the second value of the characteristic, and a comparison (CMP_S) of the difference to a predefined threshold, the product being considered to have been removed from said enclosure between the first detection and the second detection if the difference is greater than the predefined threshold, or - the use of an artificial neural network (ANN) having as inputs at least the first and second value of the characteristic or the difference between the first and second value of the characteristic, and as output an indicator whose value indicates if the product has been removed from said enclosure between the first detection and the second detection.