Supplementary access control system for an RFID-controlled cat flap and method for supplementary control of access through an RFID-controlled cat flap

Abstract

Supplementary access control system (2) for an RFID-controlled cat flap (100), comprising: a sensor (8) for detecting an approaching cat (130) and a camera (10) for capturing at least one image of the approaching cat (130); an image processing unit (18) coupled to the camera (10) and configured to determine, based on the at least one image, whether the cat (130) is a predefined, specific cat; a radio frequency signal generator (14) configured to send an identification signal, the identification signal being suitable for enabling the opening of the RFID-controlled cat flap (100);and a control unit (20) coupled to the sensor (8), the image processing unit (18) and the radio frequency signal generator (14), wherein the control unit (20) is configured, based on the determination by the image processing unit (18) that the cat (130) is the predefined, specific cat, to cause the radio frequency generator (14) to send the identification signal.

Description

The present invention relates to cat flaps. In particular, the present invention relates to RFID-controlled cat flaps. Cat flaps are technical systems that allow a cat access to a building without requiring a front door, balcony door, window, or similar opening to be open. Cat flaps typically allow cats to both enter and exit a building. RFID-controlled cat flaps are widely used. Their purpose is to restrict access to specific cats, limiting it to one or more pre-defined cats. RFID-controlled cat flaps remain closed until a cat approaching the flap is identified by an RFID tag and the flap's control system determines that the detected tag belongs to a pre-defined cat, thus enabling the opening of a hinged door panel. DE 10 2006 048 321 A1 discloses an identification system for animals, in particular cats. This system includes a biometric system for recording the pet's biometric data. While RFID-controlled cat flaps represent a significant improvement over simple mechanical ones, they are not always ideal for controlling access to a building. For example, many RFID-controlled cat flaps rely on the approaching cat carrying an RFID tag. However, carrying the RFID tag in the collar is not ideal due to the risk of losing the collar. Implanting RFID tags is not always desirable for animal welfare reasons. Furthermore, many existing RFID-controlled cat flaps are unable to prevent cats carrying prey from entering the house. Therefore, it would be desirable to provide systems and procedures that can overcome one or more of the shortcomings described above in RFID-controlled cat flaps. Exemplary embodiments of the invention include a supplementary access control system for an RFID-controlled cat flap, comprising: a sensor for detecting an approaching cat and a camera for capturing at least one image of the approaching cat; an image processing unit coupled to the camera and configured to determine, based on the at least one image, whether the cat is carrying prey in its mouth; a radio frequency signal generator configured to send a jamming signal, wherein the jamming signal is suitable for preventing the RFID-controlled cat flap from opening; and a control unit coupled to the sensor, the image processing unit, and the radio frequency signal generator, wherein the control unit is configured to cause the radio frequency signal generator to send the jamming signal based on the detection of the approaching cat.and, based on the determination by the image processing unit that the cat is not carrying prey in its mouth, to instruct the radio frequency signal generator to cease transmitting the interference signal. Exemplary embodiments of the invention allow an RFID-controlled cat flap to be overridden until it is confirmed that the cat is not carrying prey in its mouth. No modifications to the technical system consisting of the RFID-controlled cat flap and the RFID tag worn by the cat are necessary. Existing RFID-controlled cat flaps can be easily retrofitted with the functionality to deny access to a cat that is otherwise authorized but is carrying prey in its mouth. The unwanted introduction of live or dead prey, such as mice, birds, etc., can be very effectively implemented in an existing system architecture consisting of an RFID-controlled cat flap and RFID tag. The access control system is described as a supplementary access control system because it works in conjunction with the RFID-controlled cat flap's access control system, which functions independently but has more limited functionality. The supplementary access control system adds functionality. It can also be referred to as an additional access control system or a higher-level access control system. It builds upon the functioning access control system consisting of the RFID-controlled cat flap and the RFID tag worn by the cat. The control unit of the supplementary access control system is coupled with the sensor, the image processing unit, and the radio frequency signal generator. It is configured to trigger the radio frequency generator to transmit a jamming signal upon detecting an approaching cat, and to cease transmitting the jamming signal once the image processing unit determines that the cat is not carrying prey. Thus, the control unit is configured to manage the transmission of the jamming signal by the radio frequency generator in such a way that the RFID-controlled cat flap will not open until the image processing unit determines that the cat is not carrying prey.The mechanism of emitting a jamming signal upon detection of an approaching cat, and ceasing transmission once the image processing unit determines that the cat is not carrying prey, effectively prevents the RFID-controlled cat flap from opening immediately upon identification of an otherwise authorized cat. The supplementary access control system first establishes a secure default state in which the RFID-controlled cat flap denies the cat access. Only after actively determining that the cat is not carrying prey and after the jamming signal has ceased can the RFID-controlled cat flap and the cat's RFID tag communicate with each other in such a way that the RFID-controlled cat flap can open. The image processing unit is configured to determine, based on at least one image, whether the cat is carrying prey in its mouth. In other words, the image processing unit is configured to analyze the at least one image to determine whether the cat is carrying prey in its mouth. The image processing unit makes a determination, which can be, for example, a yes / no decision. It is also possible for the image processing unit to output a relative indicator, such as a probability value, indicating whether the cat is carrying prey in its mouth. The control unit can then control the radio frequency signal generator using appropriate logic, for example, by using a corresponding threshold value for the probability value. The image processing unit is configured to determine, based on at least one image, whether the cat is carrying prey in its mouth.The image processing unit does not need to deliver a result that is correct in 100% of cases. It is evident that 100% reliability cannot always be achieved given the multitude of degrees of freedom present when capturing the at least one image. The supplementary access control system includes a camera for capturing at least one image of an approaching cat. The camera is a digital camera. The camera may be equipped with, or coupled to, a light source for illuminating the approaching cat. The light source may be an LED, particularly one emitting white light. The light source may be controlled to emit light for a predetermined time after detecting the approaching cat, ensuring continuous exposure during this period. Alternatively, the light source may be linked to the trigger point for capturing the at least one image, similar to a camera flash. The light source may be switched on or off depending on the ambient light level. According to another embodiment, the image processing unit is configured to analyze at least one image using artificial intelligence to determine whether the cat is carrying prey in its mouth. The inventors have discovered that an image processing unit can be implemented using artificial intelligence, which can reliably determine whether or not the cat is carrying prey in its mouth. It has been found that artificial intelligence is very well suited to handling the many degrees of freedom in images of the approaching cat under various environmental conditions and to achieving a high degree of accuracy in evaluating the captured images. Here, the term artificial intelligence refers to a data processing entity that was not created through conventional programming, but is the product of a machine learning process using training images. According to another embodiment, the artificial intelligence comprises a neural network. In particular, the artificial intelligence can comprise a convolutional neural network. The term "convolutional neural network" is well-established and can be translated into German as "faltendes neuronales Netz" (folding neural network). The neural network described herein is an artificial neural network. An example of a suitable neural network is MobileNet from Keras. It is understood that this example is purely illustrative and that there are many different neural networks that can be used in the image processing unit of the present invention. According to another embodiment, the artificial intelligence is a pre-trained artificial intelligence. In particular, the artificial intelligence can be a pre-trained artificial intelligence that has been trained with at least 200 images of cats without prey in their mouths and at least 200 images of cats with prey in their mouths, which are appropriately annotated. Furthermore, the artificial intelligence can be trained with between 500 and 1500 images of cats without prey in their mouths and between 500 and 1500 images of cats with prey in their mouths, which are appropriately annotated. The images of cats without prey in their mouths can be images of more than 20 different cats. Likewise, the images of cats with prey in their mouths can be images of more than 20 different cats.Because prey animals are usually only prey once, a comparable number of different prey animals can be present in the images of cats with prey in their mouths. Using the aforementioned number of images of cats without prey in their mouths and images of cats with prey in their mouths, a very good basic robustness of the artificial intelligence can be achieved. After training with the aforementioned number of images of cats without prey in their mouths and images of cats with prey in their mouths, it may be possible to achieve a success rate of more than 95%, and in particular a success rate of more than 99%, in determining whether the cat is carrying prey in its mouth when analyzing a new image. It is evident that there is a correlation between the number of images available for training the artificial intelligence and the success rate achievable in operation.If a lower hit rate is acceptable, a smaller number of images may suffice for training the artificial intelligence. Conversely, increasing the number of annotated images available for training can improve the hit rate achievable during operation. According to another embodiment, the images of cats with prey in their mouths are images of cats with mice in their mouths and / or images of cats with birds in their mouths and / or images of cats with other prey in their mouths. According to another embodiment, the control unit is configured to cause the radio frequency generator to transmit the jamming signal during predetermined blocking periods, regardless of whether the image processing unit detects that the cat is carrying prey in its mouth. In this way, the supplementary access control system can implement blocking periods for the RFID-controlled cat flap without requiring such functionality in the RFID-controlled