Intelligent cabinet control method based on intention recognition, intelligent cabinet and storage medium

Abstract

The application discloses an intelligent cabinet control method based on intention recognition, an intelligent cabinet and a storage medium, relates to the technical field of intention recognition and intelligent cabinet control, and comprises the following steps: identifying whether a user in the vicinity of a smart cabinet in a sensing range of the smart cabinet is an authorized user; if the user in the vicinity of the smart cabinet is an authorized user, obtaining behavior characteristic information and movement trajectory data of the authorized user; generating an open-door intention signal according to the behavior characteristic information and the movement trajectory data; obtaining a current service state of the smart cabinet; and based on the current service state and the open-door intention signal, controlling opening and closing of a cabinet door of a corresponding target compartment of the smart cabinet, so as to complete a delivery process or a pick-up process. The scheme solves the problem that the overall operation process of an existing intelligent cabinet is relatively complicated, realizes safe, efficient and non-sensitized intelligent package delivery and pick-up, and improves the delivery efficiency of a logistics terminal and the user experience.

Description

Technical Field

[0001] This invention relates to the field of intent recognition and smart cabinet control technology, and in particular to a smart cabinet control method based on intent recognition, a smart cabinet, and a storage medium. Background Technology

[0002] Smart parcel lockers, as a key facility for solving the "last mile" delivery problem, have been widely used in communities, schools, office buildings, and other scenarios in recent years. While existing smart parcel lockers have achieved basic storage and retrieval functions, there are still significant technical shortcomings: whether a courier delivers a parcel or a user picks it up, a series of active interactive operations must be performed, such as scanning a QR code, entering a retrieval code, clicking the screen to confirm, or using a mobile app to open the door, making the overall operation process rather cumbersome.

[0003] Therefore, how to achieve seamless operation of smart cabinets has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0004] This invention provides an intelligent cabinet control method based on intent recognition, an intelligent cabinet, and a storage medium to solve the problem of the cumbersome overall operation process of existing intelligent cabinets.

[0005] Firstly, an intention-based intelligent cabinet control method is provided, including: Identify whether a user approaching the smart locker within its sensing range is an authorized user; If the user being approached is the authorized user, then the behavioral characteristics information and movement trajectory data of the authorized user are obtained; Based on the behavioral characteristic information and the movement trajectory data, a door opening intention signal is generated; Obtain the current business status of the smart cabinet; Based on the current business status and the door opening intention signal, the cabinet door of the corresponding target compartment on the smart cabinet is controlled to open and close to complete the delivery process or the pickup process.

[0006] Secondly, a smart cabinet is provided, which includes an identity recognition module, a control module, a sensing module, a status management module, and an execution module; The identity recognition module is used to identify whether a user approaching the smart cabinet within its sensing range is an authorized user; The control module is further configured to, if the approaching user is the authorized user, acquire the behavioral characteristic information and movement trajectory data of the authorized user collected by the perception module; The control module is also used to generate a door opening intention signal based on the behavioral feature information and the movement trajectory data; The control module is also used to obtain the current business status of the smart cabinet maintained by the status management module; The control module is also used to control the execution module to open and close the cabinet door of the corresponding target compartment on the smart cabinet based on the current business status and the door opening intention signal, so as to complete the delivery process or the pickup process.

[0007] Thirdly, a smart cabinet is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described intention-based smart cabinet control method.

[0008] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described intelligent cabinet control method based on intent recognition.