cat flap itself. Accordingly, a cat's entry into and / or exit from a building can be categorically prevented during times that represent preferred hunting periods for cats. This further reduces the risk of a cat entering with prey in its mouth. Blocking periods can also be conveniently implemented for other purposes without having to modify the RFID-controlled cat flap.During the predetermined lockout periods, the control unit can cause the radio frequency signal generator to transmit the jamming signal for as long as the sensor detects the cat's presence. It is also possible for the control unit to cause the radio frequency signal generator to transmit the jamming signal continuously during the predetermined lockout periods. Furthermore, the supplementary access control system may have sensors that detect communication between the RFID-controlled cat flap and an RFID tag, and upon detecting such communication during the predetermined lockout periods, the control unit may cause the radio frequency signal generator to transmit the jamming signal. According to another embodiment, the control unit is configured to cause the radio frequency generator to send the jamming signal based on one or more of the input variables: time of day, weather data, season, and temperature, regardless of whether the image processing unit detects that the cat is carrying prey in its mouth or not. This allows for the implementation of additional locking functionality using the supplementary access control system, without requiring such functionality to be present in the RFID-controlled cat flap itself. For example, the control unit can ensure that the RFID-controlled cat flap remains closed during rain unless the temperature drops below a critical value. The logic implemented in the control unit for locking the RFID-controlled cat flap can consider one or more of the aforementioned input variables in any combination. According to another embodiment, the radio frequency signal generator is configured to emit a jamming signal with a fundamental frequency between 100 kHz and 200 kHz, in particular with a fundamental frequency between 120 kHz and 140 kHz. In this way, the radio frequency signal generator is very well suited to effectively jamming commonly used RFID frequencies of 125 kHz and 134 kHz. According to another embodiment, the radio frequency signal generator has a resonant circuit. The components of the resonant circuit, in particular one or more capacitors and one or more inductors, can be designed to provide an interference signal with a fundamental frequency in the frequency ranges mentioned above. According to another embodiment, the radio frequency signal generator is configured to modulate a random or pseudo-random identification number onto a fundamental oscillation of the interference signal. In this way, the radio frequency signal generator can simulate an identification number that is highly unlikely to be stored in the RFID-controlled cat flap and therefore cannot lead to the opening of the RFID-controlled cat flap. The pseudo-random identification number can be predefined or generated when the interference signal is emitted. According to another embodiment, the radio frequency signal generator is configured to modulate a random or pseudorandom bit pattern onto a fundamental oscillation of the interference signal. In this way, the radio frequency signal generator can provide a data pattern that is undecipherable or does not correspond to an expected communication protocol of the RFID-controlled cat flap and is therefore meaningless from the perspective of the RFID-controlled cat flap. The pseudorandom bit pattern can be predefined or generated when the interference signal is transmitted. The modulation of the fundamental frequency of the interfering signal can be amplitude modulation, frequency modulation, phase modulation, or any other suitable modulation. The radio frequency signal generator can prevent the cat flap from opening in two ways. Firstly, by superimposing its interference signal on the cat's RFID tag's transmitted identification, the radio frequency signal generator can create a combined signal that the RFID-controlled cat flap cannot meaningfully decode. Secondly, the radio frequency signal generator can prevent the cat flap from opening by actively transmitting an identification signal that is not stored in the RFID-controlled cat flap. The radio frequency signal generator can actively transmit the interference signal or provide it as a passive, or essentially passive, response to a read signal transmitted by the RFID-controlled cat flap. When passively providing the interference signal, the radio frequency signal generator can, for example, switch its resonant circuit on and off in a random or pseudo-random sequence, or change the parameters of its resonant circuit in a random or pseudo-random sequence. Depending on the technology used, such changes to the radio frequency signal generator's resonant circuit are interpreted by the RFID-controlled cat flap as either modulation of the inductive load or modulation of the reflected signal, i.e., modulation of the so-called backscatter.Since passive RFID tags also provide this type of feedback to the RFID-controlled cat flap, the behavior of a radio frequency signal generator operating passively, as described above, is perceived as a source of interference. The passively operated radio frequency signal generator can be considered an entity that emulates the behavior of a passive RFID tag without transmitting any meaningful information. Regardless of whether the radio frequency signal generator has an active or passive implementation, it is considered here to be a component that emits an interference signal. From the perspective of the RFID-controlled cat flap, the influence of the radio frequency signal generator on the communication between the RFID-controlled cat flap and the RFID tag carried by a cat is perceived as receiving an interference signal. According to another embodiment, the radio frequency signal generator has a transmission power that is on the same order of magnitude as, or greater than, the transmission power of RFID tags commonly used for animal identification. A comparable transmission power can be sufficient to effectively disrupt the communication between an RFID-controlled cat flap and the RFID tag, and such a transmission power level can be provided in a very energy-efficient manner. According to another embodiment, the supplementary access control system also has a network connection. In particular, the supplementary access control system can have a wireless network connection, such as a WLAN connection. The supplementary access control system can be conveniently parameterized via a network connection. The network connection can also be used to implement the supplementary access control system as a distributed system, with some components located near the RFID-controlled cat flap (referred to here as local components) and other components located away from the RFID-controlled cat flap. The network connection allows those components of the supplementary access control system located near the RFID-controlled cat flap to communicate with an external data processing entity, such as an external server.The supplementary access control system can also be monitored and / or controlled via a user interface, such as a smartphone or tablet app. A partially cloud-based implementation of the supplementary access control system is possible. According to a further embodiment, the supplementary access control system is a distributed system in which the image processing unit, or a part thereof, is located remotely from the sensor, camera, radio frequency signal generator, and control unit. In particular, the sensor, camera, radio frequency signal generator, control unit, and optionally a part thereof, can be configured as an integrated device, which may be located near the RFID-controlled cat flap. This device can be referred to as the local device or the local part of the supplementary access control system. The local device can include the network connection through which it can communicate with the remotely implemented image processing unit or part thereof.In this way, a computationally intensive implementation of the image processing unit can be distributed between a local device and a backend dedicated solely to data processing, such as an external server. An optimized compromise between local energy consumption, device size near the RFID-controlled cat flap, and implementation costs can thus be achieved. According to a further embodiment, the supplementary access control system is configured to transmit at least one image via the network connection to the remotely located image processing unit and to receive the image processing unit's determination, via the same network connection, of whether the cat is carrying prey in its mouth. In this way, the image processing unit can be implemented without the technical limitations of a local device near the RFID-controlled cat flap, in particular without the restrictions regarding local computing power and local storage capacity. According to another embodiment, the control unit and the image processing unit are integrated. In particular, the control unit and the image processing unit can be integrated together in the local device, which can be positioned near the RFID-controlled cat flap. The control unit and the image processing unit can, for example, be implemented in a common microcontroller. According to another embodiment, if the image processing unit is implemented locally, it is configured to receive an installation of an artificial intelligence trained with an expanded database via the network connection. In other words, even with a locally implemented image processing unit, it may be possible to achieve greater reliability in determining whether the cat is carrying prey in its mouth through appropriate updates. According to another embodiment, the supplementary access control system is configured to receive modified lockout times via the network connection. In this way, the supplementary access control system can be conveniently parameterized in another dimension during operation. According to another embodiment, the supplementary access control system is configured to receive the time and / or date and / or weather data and / or temperature data via the network connection. In this way, additional data, which may be used for additional functionality of the supplementary access control system, can be conveniently provided externally. According to another embodiment, the sensor is a passive infrared sensor. A passive infrared sensor enables particularly reliable detection of an approaching cat, and its implementation is very energy-efficient. According to another embodiment, the passive infrared sensor is configured to trigger the supplementary access control system to switch from a standby state to an operating state based on the detection of an approaching cat. In this way, the passive infrared sensor can wake up the supplementary access control system as needed, thus enabling extended operation in the energy-saving standby state. An implementation with low overall power consumption can be achieved. According to another embodiment, the sensor is designed as a distance sensor. Additionally or alternatively, the supplementary access control system can have an additional distance sensor that is coupled to the control unit. By providing a distance sensor, a quantitative statement can be made about how close the cat is to the distance sensor and thus to the other components of the supplementary access control system. This information can be taken into account when controlling the other components. It is also possible to determine whether the cat is within a specific area of ​​the supplementary access control system by providing a distance sensor. The distance sensor can be designed as a type of light barrier. This information can also be taken into account when controlling the other components. According to a further embodiment, the sensor or the additional distance sensor provides a signal indicating the distance of the approaching cat, and the control unit is configured to control the trigger time for capturing the at least one image based on the distance of the approaching cat. In this way, relatively constant conditions can be achieved for the size of the cat's head in the at least one image and for the illumination of the cat's head in the at least one image. This, in turn, can contribute to a particularly high success rate in determining whether the cat is carrying prey in its mouth. According to a further embodiment, the sensor designed as a distance sensor, or the additional distance sensor, is configured to continuously monitor the presence of the approaching cat. In particular, the control unit can be configured to control the radio frequency signal generator based on the continuous monitoring by the sensor or the additional distance sensor. Specifically, the control unit can be configured to cause the radio frequency signal generator to transmit the jamming signal for as long as the presence of the approaching cat is indicated by the sensor or the additional distance sensor, unless the image processing unit determines that the cat is not carrying prey in its mouth. By coupling this to the actual presence of the cat, as detected by the sensor or the additional distance sensor, the system can be configured to transmit the jamming signal for as long as the presence of the approaching cat is indicated by the sensor or the additional distance sensor, unless the image processing unit determines that the cat is not carrying prey in its mouth.Detected by the additional distance sensor, the transmission of the jamming signal can be implemented without any assumption about the communication between the RFID-controlled cat flap and the RFID tag carried by the cat, and an extremely reliable locking of the RFID-controlled cat flap can be achieved when the cat is carrying prey in its mouth. According to another embodiment, the supplementary access control system is an autonomously operating system. The term "autonomously operating system" here means that the supplementary access control system does not require any control or commands from outside the system during operation. In particular, it is not necessary for a data or control connection to exist between the RFID-controlled cat flap and the supplementary access control system. The supplementary access control system can fulfill its additional functionality by sending or not sending the jamming signal and does not need to exchange any further information with the RFID-controlled cat flap. According to another embodiment, the supplementary access control system has an internal power supply. In particular, the supplementary access control system can have a rechargeable battery. In this way, the supplementary access control system can operate independently of a power connection, especially independent of a connection to a building's electrical network. It is also possible for the supplementary access control system to be supplied with electrical energy continuously or intermittently as needed. For this purpose, the supplementary access control system can have a standard plug that can be inserted into a household electrical outlet. According to another embodiment, the supplementary access control system can be attached to the RFID-controlled cat flap or to a building structure. "Attachable" here means that the supplementary access control system can be mechanically attached to the RFID-controlled cat flap or to the building structure. No electrical connection is required. The supplementary access control system can be configured as a standalone system to work in conjunction with the RFID-controlled cat flap. The supplementary access control system can be retrofitted to existing RFID-controlled cat flaps. Retrofitting a supplementary access control system into the architecture of existing RFID-controlled cat flaps is possible. According to a further embodiment, the supplementary access control system has a frame structure that can be inserted into a doorway, window opening, or wall opening, and / or that can be attached to an RFID-controlled cat flap. In particular, the frame structure of the supplementary access control system can be attached to a frame structure of the RFID-controlled cat flap. In this way, the frame structures of the RFID-controlled cat flap and the supplementary control system can form a common passage / tunnel through which the cat gains access to the building. The mechanical architecture of the supplementary control system can be adapted to the mechanical architecture of the RFID-controlled cat flap in such a way that neither a change in the cat's usage pattern is required, nor does the supplementary access control system cause any visual impairment for humans. According to another embodiment, the radio frequency signal generator has a coil as an antenna, which is arranged in the frame structure of the supplementary access control system. In this way, the frame structure can fulfill the dual function of a passage adapted to the RFID-controlled cat flap and a support structure for the antenna of the radio frequency signal generator. According to an alternative embodiment, the radio frequency signal generator has a coil as an antenna, which is not located within the frame structure. The complete radio frequency signal generator, including the antenna, can be housed, for example, in an electronics enclosure. According to another embodiment, the camera is positioned on the side of the frame structure facing the interior of the building and is directed through the frame structure. In this way, the frame structure can act as an aperture, ensuring that the conditions for recording approaching cats are as similar as possible. This, in turn, can contribute to a particularly high success rate in determining whether the cat is carrying prey in its mouth. Exemplary embodiments of the invention further comprise an RFID-controlled cat flap comprising: a mounting device with which the RFID-controlled cat flap can be attached to a building wall and / or to a building door and / or to a building window; a door element pivotable relative to the mounting device; a locking mechanism which selectively enables the opening of the pivotable door element; an RFID-based controller which is configured to communicate with an RFID tag carried by a cat approaching the RFID-controlled cat flap and which is configured to cause the locking mechanism to enable the opening of the pivotable door element when it detects an approved RFID tag; and a supplementary access control system according to one of the preceding embodiments.The additional features, modifications and effects, as described above with reference to the supplementary access control system, are applicable analogously to the RFID-controlled cat flap which has such a supplementary access control system. According to another embodiment, the supplementary access control system is integrated into the RFID-controlled cat flap. The components of the supplementary access control system can, for example, be mechanically embedded in the architecture of the RFID-controlled cat flap. The RFID-controlled cat flap and the supplementary access control system can also share a common power supply. According to an alternative embodiment, the supplementary access control system is designed as an independent system and is arranged relative to the RFID-based controller in such a way that the camera of the supplementary access control system is able to capture at least one image of the cat approaching the RFID-controlled cat flap. The supplementary access control system is also considered an independent system if there is a detachable mechanical connection, such as a clip connection, between the RFID-controlled cat flap and the supplementary access control system. Exemplary embodiments of the invention further include a method for supplementary control of access through an RFID-controlled cat flap, wherein the RFID-controlled cat flap is configured to communicate with an RFID tag carried by a cat approaching the RFID-controlled cat flap and to selectively allow the opening of the RFID-controlled cat flap, wherein the method comprises: detecting a cat approaching the RFID-controlled cat flap; based on the detection of the cat approaching the RFID-controlled cat flap, emitting a radio frequency jamming signal suitable for preventing the RFID-controlled cat flap from opening; taking at least one image of the cat approaching the RFID-controlled cat flap; and, based on the at least one image, determining whether the cat is carrying prey in its mouth.and, based on the determination that the cat is not carrying prey in its mouth, the transmission of the radio frequency jamming signal is terminated. The aforementioned steps can be carried out in the order stated, or in any other sensible order, or, where reasonably possible, in parallel. The additional features, modifications, and effects, as described above with reference to the supplementary access control system, are applicable analogously to the procedure for supplementary access control via an RFID-controlled cat flap. Exemplary embodiments of the invention further comprise a computer program / computer program product containing program instructions which, when executed on a computer-based device, cause the following steps: providing a graphical user interface; on the graphical user interface, displaying at least one image of an approaching cat, which has been taken by a camera of a supplementary access control system for an RFID-controlled cat flap or by a camera of an RFID-controlled cat flap; on the graphical user interface, displaying a detection by an image processing unit of the supplementary access control system or...The system comprises a determination by an image processing unit of the RFID-controlled cat flap as to whether the cat is carrying prey in its mouth in at least one image; receiving user input, wherein the user input provides an indication as to whether the cat is carrying prey in its mouth in at least one image, or wherein the user input provides an indication as to whether the image processing unit's determination is correct; and providing the user indication for training the image processing unit with an expanded database. The computer program or computer program product allows for the convenient and efficient collection of user feedback on the accuracy of the image processing unit's determinations regarding the presence of prey in the cat's mouth. In particular, incorrect determinations by the image processing unit can be conveniently and efficiently identified.The feedback allows for future improvements to the image processing unit's hit rate. For example, if a certain number of user indications are present, the image processing unit can be retrained. These user indications can originate from a single user or be collected from multiple users employing multiple identical or similar computer programs or software products for retraining. The supplementary access control system for the RFID-controlled cat flap can be a supplementary access control system according to one of the embodiments of the invention described above. The RFID-controlled cat flap can be an RFID-controlled cat flap according to one of the embodiments of the invention described above. The computer program or computer program product is designed to run on a computer-based device. This computer-based device can be a user device, such as a smartphone, tablet, laptop, or PC. It is also possible for the computer program to run on a server, such as a cloud server, and for a user to access the computer program or its graphical user interface via a web interface, such as a browser. The at least one image of the approaching cat and the image processing unit's detection can be displayed together. The image processing unit's detection of whether the cat is carrying prey in its mouth can be displayed textually, via color coding (e.g., green for no prey in mouth and red for prey in mouth), or in any