[0009] The beneficial effects of the technical solution provided by this invention are as follows: First, by determining whether the user approaching the smart locker within its sensing range is an authorized user, unauthorized entities are filtered out at the source, effectively eliminating the risk of unauthorized operation. Second, after confirming that the user is an authorized user, by acquiring the authorized user's behavioral characteristics and movement trajectory data, and generating an intention signal to open the door, the system can accurately distinguish between the authorized user's "intentional use" and "unintentional passing" behavior. This effectively avoids the cumbersome process of traditional smart lockers relying on scanning QR codes and entering pickup codes, solving problems such as inconvenience for users, low efficiency of bulk delivery by couriers, and risks of package stacking in specific scenarios. It also reduces the probability of accidental door triggering due to misjudgment of intent. Finally, by combining the acquired current business status of the smart locker, the system controls the automatic opening and closing of the locker doors of the corresponding target compartments, ensuring that the delivery or pickup process is precisely matched with the actual business status of the locker. This guarantees the rigor and reliability of the business process, ultimately achieving safe, efficient, and seamless smart package delivery and pickup, improving the delivery efficiency at the last mile of logistics and the user experience. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For users of ordinary skills in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram of a smart cabinet in one embodiment of the present invention; Figure 2This is a flowchart of an intelligent cabinet control method based on intent recognition in one embodiment of the present invention; Figure 3 This is a schematic diagram of the sensing range of the smart cabinet in one embodiment of the present invention; Figure 4 This is another schematic diagram of the sensing range of the smart cabinet in one embodiment of the present invention; Figure 5 This is another schematic diagram of the sensing range of the smart cabinet in one embodiment of the present invention; Figure 6 This is another schematic diagram of the sensing range of the smart cabinet in one embodiment of the present invention; Figure 7 This is another flowchart of an embodiment of the intelligent cabinet control method based on intent recognition in this invention; Figure 8 This is another flowchart of an embodiment of the intelligent cabinet control method based on intent recognition in this invention; Figure 9 This is another flowchart of an embodiment of the intelligent cabinet control method based on intent recognition in this invention; Figure 10 This is a schematic diagram of an authorized user's body facing the smart cabinet in one embodiment of the present invention; Figure 11 This is another flowchart of an embodiment of the intelligent cabinet control method based on intent recognition in this invention; Figure 12 This is a schematic diagram of an authorized user moving toward the smart cabinet in one embodiment of the present invention; Figure 13 Another flowchart of an intelligent cabinet control method based on intent recognition in one embodiment of the present invention; Figure 14 Another flowchart of an intelligent cabinet control method based on intent recognition in one embodiment of the present invention; Figure 15 This is another schematic diagram of the smart cabinet in one embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by users of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0013] While existing smart lockers have achieved basic storage and retrieval functions, they still have significant technical shortcomings: Firstly, whether it is a courier delivering a package or a user picking it up, a series of active interactive operations must be performed, such as scanning a QR code, entering a retrieval code, clicking on the screen to confirm, or using a mobile app to open the door. The overall operation process is relatively cumbersome, which increases the user's time cost and reduces the overall efficiency of storage and retrieval. Secondly, the user experience is significantly limited in specific scenarios. When users are carrying heavy objects, holding infants, or using umbrellas in the rain, making it difficult to operate electronic devices or the locker screen, traditional active interaction methods become extremely difficult, severely impacting the user experience. Furthermore, when couriers make bulk deliveries, they must repeatedly perform scanning and opening operations, further reducing delivery efficiency and making them more inclined to simply pile packages at the locker's doorway, increasing the potential risk of lost or damaged packages. Thirdly, existing smart lockers lack the ability to accurately determine the user's actual intentions, making it difficult to effectively distinguish between "intentional use" and "unintentional passing by." Therefore, the automatic door opening function cannot be activated rashly, as this could easily lead to accidental triggering and other safety issues, resulting in significant security risks. Addressing these issues, how to achieve seamless operation of smart lockers while ensuring user safety has become a pressing technical problem to be solved in this field.

[0014] Compared to the closest existing technology, this invention achieves the following technical effects: First, by determining whether the user approaching within the smart locker's sensing range is an authorized user, unauthorized entities are filtered out at the source, effectively eliminating the risk of unauthorized operation. Second, after confirming that the approaching user is an authorized user, by acquiring the authorized user's behavioral characteristics and movement trajectory data, and generating an intention signal to open the door accordingly, it can accurately distinguish between the authorized user's "intentional use" and "unintentional passing by" behavior. This effectively avoids the cumbersome process of traditional smart lockers relying on scanning QR codes and entering pickup codes, solving problems such as inconvenience for users, low efficiency of bulk delivery by couriers, and risks of package stacking in specific scenarios, while also reducing the probability of door mis-triggering due to misjudgment of intent. Finally, by combining the acquired current business status of the smart locker, the locker door of the corresponding target compartment on the smart locker is automatically opened and closed, ensuring that the delivery process or pickup process is accurately matched with the actual business status of the locker, guaranteeing the rigor and reliability of the business process, and ultimately achieving safe, efficient, and seamless smart package delivery and pickup, improving the delivery efficiency of the last mile of logistics and the user experience.