other suitable manner. User input can be received via corresponding buttons in the graphical user interface, which the user can, for example, click or select via a touchscreen. Before the at least one image of the approaching cat and the image processing unit's detection are displayed, the relevant data may have been received from the supplementary access control system or the RFID-controlled cat flap. Providing user input for training the image processing unit with an expanded database can involve sending the user input to an external data processing facility, such as an external server in the cloud. On the external server, an updated version of the image processing unit, or updated versions of parts thereof, can be created using the feedback data. This updated version can then be deployed as an update to the supplementary access control system or the RFID-controlled cat flap. Specifically, an artificial intelligence can be retrained on the external server using the feedback data, and the result of this new training run can be applied to the image processing unit of the supplementary access control system.the RFID-controlled cat flap. Exemplary embodiments of the invention further include a supplementary access control system for an RFID-controlled cat flap, comprising: a sensor for detecting an approaching cat and a camera for capturing at least one image of the approaching cat; an image processing unit coupled to the camera and configured to determine, based on the at least one image, whether the cat is a predefined, specific cat; a radio frequency signal generator configured to send an identification signal, wherein the identification signal is suitable for enabling the opening of the RFID-controlled cat flap;and a control unit coupled to the sensor, the image processing unit and the radio frequency signal generator, wherein the control unit is configured, based on the determination by the image processing unit that the cat is the predefined, specific cat, to cause the radio frequency generator to send the identification signal. Exemplary embodiments of the invention enable the emulation of an RFID tag, which allows access through the RFID-controlled cat flap. Thus, exemplary embodiments of the invention can extend the selective opening functionality of an RFID-controlled cat flap to cats that neither have an implanted RFID tag nor carry an RFID tag in any other way, for example, in a collar. No modifications to the technical system consisting of the RFID-controlled cat flap and the RFID tag worn by the cat are necessary. Existing RFID-controlled cat flaps can be easily supplemented with the functionality to grant access to authorized cats that do not carry an RFID tag. The potentially undesirable implantation of RFID tags or the risk of loss associated with carrying an RFID tag in a collar can be avoided. In such embodiments, the radio frequency signal generator is configured to send an identification signal suitable for enabling the opening of the RFID-controlled cat flap. In other words, the radio frequency signal generator is configured to send an identification signal that encodes an identity stored in the RFID-controlled cat flap as granting access. The RFID-controlled cat flap may have been pre-programmed with a corresponding identity. It is also possible that the radio frequency signal generator is configured to send an identification signal containing a master identity of the RFID-controlled cat flap. The additional features, modifications and effects, as described above with reference to the supplementary access control system for the analysis concerning prey in the cat's mouth, are analogously applicable to the supplementary access control system for the analysis concerning a predefined, specific cat. According to another embodiment, the image processing unit is set up to analyze the at least one image with artificial intelligence to determine whether the cat is the predefined, specific cat. According to another embodiment, the artificial intelligence has a neural network, in particular a convolutional neural network. According to another embodiment, the artificial intelligence is pre-trained, having been trained with at least 200 images of the predefined, specific cat and at least 200 images of other cats, which are appropriately annotated. In particular, training with between 200 and 500 images of the predefined, specific cat already achieves a very good basic robustness of the artificial intelligence. The number of images of other cats can be, but does not have to be, many times higher. By creating corresponding databases when operating several supplementary access control systems, it is relatively quick to build up a broad database of images of other cats. This broad database of images of other cats can then be incorporated into the training of the artificial intelligence with regard to the predefined, specific cat.After training with the aforementioned number of images of the predefined, specific cat and other cats, it may be possible to achieve a hit rate of more than 95%, and in particular a hit rate of more than 99%, in a newly analyzed image with regard to determining whether the cat is the predefined, specific cat. Exemplary embodiments of the invention further comprise an RFID-controlled cat flap comprising: a mounting device with which the RFID-controlled cat flap can be attached to a building wall and / or to a building door and / or to a building window; a door element pivotable relative to the mounting device; a locking mechanism which selectively enables the opening of the pivotable door element; an RFID-based controller which is configured to communicate with an RFID tag carried by a cat approaching the RFID-controlled cat flap and which is configured to cause the locking mechanism to enable the opening of the pivotable door element when it detects an approved RFID tag; and a supplementary access control system according to one of the preceding embodiments.The additional features, modifications and effects, as described above with reference to the supplementary access control system, are applicable analogously to the RFID-controlled cat flap which has such a supplementary access control system. According to another embodiment, the supplementary access control system is integrated into the RFID-controlled cat flap. According to an alternative embodiment, the supplementary access control system is designed as an independent system and arranged in such a way as to the RFID-based control system that the camera of the supplementary access control system is able to capture at least one image of the cat approaching the RFID-controlled cat flap. Exemplary embodiments of the invention further include a method for supplementary control of access through an RFID-controlled cat flap, wherein the RFID-controlled cat flap is configured to communicate with an RFID tag carried by a cat approaching the RFID-controlled cat flap and to selectively enable the opening of the RFID-controlled cat flap, wherein the method comprises: detecting a cat approaching the RFID-controlled cat flap; taking at least one image of the cat approaching the RFID-controlled cat flap; based on the at least one image, determining whether the cat is a predefined, specific cat; and based on the determination that the cat is a predefined, specific cat, sending an identification signal suitable for enabling the opening of the RFID-controlled cat flap.The steps mentioned can be carried out in the order given, in any other sensible order, or, where practically feasible, in parallel. The additional features, modifications, and effects, as described above with reference to the supplementary access control system, are analogous to the procedure for supplementary access control via an RFID-controlled cat flap. Exemplary embodiments of the invention further comprise a computer program / computer program product containing program instructions which, when executed on a computer-based device, cause the following steps: providing a graphical user interface; on the graphical user interface, displaying at least one image of an approaching cat, which has been taken by a camera of a supplementary access control system for an RFID-controlled cat flap or by a camera of an RFID-controlled cat flap; on the graphical user interface, displaying a detection by an image processing unit of the supplementary access control system or...The system comprises a determination by an image processing unit of the RFID-controlled cat flap as to whether the cat in the at least one image is a predefined, specific cat; receiving user input, wherein the user input provides an indication as to whether the cat in the at least one image is the predefined, specific cat, or wherein the user input provides an indication as to whether the image processing unit's determination is correct; and providing the user indication for training the image processing unit with an expanded database. The computer program or computer program product allows for the convenient and efficient collection of user feedback on whether the image processing unit's determinations regarding the predefined, specific cat are correct. In particular, incorrect determinations by the image processing unit can be conveniently and efficiently identified.The feedback can be used to improve the image processing unit's accuracy in the future. For example, if a certain number of user indications are present, the image processing unit can be retrained. These user indications can originate from the user whose supplementary access control system or RFID-controlled cat flap regulates access for a specific, predefined cat. Images from other users whose supplementary access control system or RFID-controlled cat flap regulates access for other cats and who use the same or similar computer programs or software can be used as additional images of other cats for retraining. The supplementary access control system for the RFID-controlled cat flap can be a supplementary access control system according to one of the embodiments of the invention described above.The RFID-controlled cat flap can be an RFID-controlled cat flap according to one of the embodiments of the invention described above. The computer program or computer program product is designed to run on a computer-based device. This computer-based device can be a user device, such as a smartphone, tablet, laptop, or PC. It is also possible for the computer program to run on a server, such as a cloud server, and for a user to access the computer program or its graphical user interface via a web interface, such as a browser. The image of the approaching cat and the image processing unit's identification can be displayed together. The image processing unit's identification of the cat as a predefined, specific cat can be represented textually, via color coding (e.g., green for the predefined, specific cat and red for other cats), or in any other suitable manner. User input can be received via corresponding buttons in the graphical user interface, which the user can click or select via a touchscreen. Prior to displaying the image of the approaching cat and the image processing unit's identification, the relevant data may have been received from the supplementary access control system or the RFID-controlled cat flap. Providing user input for training the image processing unit with an expanded database can involve sending the user input to an external data processing facility, such as an external server in the cloud. On the external server, an updated version of the image processing unit, or updated versions of parts thereof, can be created using the feedback data. This updated version can then be deployed as an update to the supplementary access control system or the RFID-controlled cat flap. Specifically, an artificial intelligence can be retrained on the external server using the feedback data, and the result of this new training run can be applied to the image processing unit of the supplementary access control system.the RFID-controlled cat flap. The computer program or computer program product may not only be