[0015] It should be noted that the smart locker described in this invention includes parcel storage devices with automatic identification, network communication, and remote control capabilities, such as smart express lockers, smart fresh food lockers, smart home parcel lockers, smart office document exchange lockers, smart shared resource requisition lockers, and smart vending machines, etc., which are not limited here. The smart locker typically consists of multiple independently controllable storage compartments (slots), and its hardware system architecture may include, but is not limited to, [examples of such devices]. Figure 1 The components shown include: an identity recognition module 101 (e.g., a Bluetooth beacon identification unit, a UWB identity binding module, etc.), a control module 102, a sensing module 103, a status management module 104 (e.g., a status register, etc.), an execution module 105 (e.g., an electronic lock, etc.), an in-cabin status monitoring module 106 (e.g., a weight sensor, etc.), a communication module 107 (e.g., a 4G module, a 5G module, a Wi-Fi module, etc.), and an alarm device 108 (e.g., a combination of LED status lights and a buzzer alarm). The hardware modules interact in real time via a system bus or other data transmission methods, and through collaborative work, they jointly execute the steps of the intent-based intelligent cabinet control method.

[0016] The control module 102, as the core processing unit of the intelligent cabinet, can be flexibly determined according to the complexity of data processing, power consumption requirements, and cost budget. For example, it can be a central processing unit (CPU), a microcontroller unit (MCU), or a dedicated controller (such as a field-programmable gate array (FPGA), a digital signal processor (DSP), etc.), etc., without limitation here. The sensing module 103 can be obtained by combining non-visual sensing components with privacy protection as the core. Specifically, it can be composed of millimeter-wave radar, depth sensors (e.g., infrared array sensors, ultrasonic distance sensors, etc.), UWB positioning acquisition units, or other non-visual sensing modules, which can be arbitrarily combined to form a data acquisition terminal (i.e., sensing module 103).

[0017] It is worth emphasizing that the core advantages of the non-visual sensing component combination used in the sensing module 103, with privacy protection as its core, are reflected in two aspects: First, in terms of privacy protection, the non-visual sensing mode does not collect sensitive biometric information such as facial features, clothing, or skin color, but only acquires numerical data related to user behavior and spatial location, completely eliminating the risk of privacy leakage from the technical source; second, in terms of environmental adaptability, for example, the operation of components such as millimeter-wave radar and infrared array sensors is not affected by environmental factors such as light intensity, rain, fog, or dust interference, and can achieve stable operation in all weather conditions. These characteristics enable the sensing module 103 to balance high-level privacy protection with industrial-grade operational reliability, achieving an organic unity between technical ethics and practical performance.

[0018] Please refer to the following as well. Figures 2 to 14This paper elaborates on an intent-based smart cabinet control method provided by an embodiment of the present invention. It should be noted that, for ease of explanation of the embodiments of the present invention, the relevant hardware components of the smart cabinet will not be identified in the subsequent method description.

[0019] Example A Firstly, such as Figures 2 to 6 As shown, an intention recognition-based intelligent cabinet control method is provided, which can be applied to applications such as... Figure 1 The following steps are used as an example of the smart cabinet shown: S201: Identify whether a user approaching the smart cabinet within its sensing range is an authorized user; S202: If the user closest to you is an authorized user, then obtain the authorized user's behavioral characteristics and movement trajectory data.

[0020] As an example, the sensing range of the smart cabinet can be a circular area with the cabinet body as the reference point and a radius of a first preset distance (e.g., 2.5 meters); or a rectangular area formed by extending forward and backward by a second preset distance (e.g., 2.5 meters) and to the left and right by a third preset distance (e.g., 1.5 meters) from the cabinet body as the reference point, without limitation here. The sensing range may include a preset interaction area located in front of the cabinet body, which can be a circular area with a radius of a fifth preset distance (e.g., 2.5 meters) extending forward by a fourth preset distance (e.g., 0.5 meters) from the center of the cabinet door; or a rectangular area formed by extending forward and backward by a sixth preset distance (e.g., 1.5 meters) and to the left and right by a seventh preset distance (e.g., 2.0 meters) from the center of the cabinet door as the reference point, without limitation here. It should be understood that the above values ​​and the formed areas are merely examples and do not constitute a limitation of the present invention.

[0021] As mentioned earlier, the two forms of the sensing range and the two forms of the preset interaction area can be combined in pairs to form, for example, Figure 3 , Figure 4 , Figure 5 and Figure 6 The four specific presentation formats shown are: for example, a circular perception range + a circular preset interaction area (corresponding to...). Figure 3 ), circular perception range + rectangular preset interaction area (corresponding to) Figure 4 ), rectangular perception range + circular preset interaction area (corresponding to) Figure 5 ), rectangular perception range + rectangular preset interaction area (corresponding to) Figure 6 (The following is merely an example and does not constitute a limitation on the present invention.)

[0022] It should be noted that the preset distances for extending forward, backward, left, and right as described above may all be different. This is only an example and does not constitute a limitation of the present invention.