intended for collecting user feedback for retraining the image processing unit. It is also possible that the computer program or computer program product is configured for compiling annotated images for the initial training of the image processing unit.In a more ambitious manner, exemplary embodiments of the invention further comprise a computer program / computer program product containing program instructions that, when executed on a computer-based device, initiate the following steps: providing a graphical user interface; on the graphical user interface, displaying at least one image of an approaching cat, which has been captured by a camera of a supplementary access control system for an RFID-controlled cat flap or by a camera of an RFID-controlled cat flap; receiving a user input, wherein the user input includes a user indication as to whether the cat in the at least one image is the predefined, specific cat; and providing the user indication for the initial training of the image processing unit.The additional features, modifications, and effects described above regarding the retraining of the image processing unit are analogously applicable to the initial training. Images of other cats can be used for the initial training from a database. This database can, in particular, contain images of other users who use the same or similar computer programs or software. In this way, after commissioning a supplementary access control system for an RFID-controlled cat flap, a database can be quickly established that allows for a high success rate in determining whether the cat is a predefined, specific cat. In the embodiments described so far, the aspects of determining whether the cat is carrying prey in its mouth and determining whether the cat is a predefined, specific cat have been described separately. However, it is possible for both aspects to be implemented together. In particular, it is possible, and hereby expressly disclosed, that supplementary access control systems, RFID-controlled cat flaps, and methods for supplementary access control via an RFID-controlled cat flap according to exemplary embodiments of the invention have features of both aspects. Furthermore, it is particularly possible that the same or different subcombinations of features of both aspects are implemented in the supplementary access control systems, RFID-controlled cat flaps, and methods for supplementary access control via an RFID-controlled cat flap. Furthermore, it is hereby emphasized and expressly disclosed that the aspect of determining whether the cat is a predefined, specific cat constitutes an invention, independent of the implementation of the cat flap as an RFID-controlled cat flap and independent of the implementation of a supplementary access control system. In other words, the aspect of determining whether a cat approaching the cat flap is a predefined, specific cat and controlling the cat flap based on this determination is considered an independent invention. Formulated in claim form, exemplary embodiments of the invention comprise a cat flap which has: a mounting device with which the cat flap can be attached to a building wall and / or a building door and / or a building window; a door element that pivots relative to the mounting device;a locking mechanism that selectively allows the opening of the pivoting door element; a sensor for detecting an approaching cat and a camera for capturing at least one image of the approaching cat; an image processing unit coupled to the camera and configured to determine, based on the at least one image, whether the cat is a predefined, specific cat;and a control unit coupled to the sensor and the image processing unit, wherein the control unit is configured, based on the image processing unit's determination that the cat is the predefined, specific cat, to initiate the locking mechanism and allow the opening of the pivoting door element. The additional features, modifications, and effects, as described above with reference to the supplementary access control system and the RFID-controlled cat flap, are, where appropriate, applicable analogously to such a cat flap. A corresponding method for operating a cat flap is also expressly disclosed herewith. Further exemplary embodiments of the invention are described below with reference to the accompanying drawings. Fig. 1 shows an RFID-controlled cat flap with which a supplementary access control system according to an exemplary embodiment of the invention can cooperate or into which a supplementary access control system according to an exemplary embodiment of the invention can be integrated, partly in a schematic cross-sectional view and partly as a block diagram. Fig. 2 shows a supplementary access control system according to an exemplary embodiment of the invention in a schematic perspective view. Fig. 3 shows a supplementary access control system according to an exemplary embodiment of the invention in a block diagram. Fig. 4 illustrates the operation of a supplementary access control system according to an exemplary embodiment of the invention by means of a flowchart.Figure 5 shows two exemplary images of a cat without prey in its mouth and the same cat with prey in its mouth, as they can be captured and processed by a supplementary access control system according to an exemplary embodiment of the invention. Figure 6 illustrates the operation of a supplementary access control system according to an exemplary embodiment of the invention by means of a flowchart. Figure 7 shows two exemplary images of two different cats, as they can be captured and processed by a supplementary access control system according to an exemplary embodiment of the invention. Fig. 1 shows an RFID-controlled cat flap 100 according to an exemplary embodiment, partly in a schematic cross-sectional view and partly as a block diagram. The RFID-controlled cat flap 100 can be equipped with a supplementary access control system according to an exemplary embodiment of the invention, or can cooperate with a supplementary access control system according to an exemplary embodiment of the invention. The RFID-controlled cat flap 100 can thus be considered a starting point for the extension by means of the supplementary access control system according to exemplary embodiments of the invention. The RFID-controlled cat flap 100 has a frame structure 102, which is shown in cross-section in Fig. 1. The frame structure 102 forms a passage / tunnel through which a cat 130 can enter and exit a building, such as a house. Fig. 1 shows a situation in which the cat 130 is about to enter the building from the outside through the tunnel formed by the frame structure 102. The frame structure 102 can be inserted into a corresponding opening in a front door or in a corresponding opening in another part of the building. The RFID-controlled cat flap 100 features a pivoting door element 104. Provided the pivoting door element 104 is not locked, the cat 130 can push it open with its snout / head and gain access to the building. The pivoting door element 104 is essentially dimensioned to completely close the tunnel through the frame structure 102. In the exemplary embodiment of Fig. 1, the RFID-controlled cat flap has a locking mechanism which is composed of several components. On the one hand, the locking mechanism has a first pivoting stop 106 and a second pivoting stop 108, wherein the second pivoting stop 108 is movable by an actuator 110 and / or can be fixed in its upright position by the actuator 110. The first pivoting stop 106 and the second pivoting stop 108 are arranged on the frame structure 102 in an area where the lower end of the pivoting door element 104 is positioned when the pivoting door element 104 is in its equilibrium position. Upon entering the building, the cat 130 can move the second pivoting stop 108 into a horizontal position with the aid of the pivoting door element 104, provided that the second pivoting stop 108 is not fixed in its upright position. When exiting the building, the cat 130 can bring the first swiveling stop 106 into a horizontal position using the swiveling door element 104.In a closed state, the lower end area of ​​the pivoting door element 104 is arranged between the first pivoting stop 106 and the second pivoting stop 108. On the other hand, the locking mechanism has a locking pin 113, which is located in an upper area of ​​the frame structure 102 above the pivoting door element 104 and can be inserted into a corresponding recess 105 in the upper area of ​​the pivoting door element 104 by means of an actuator 112 and can be withdrawn from it. With the aid of the locking pin 113 and the second pivoting stop 108, which can be fixed in the upright position, the pivoting door element 104 can be fixed in the closed position. In other words, the RFID-controlled cat flap 100 is able to selectively allow or prevent the opening of the pivoting door element. It is emphasized that the locking mechanism described above is purely exemplary. Any other suitable type of locking mechanism may be provided. It is also possible that the locking is achieved solely by means of the locking pin 113 or by means of the second pivoting stop 108. The RFID-controlled cat flap 100 can be configured to be generally closed, meaning that the locking pin 113 is inserted into the receptacle 105 and the second pivoting stop 108 is fixed in the upright position. Selective opening of the pivoting door element 104 is enabled during operation by identifying the cat 130 and releasing the locking mechanism. The cat 130 is identified as follows. In the exemplary embodiment shown in Fig. 1, the cat 130 has an implanted RFID tag 132, also referred to as an RFID chip. It is understood that the RFID tag can also be part of the cat 130's collar. When the cat 130 approaches the RFID-controlled cat flap 100, a control unit 116 of the RFID-controlled cat flap 100 initiates communication with the RFID tag 132. For this purpose, the RFID-controlled cat flap 100 has a coil 118. The coil 118 is wound around a central area of ​​the frame structure 102 and acts as an antenna on the side of the RFID-controlled cat flap 100 for communication with the RFID tag 132. The control unit 116 and the RFID tag 132 communicate using a suitable protocol to read the identification information stored in the RFID tag 132.The retrieved identification information is compared with the data stored in control unit 116 for authorized cats. If control unit 116 identifies cat 130 as authorized, it causes actuator 110 and actuator 112 to release the second pivoting stop 108 and retract the locking pin 113. Consequently, cat 130 can open the pivoting door element 104 and enter the building. The RFID-controlled cat flap 100 also features a power supply 114, which provides electrical energy to the control unit 116 and the actuators 110 and 112. The power supply 114 can be a local power source, such as a rechargeable battery, or a connection to a household electrical outlet or any other suitable type of electrical power supply. Fig. 2 shows a schematic perspective view of a supplementary access control system 2 for an RFID-controlled cat flap, such as the RFID-controlled cat flap 100 of Fig. 1. The supplementary access control system 2 has a frame structure 4. The frame structure 4 can be attached to the frame structure of an existing RFID-controlled cat flap, such as the frame structure 102 of the RFID-controlled cat flap 100 of Fig. 1, via clip connections 5. Thus, the frame structure 4 of the supplementary access control system 2 can form an extension of the tunnel through which a cat passes when entering or leaving a building. In the lower part of the frame structure 4 of the supplementary access control system 2, an electronics housing 6 is arranged, which contains various components of the supplementary access control system 2, described in detail below. A sensor 8 for detecting an approaching cat is arranged on the outward-facing front face of the electronics housing 6. The sensor 8 can be a passive infrared sensor. Furthermore, an additional distance sensor 12 is provided on the top of the electronics housing 6, close to the front face. In the exemplary embodiment shown in Fig. 2, the additional distance sensor is oriented upwards and comprises an IR light source, such as an IR LED, and an IR photosensor, such as an IR photodiode. By emitting IR light via the IR light source and evaluating the measured values ​​of the IR photosensor, the additional distance sensor 12 can provide an indication of whether the approaching cat is at least partially within the frame structure 4. A cat within the frame structure 4 reflects a significant amount of IR light, which is reflected in the measured values ​​of the IR photosensor. The additional distance sensor can function as a kind of light barrier, providing an indication of how close the cat has come to the supplementary access control system 2 or whether the cat is still within / at the supplementary access control system 2. On the side of the electronics housing 6 facing the interior of the building, a structure is provided in which a camera 10 and a light source 11 are integrated. The camera 10 and the light source 11 are directed obliquely upwards. In this way, the camera 10 and the light source 11 are directed towards the area where the head of an approaching cat is expected to be when it is about to enter the building through the frame structure 4 of the supplementary access control system 2. The camera 10 is a digital camera. The light source 11 is intended to illuminate the approaching cat and can, for example, be implemented as an LED. The functional linkage of the described components of the supplementary access control system 2 and their interaction during operation are described with reference to the following figures. Fig. 3 shows a supplementary access control system 2 according to an exemplary embodiment of the invention in a block diagram. The local components of the supplementary access control system 2 can be arranged in the device structure shown in Fig. 2. However, it is also possible that the components of the supplementary access control system 2 shown in Fig. 3 are assembled in another suitable arrangement. The sensor 8, the camera 10 with the light source 11, and the additional distance sensor 12 are connected to a microcontroller 16. In the exemplary arrangement of Fig. 2, the microcontroller 16 can be housed in the electronics enclosure 6. In the exemplary embodiment of Fig. 3, part of an image processing unit 18 and a control unit 20 are implemented on the microcontroller 16. The microcontroller 16 is connected to a radio frequency signal generator 14. The radio frequency signal generator 14 is configured to emit a jamming signal suitable for disrupting the communication between an RFID-controlled cat flap and an RFID tag carried by a cat, such that the cat cannot be identified by the RFID-controlled cat flap. The radio frequency signal generator 14 can have a resonant circuit comprising one or more capacitors and one or more inductors. The one or more inductors can serve as antennas for emitting the jamming signal. In the exemplary arrangement shown in Fig. 2, the one or more capacitors can be arranged in the electronics housing 6, and the one or more inductors can be coils running around the cat's passage area in the frame structure 4.This coil(s) functions as an antenna for transmitting the interference signal, similar to the coil 118 shown in Fig. 1, which serves as the antenna for the RFID-controlled cat flap 100 for communication with the RFID tag 132. It is also possible that the one or more capacitors and the one or more inductors are housed within the electronics enclosure 6. In addition to or as an alternative to the ability of the radio frequency signal generator 14 to emit the interference signal described above, the radio frequency signal generator 14 can be configured to send an identification signal suitable for enabling the opening of the RFID-controlled cat flap. The microcontroller 16 is also connected to a network port 22. The network port 22 can be, in particular, a wireless network port, such as a WLAN port. The microcontroller 16 is connected via the network port 22 to external data processing devices, shown here as part of a cloud 26. The microcontroller 16 can exchange data via the network port 22 with external entities, such as a remotely located server 28 and / or with monitoring and control programs, such as smartphone or tablet apps. In the exemplary embodiment shown in Fig. 3, part of the image processing unit 18 is implemented in the remotely located server 28. The supplementary access control system 2 also includes an internal power supply 24. The internal power supply 24 provides electrical power to the camera 10, the light source 11, the additional distance sensor 12, the microcontroller 16, and optionally the radio frequency signal generator 14 and the network port 22. The internal power supply 24 can be a rechargeable battery or a connection to a household electrical outlet or any other suitable type of electrical power supply. In the exemplary embodiment shown in Fig. 3, the image processing unit 18 is implemented partly on the microcontroller 16 and partly on the remotely located server 28. However, it is also possible for the image processing unit 18 to be implemented entirely on the microcontroller 16 or entirely in the cloud 26. If part or all of the image processing unit is implemented in the cloud 26, the supplementary access control system is a distributed system. In such a distributed system, some of the components of the supplementary access control system 2, namely those components shown to the right of the network connector 22 in the block diagram of Fig. 3, are located locally in the device structure shown in Fig. 2, while other components are implemented remotely from the device structure shown in Fig. 2, for example, in the cloud 26.The interaction and cooperation of the components of the supplementary access control system 2 of Fig. 3 is described below with reference to the flowchart of Fig. 4. Fig. 4 shows a flowchart illustrating the operation of a supplementary access control system 2 according to an exemplary embodiment of the invention, in particular the operation of the supplementary access control system 2 of Fig. 3, by means of an exemplary process run. In step 40, sensor 8 detects a cat approaching the supplementary access control system 2. Based on this detection, sensor 8 sends a signal to microcontroller 16, which switches it from a standby state to an operating state. This can also be described as waking up microcontroller 16, shown as step 42. Based on the knowledge of a cat's presence, the control unit 20 of microcontroller 16 causes radio frequency signal generator 14 to emit the interference signal. This is summarized in Fig. 4 in step 44 as the emission of the interference signal. In step 46, the additional distance sensor 12 monitors whether the approaching cat is within the frame structure 4, thus ensuring that the distance between the cat and the camera 10 is within a suitable range. Once the distance is suitable for capturing one or more images, the control unit 20, which continuously receives an indication of the cat's distance from the additional distance sensor 12, transmits a trigger signal to the camera 10. The camera 10 then captures at least one image of the cat in step 48. The trigger signal from the control unit can also be transmitted to the light source 11 or forwarded from the camera 10 to the light source 11. It is possible that the light source 11 activates the cat's exposure in response to the trigger signal. It is also possible that such exposure occurs only if the ambient light conditions require it.The camera 10 can be equipped with appropriate sensors and logic and control the light source 11 accordingly. In step 50, the image captured by the cat is sent via network port 22 to the external server 28 in the cloud 26. In step 52, the image analysis to determine whether the approaching cat is carrying prey in its mouth is performed in the part of the image processing unit 18 implemented on the external server 28 in the cloud 26. As described above, it is also possible for the image processing unit 18 to be implemented entirely locally in the microcontroller 16. In other words, step 52 can be performed either in the cloud 26 or locally in the microcontroller 16. Accordingly, step 50, i.e., sending the image via network port 22, can take place or be omitted. In step 52, the image processing unit determines whether the cat is carrying prey in its mouth, as shown in the captured image. This determination is made using artificial intelligence. Specifically, the image processing unit analyzes whether the cat is carrying prey using a pre-trained neural network. The neural network, which has been trained with appropriately annotated images of cats without prey in their mouths and cats with prey in their mouths, arrives at a result indicating whether the cat is carrying prey or not. The result can be a yes / no decision, a probability value indicating whether the cat is carrying prey, or another suitable form. The part of the image processing unit 18 implemented on the server in the cloud 26 sends the result of the image analysis back to the microcontroller 16 via the network connection 22. In step 54, the control unit 20 of the microcontroller 16 receives a signal via network connection 22 indicating whether the cat is carrying prey in its mouth. Based on this signal, the control unit 20 makes a decision in step 56. If it has been determined that the cat is carrying prey in its mouth, the additional distance sensor 12 continuously monitors in step 58 whether the cat remains near the supplementary access control system 2. If so, the control unit 20 ensures that the radio frequency signal generator 14 continues to transmit the jamming signal. This is illustrated in step 60. If the additional distance sensor 12 determines in step 58 that the cat is no longer near the supplementary access control system 2, the control unit 20 instructs the radio frequency signal generator 14 to cease transmitting the jamming signal in step 62.The control unit 20 also causes the radio frequency signal generator 14 to stop sending the jamming signal when it has been determined in step 56 that the cat is not carrying prey in its mouth. The two different paths between step 56 and step 62 achieve two things. Firstly, the jamming signal is stopped when the cat is not carrying prey in its mouth, allowing the cat to enter the building through the RFID-controlled cat flap. Secondly, based on the detection that the cat is carrying prey in its mouth, the jamming signal is emitted until the cat has moved away from the supplementary access control system, thus preventing entry into the building with prey in its mouth. After the interference signal is terminated in step 62, the supplementary access control system 2, in particular the microcontroller 16, enters a standby state in step 64. The supplementary access control system 2 wakes up from this standby state upon the next detection of an approaching cat by sensor 8, as described above with reference to step 40. Fig. 5 shows two exemplary images of a cat without prey in its mouth and the same cat with prey in its mouth, as they can be captured and processed in a supplementary access control system according to an exemplary embodiment of the invention. Fig. 5A shows an image of a cat 130 without prey in its mouth. In the image, the cat's head is framed by the structure 4 of a supplementary access control system 2, as shown, for example, in Fig. 2. A section of the surroundings of the supplementary access control system 2 can be seen as the background between the cat's head and the structure 4. Fig. 5B shows an image of the same cat 130 as in Fig. 5A, but with prey in its mouth. In the image of Fig. 5B, the cat 130 has a mouse 140 in its mouth. Again, the head of the cat 130 in the image shown is framed by the frame structure 4, and a section of the surroundings of the supplementary access control system 2 can be seen as the background. As can be seen from the comparison