[0023] It should also be emphasized that the specific range and shape of the sensing range and the preset interaction area can be adjusted according to the needs of the actual application scenario, including but not limited to changing the distance parameter or adjusting the shape of the area, which is not limited here and does not constitute a limitation of the present invention.

[0024] Authorized users may include, but are not limited to, authorized couriers and authorized pickup users. This article is only for illustrative purposes and does not limit the specific types of authorized users.

[0025] Behavioral feature information refers to the behavioral data generated by authorized users within a preset interaction area, which can be captured by the sensing module and associated with the user's intention, with privacy protection as the core. This data may include, but is not limited to, dwell time and body orientation characteristics. Such data can be collected in real time by sensing units such as infrared array sensors and millimeter-wave radar. The specific sensing methods are not limited in this invention.

[0026] Mobile trajectory data refers to the spatial movement information of an authorized user from entering to leaving a preset interactive area, with privacy protection as the core. This may include, but is not limited to, the direction of movement (e.g., directly towards the cabinet door or diagonally towards the central axis of the cabinet), movement speed (e.g., common walking speed of 0.3-1.5 m / s), and path shape (e.g., approximately straight line or curve). This type of trajectory data can be collected by millimeter-wave radar, UWB positioning unit, or other non-visual positioning devices, and the specific implementation method is not limited here.

[0027] As an example, when the identity recognition module detects someone entering the smart locker's sensing range, it first marks the person as approaching a user. Then, it verifies the user's identity, determining if they are an authorized user. Only after being determined to be an authorized user does the sensing module detect whether the authorized user has entered a preset interaction area. Upon entering the preset interaction area, it collects the authorized user's behavioral characteristics and movement trajectory data within that area. This effectively avoids the collection of large amounts of data from irrelevant passersby or unauthorized entities, saving computing and communication resources and reducing privacy risks.

[0028] S203: Generate a door opening intention signal based on behavioral characteristic information and movement trajectory data.

[0029] As an example, an intent analysis model can be used to perform multimodal fusion analysis on behavioral feature information and movement trajectory data to determine whether an authorized user has the intent to deliver or pick up a package, and generate an open door intent signal when a valid intent is confirmed. The intent analysis model can be any applicable implementation, such as a rule-based model based on a weighted fusion algorithm (which enhances the confirmation of the final access intent by introducing a weighted fusion algorithm to make collaborative decisions on various features, rather than using simple conditional "AND" logic, thus achieving more accurate and robust recognition of the user's access intent), a neural network model based on deep learning, or an inference / fusion model based on a large language model, etc., without limitation here. By introducing multiple optional intelligent analysis models, the optimal intent recognition strategy can be selected according to different application scenarios, ensuring recognition accuracy while taking into account system resource efficiency and real-time requirements. In this way, by comprehensively analyzing the behavioral characteristics and movement trajectory data of authorized users through the intent recognition model, the accuracy of intent recognition can be ensured, effectively avoiding the risk of misjudgment. This allows for precise differentiation between the "intentional use" and "unintentional passing" behaviors of authorized users. It effectively avoids the cumbersome process of traditional smart lockers relying on active operations such as scanning QR codes and entering pickup codes, solving problems such as inconvenience for users, low efficiency of bulk delivery by couriers, and risks of package stacking in specific scenarios. It also reduces the probability of locker doors being accidentally triggered due to misjudgment of intent.

[0030] S204: Obtain the current business status of the smart cabinet; S205: Based on the current business status and the door opening intention signal, control the opening and closing of the cabinet door of the corresponding target compartment on the smart cabinet to complete the delivery process or the pickup process.

[0031] As an example, the current business status refers to the logical operating stage or semantic state of the smart locker at any given moment, which may include, but is not limited to, idle state, delivery in progress state, delivered state, retrieval in progress state, and locked / abnormal state. This current business status can be stored in the state machine record of the state management module in the form of an enumerated value or identifier, and is updated in real time with each locker door operation.

[0032] As an example, the control module can read the state machine records maintained by the state management module before and after receiving the door opening intention signal to obtain the latest current business status. Next, the control module first determines the target compartment number based on the user identifier or task identifier contained in the door opening intention signal. Based on the current business status, it controls the automatic opening of the door of the corresponding target compartment on the smart locker. After delivery or retrieval is completed, it controls the automatic closing of the door of the corresponding target compartment on the smart locker, thus completing the delivery or retrieval process. This ensures that the delivery or retrieval process accurately matches the actual business status of the locker, guaranteeing the rigor and reliability of the business process. Ultimately, it achieves safe, efficient, and seamless smart parcel delivery and retrieval, improving the delivery efficiency at the last mile of logistics and the user experience.