of the images in Fig. 5A and Fig. 5B, images taken with camera 10 of a supplementary access control system 2 can differ in many respects, such as in the lighting, the direction of the cat's gaze 130, the blurriness of the image, the brightness of the background, etc. Despite this multitude of degrees of freedom in image composition, it has been found that an artificial intelligence trained with approximately 1000 images of cats without prey in their mouths and approximately 1000 images of cats with prey in their mouths can achieve a success rate of greater than 95% or even greater than 99% when analyzing newly taken images to determine whether the cat is carrying prey in its mouth. Fig. 6 shows a flowchart illustrating the operation of a supplementary access control system 2 according to an exemplary embodiment of the invention, in particular the operation of the supplementary access control system 2 of Fig. 3, by means of an exemplary process flow. Some steps of the flowchart in Fig. 6 are analogous to the corresponding steps of the flowchart in Fig. 4. These steps are described again below, but reference is also made to the description of Fig. 4. In step 70, sensor 8 detects a cat approaching the supplementary access control system 2. Based on this detection, sensor 8 sends a signal to microcontroller 16, which switches it from a standby state to an operating state. This can also be described as waking up microcontroller 16, represented as step 72. In step 74, the additional distance sensor 12 monitors whether the approaching cat is within the frame structure 4, thus ensuring that the distance between the cat and the camera 10 is within a suitable range. Once the distance is suitable for capturing one or more images, the control unit 20, which continuously receives an indication of the cat's distance from the additional distance sensor 12, transmits a trigger signal to the camera 10. The camera 10 then captures at least one image of the cat in step 76. The trigger signal from the control unit can also be transmitted to the light source 11 or forwarded from the camera 10 to the light source 11. It is possible that the light source 11 activates the cat's exposure in response to the trigger signal. It is also possible that such exposure occurs only if the ambient light conditions require it.The camera 10 can be equipped with appropriate sensors and logic and control the light source 11 accordingly. In step 78, the image analysis to determine whether the approaching cat is a predefined, specific cat is performed in the part of the image processing unit 18 that is implemented locally in the microcontroller 16. It is also possible for step 78 to be performed in a part of the image processing unit 18 located in the cloud 26, for example, in a part of the image processing unit 18 implemented on an external server 28. For this to be the case, before step 78 is performed, the image captured of the cat would be sent to the external server 28 in the cloud 26 via network port 22. In other words, step 78 can be performed both in the cloud 26 and locally in the microcontroller 16. In step 78, the image processing unit determines whether the cat is a predefined, specific cat according to the captured image. The identification is performed using artificial intelligence. Specifically, the image processing unit analyzes, using a pre-trained neural network, whether the cat in the captured image is a predefined, specific cat. The neural network, which has been trained with appropriately annotated images of the predefined, specific cat and images of other cats, arrives at a result as to whether the cat in the captured image is indeed the predefined, specific cat. The result can be a yes / no decision, a probability value indicating whether the cat is the predefined, specific cat, or another suitable form.When step 78 is implemented in Cloud 26, the part of the image processing unit 18 implemented on the server in Cloud 26 sends the result of the image analysis back towards the microcontroller 16 via the network port 22. In step 80, the control unit 20 of the microcontroller 16 receives confirmation that the cat is the predefined, specific cat. If it has been determined that the cat is the predefined, specific cat, the control unit 20, in step 82, instructs the radio frequency signal generator 14 to send an identification signal suitable for opening the RFID-controlled cat flap. The radio frequency signal generator 14 can send an identification signal that encodes an identity stored in the RFID-controlled cat flap as authorizing access. The RFID-controlled cat flap may have been pre-programmed with a corresponding identity. It is also possible that the radio frequency signal generator is configured to send an identification signal containing a master identity of the RFID-controlled cat flap. The method illustrated in Fig. 6 allows a cat that should have access to a building, but does not have an implanted or carry an RFID tag, to gain access through visual identification. Access control based on the biometric identification of a predefined, specific cat can be implemented via the supplementary access control system 2. After the identification signal is sent in step 82, the supplementary access control system 2, specifically the microcontroller 16, enters a standby state in step 84. The supplementary access control system 2 awakens from this standby state upon the next detection of an approaching cat by sensor 8, as described above with reference to step 70. The supplementary access control system 2 also enters a standby state in step 84 after determining that the cat is not the predefined, specific cat. Fig. 7 shows two exemplary images of two different cats as they can be captured and processed by a supplementary access control system according to an exemplary embodiment of the invention. Fig. 7A shows an image of a first cat 130, which, in the nomenclature used herein, is a predefined, specific cat. The head of the cat 130 is framed in the image by the frame structure 4 of a supplementary access control system 2, as shown, for example, in Fig. 2. A section of the surroundings of the supplementary access control system 2 can be seen as the background between the head of the cat 130 and the frame structure 4. Fig. 7B shows an image of a second cat 130', which, in the nomenclature used herein, is not the predefined, specific cat. Again, the head of cat 130 in the image shown is framed by the frame structure 4, and a section of the surroundings of the supplementary access control system 2 can be seen as the image background. As can be seen from the comparison of the images in Fig. 7A and Fig. 78, images taken with camera 10 of a supplementary access control system 2 can differ in many respects, such as in the lighting, the direction of the cat's gaze (130 or 130'), the blurriness of the image, the brightness of the background, etc. Despite this multitude of degrees of freedom in image composition, it has been found that an artificial intelligence trained with approximately 400 images of a predefined, specific cat and approximately 400 images of other cats can achieve a success rate of greater than 95% or even greater than 99% when analyzing newly taken images to determine whether the cat is the predefined, specific cat. This is even the case when the artificial intelligence is supposed to distinguish between very similar cats, such as the brothers shown in Fig. 7A and Fig. 7B. With reference to Figures 4 and 5 on the one hand, and Figures 6 and 7 on the other, the aspects of determining whether the cat is carrying prey in its mouth and determining whether the cat is a predefined, specific cat have been described separately. It is expressly emphasized again that it is possible for both aspects to be implemented together. The procedures of Figures 4 and 6 can be implemented together in a supplementary access control system. In this case, some steps can be easily performed and used for both aspects. As described in Figures 4 and 6 above, the part of the image processing unit that determines whether the cat is carrying prey in its mouth is implemented in the cloud, while the part that determines whether the cat is a predefined, specific cat is implemented locally. This separation is highly efficient because the artificial intelligence pre-trained for the former determination can be used for a variety of supplementary access controls that work with a variety of RFID-controlled cat flaps. Therefore, it is possible to maintain, update, and manage only one implementation of the artificial intelligence, while making the functionality available to a large number of RFID-controlled cat flaps.The second point is by definition linked to a predefined, specific cat and generally only makes sense in conjunction with a specific RFID-controlled cat flap. Therefore, the reusability of a suitably trained artificial intelligence is limited, and a local implementation may be desirable. However, it is explicitly emphasized that other divisions between cloud and local implementation are possible, and that a complete implementation of the image processing unit in the cloud or locally is also feasible. Furthermore, it is expressly emphasized that the functionality of the supplementary access control system, as described with reference to Figures 2, 3, 4, 5, 6 to 7, can be integrated into an RFID-controlled cat flap, such as the RFID-controlled cat flap of Figure 1. The RFID-controlled cat flap of Figure 1 can be an RFID-controlled cat flap according to an exemplary embodiment of the invention. Further exemplary embodiments are disclosed in the following numbered paragraphs: 1. Supplementary access control system (2) for an RFID-controlled cat flap (100), comprising: a sensor (8) for detecting an approaching cat (130) and a camera (10) for capturing at least one image of the approaching cat (130); an image processing unit (18) coupled to the camera (10) and configured to determine, based on the at least one image, whether the cat (130) is carrying prey (140) in its mouth; a radio frequency signal generator (14) configured to send a jamming signal, wherein the jamming signal is suitable for preventing the RFID-controlled cat flap (100) from opening; and a control unit (20) coupled to the sensor (8), the image processing unit (18), and the radio frequency signal generator (14).wherein the control unit (20) is configured: based on the detection of the approaching cat (130), to cause the radio frequency signal generator (14) to send the jamming signal; and based on the determination by the image processing unit (18) that the cat (130) is not carrying prey (140) in its mouth, to cause the radio frequency signal generator (14) to stop sending the jamming signal. 2. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to paragraph 1, wherein the image processing unit (18) is configured to analyze the at least one image with artificial intelligence to determine whether the cat (130) is carrying prey (140) in its mouth. 3. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to paragraph 2, wherein the artificial intelligence is a neural network, in particular a convolutional neural network,exhibits. 4. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to paragraph 2 or 3, wherein the artificial intelligence is a pre-trained artificial intelligence, wherein the artificial intelligence has been trained, in particular, with at least 200 images of cats without prey in their mouths and at least 200 images of cats with prey in their mouths, which are appropriately annotated. 5. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to one of the preceding paragraphs, wherein the control unit (20) is configured to cause the radio frequency generator (14) to send the jamming signal during predetermined lockout times, irrespective of whether the image processing unit (18) detects that the cat (130) is carrying prey (140) in its mouth or not. 6. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to one of the preceding paragraphs,wherein the radio frequency signal generator (14) is configured to transmit a jamming signal with a fundamental frequency of between 100 kHz and 200 kHz, in particular with a fundamental frequency of between 120 kHz and 140 kHz; and / or wherein the radio frequency signal generator (14) is configured to modulate a random or pseudo-random identification number onto a fundamental oscillation of the jamming signal, or wherein the radio frequency signal generator (14) is configured to modulate a random or pseudo-random bit pattern onto a fundamental oscillation of the jamming signal. 7. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to one of the preceding paragraphs, further comprising: a network connection (22), in particular a wireless network connection, and further in particular a WLAN connection. 8. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to paragraph 7,wherein the supplementary access control system (2) is a distributed system in which the image processing unit (18) or a part of the image processing unit (18) is located remotely from the sensor (8), the camera (10), the radio frequency signal generator (14), and the control unit (20), wherein the supplementary access control system (2) is in particular configured to transmit at least one image to the image processing unit (18) via the network connection (22) and to receive the determination of whether the cat (130) is carrying prey (140) in its mouth via the network connection (22). 9. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to paragraph 7, wherein the control unit (20) and the image processing unit (18) are integrated and wherein the image processing unit (18) is configured10. An installation of an artificial intelligence trained with an extended database to receive a signal via the network connection (22). 11. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to one of the preceding paragraphs, wherein the sensor (8) is a passive infrared sensor, wherein the passive infrared sensor (8) is configured, in particular, to cause the supplementary access control system (2) to switch from a standby state to an operating state based on the detection of the approaching cat (130). 12. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to one of the preceding paragraphs, wherein the sensor (8) is configured as a distance sensor and / or wherein the supplementary access control system (2) has an additional distance sensor (12).which is coupled to the control unit (20); and wherein the sensor (8) or the additional distance sensor (12) provides a signal indicating the distance of the approaching cat (130) and the control unit (20) is configured to control the trigger time for taking the at least one image based on the distance of the approaching cat (130); and / or wherein the sensor (8) or the additional distance sensor (12) is configured to continuously monitor the presence of the approaching cat (130). 12. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to one of the preceding paragraphs, wherein the supplementary access control system (2) is a self-contained operating system; and / or wherein the supplementary access control system (2) has an internal power supply (24), in particular a rechargeable battery,13. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to one of the preceding paragraphs, wherein the supplementary access control system (2) can be attached to the RFID-controlled cat flap (100) or to a building structure, wherein the supplementary access control system (2) in particular has a frame structure (4) that can be inserted into a door opening and / or a window opening and / or a wall opening and / or that can be attached to an RFID-controlled cat flap (100). 14. RFID-controlled cat flap (100), comprising: a mounting device (102) with which the RFID-controlled cat flap (100) can be attached to a building wall and / or to a building door and / or to a building window; a door element (104) that can be pivoted relative to the mounting device (102); a locking mechanism (108, 110; 112, 113, 115),which selectively enables the opening of the pivoting door element (104); an RFID-based controller (116, 118) configured to communicate with an RFID tag (132) carried by a cat (130) approaching the RFID-controlled cat flap (100), and configured to cause the locking mechanism (108, 110; 112, 113, 115) to enable the opening of the pivoting door element (104) when it detects an approved RFID tag (132); and a supplementary access control system (2) according to any of the preceding paragraphs. 15. RFID-controlled cat flap (100) according to paragraph 14, wherein the supplementary access control system (2) is integrated into the RFID-controlled cat flap (100), or wherein the supplementary access control system (2) is designed as an independent system and is arranged in such a way as to the RFID-based control (116, 118) that the camera (10) of the supplementary access control system (2) is able toto take at least one picture of the cat (130) approaching the RFID-controlled cat flap (100). 16. Method for supplementary control of access through an RFID-controlled cat flap (100), wherein the RFID-controlled cat flap (100) is configured to communicate with an RFID tag (132) carried by a cat (130) approaching the RFID-controlled cat flap and to selectively allow the opening of the RFID-controlled cat flap (100), comprising: detecting a cat (130) approaching the RFID-controlled cat flap (100); based on the detection of the cat (130) approaching the RFID-controlled cat flap (100), emitting a radio frequency jamming signal suitable for preventing the opening of the RFID-controlled cat flap (100); taking at least one image of the cat (130) approaching the RFID-controlled cat flap (100); based on the at least one image, determining,whether the cat (130) is carrying prey (140) in its mouth; and, based on the determination that the cat (130) is not carrying prey (140) in its mouth, terminating the transmission of the radio frequency jamming signal. 17. Computer program or computer program product containing program instructions which, when executed on a computer-based device, such as a smartphone or tablet, cause the following steps: providing a graphical user interface; displaying at least one image of an approaching cat (130) taken by a camera (10) of a supplementary access control system (2) for an RFID-controlled cat flap (100) or an RFID-controlled cat flap (100); displaying a determination by an image processing unit (18) of the supplementary access control system (2) or the RFID-controlled cat flap (100).whether the cat (130) in the at least one image is carrying prey (140) in its mouth; receiving user input, wherein the user input includes a user indication as to whether the cat (130) in the at least one image is carrying prey (140) in its mouth, or a user indication as to whether the determination of the image processing unit (18) is correct; and providing the user indication for training the image processing unit (18) with an extended database. Although the invention has been described with reference to exemplary embodiments, it is apparent to a person skilled in the art that various modifications can be made and equivalents used without departing from the scope of the invention. The invention is not intended to be limited by the specific embodiments described. Rather, it encompasses all embodiments covered by the appended claims.

Claims

Supplementary access control system (2) for an RFID-controlled cat flap (100), comprising: a sensor (8) for detecting an approaching cat (130) and a camera (10) for capturing at least one image of the approaching cat (130); an image processing unit (18) coupled to the camera (10) and configured to determine, based on the at least one image, whether the cat (130) is a predefined, specific cat; a radio frequency signal generator (14) configured to send an identification signal, the identification signal being suitable for enabling the opening of the RFID-controlled cat flap (100);and a control unit (20) coupled to the sensor (8), the image processing unit (18) and the radio frequency signal generator (14), wherein the control unit (20) is configured, based on the determination by the image processing unit (18) that the cat (130) is the predefined, specific cat, to cause the radio frequency generator (14) to send the identification signal. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to claim 1, wherein the image processing unit (18) is configured to analyze the at least one image with artificial intelligence to determine whether the cat (130) is the predefined, specific cat. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to claim 2, wherein the artificial intelligence comprises a neural network, in particular a convolutional neural network. Supplementary access control system (2) for an RFID-controlled cat flap (100) according to claim 2 or 3, wherein the artificial intelligence is a pre-trained artificial intelligence, wherein the artificial intelligence has been trained in particular with at least 200 images of the predefined, specific cat and at least 200 images of other cats which are annotated accordingly. RFID-controlled cat flap (100), comprising: a mounting device (102) with which the RFID-controlled cat flap (100) can be attached to a building wall and / or a building door and / or a building window; a door element (104) pivoting relative to the mounting device (102); a locking mechanism (108, 110; 112, 113, 115) which selectively allows the pivoting door element (104) to open; an RFID-based controller (116, 118) configured to communicate with an RFID tag (132) carried by a cat (130) approaching the RFID-controlled cat flap (100), and configured to cause the locking mechanism (108, 110; 112, 113, 115) to allow the pivoting door element (104) to open when it a permitted RFID tag (132) detected; and a complementary access control system (2) according to any one of claims 1 to 4. RFID-controlled cat flap (100) according to claim 5, wherein the supplementary access control system (2) is integrated into the RFID-controlled cat flap (100), or wherein the supplementary access control system (2) is designed as an independent system and is arranged to the RFID-based control (116, 118) in such a way that the camera (10) of the supplementary access control system (2) is able to capture the at least one image of the cat (130) approaching the RFID-controlled cat flap (100). A method for supplementary access control through an RFID-controlled cat flap (100), wherein the RFID-controlled cat flap (100) is configured to communicate with an RFID tag (132) carried by a cat (130) approaching the RFID-controlled cat flap (100) and to selectively enable the opening of the RFID-controlled cat flap (100), comprising: detecting a cat (130) approaching the RFID-controlled cat flap (100); taking at least one image of the cat (130) approaching the RFID-controlled cat flap (100); based on the at least one image, determining whether the cat (130) is a predefined, specific cat; and based on the determination that the cat (130) is a predefined, specific cat, sending an identification signal suitable for enabling the opening of the RFID-controlled cat flap (100). A computer program or computer program product containing program instructions that, when executed on a computer-based device, such as a smartphone or tablet, cause the following steps: providing a graphical user interface; displaying at least one image of an approaching cat (130) taken by a camera (10) of a supplementary access control system (2) for an RFID-controlled cat flap (100) or an RFID-controlled cat flap (100); displaying a detection by an image processing unit (18) of the supplementary access control system (2) orthe RFID-controlled cat flap (100), whether the cat (130) in the at least one image is a predefined, specific cat; receiving a user input, wherein the user input includes a user indication as to whether the cat (130) in the at least one image is the predefined, specific cat, or a user indication as to whether the determination of the image processing unit (18) is correct; and providing the user indication for training the image processing unit (18) with an extended database.

Images