[0033] In summary, the beneficial effects of the technical solution provided by this invention are as follows: First, by determining whether the user approaching the smart locker within its sensing range is an authorized user, unauthorized entities are filtered out at the source, effectively eliminating the risk of unauthorized operation. Second, after confirming the user is an authorized user, by acquiring the authorized user's behavioral characteristics and movement trajectory data, and generating an opening intention signal accordingly, the system can accurately distinguish between the authorized user's "intentional use" and "unintentional passing" behavior. This effectively avoids the cumbersome process of traditional smart lockers relying on scanning QR codes and entering pickup codes, solving problems such as inconvenience for users, low efficiency of bulk delivery by couriers, and risks of package stacking in specific scenarios. It also reduces the probability of accidental door triggering due to misjudgment of intent. Finally, by combining the acquired current business status of the smart locker, the system controls the automatic opening and closing of the locker doors of the corresponding target compartments, ensuring that the delivery or pickup process is precisely matched with the actual business status of the locker. This guarantees the rigor and reliability of the business process, ultimately achieving safe, efficient, and seamless smart package delivery and pickup, improving the delivery efficiency at the last mile of logistics and the user experience.

[0034] In one embodiment, such as Figure 7 As shown, step S201, which involves identifying whether a user approaching the smart cabinet within its sensing range is an authorized user, includes the following steps: S211: Obtain device identification information that is near the user's smart terminal; S212: Match the device identification information with the authorized device identification information in the target authorization list to obtain the matching result; S213: When the matching result is successful, determine the nearest user as the authorized user.

[0035] As an example, a smart terminal refers to a communication device that is close to a user and has a unique device identifier. It may include, but is not limited to, smartphones, wearable devices (e.g., smartwatches) or handheld terminals (e.g., PDA devices (personal digital assistants, which are portable smart devices), etc., without limitation.

[0036] Device identification information is coded information that can uniquely identify the smart terminal, such as Bluetooth MAC address, RFID serial number, UWB tag number or dynamic encryption token, etc., and is not limited here.

[0037] The target authorization list refers to a pre-configured authorization information database used to record the identification information of authorized devices permitted to operate the smart cabinet. This target authorization list can be stored in the smart cabinet's local database or in a cloud database, and can be updated in real time according to usage, which does not constitute a limitation of this invention.

[0038] As an example, an identity recognition module can automatically collect device identification information of smart terminals near users and match it item by item with authorized device identification information in a target authorization list synchronized in the cloud or locally to obtain matching results. When the same device identification information is found in the target authorization list, the match is considered successful, thus identifying the nearby user as an authorized user; when no matching authorized device identification information is found, the match is considered unauthorized, and the nearby user is identified as an unauthorized user. In the specific matching process, string matching algorithms, hash-based fast matching algorithms, or other algorithms suitable for efficient retrieval can be used, without limitation.

[0039] In one specific implementation, when the authorized user is an authorized courier, the identity recognition process can automatically collect the first device identification information (e.g., Bluetooth MAC address or UWB tag number) of the first smart terminal (e.g., the courier's work phone or the PDA device issued by the company) that is close to the user, and match the first device identification information with the first authorized device identification information in the courier authorization list synchronized in the cloud or locally to obtain a first matching result; when the match is successful, the close user is determined to be an authorized courier; when the match fails, the user is determined to be an unauthorized courier.

[0040] For example, when a courier enters the smart locker's sensing range carrying a company-issued PDA device (UWB tag number: SN-202x-0x-001), the identification module automatically retrieves the UWB tag number and compares it with authorized UWB tag numbers in the courier authorization list synchronized in the cloud or locally. After verification, if the UWB tag number is confirmed to exist in the courier authorization list, the courier is identified as an authorized courier. Therefore, in scenarios where the authorized user is also an authorized courier, this invention can achieve automatic identification and contactless identity verification during the delivery process, thereby improving delivery efficiency.

[0041] In another specific implementation, when the authorized user is an authorized pickup user, the identity recognition process can also automatically collect the second device identification information (e.g., Bluetooth MAC address or UWB tag number) of the second smart terminal (e.g., the user's smartphone or bound smartwatch) that is close to the user, and match the second device identification information with the second authorized device identification information in the user authorization list synchronized in the cloud or locally to obtain a second matching result; if the match is successful, the close user is determined to be an authorized pickup user; if the match fails, the user is determined to be an unauthorized pickup user.

[0042] For example, when a user carrying a bound smartphone (UWB tag number: SN-202x-0x-015) enters the smart locker's detection range, the identity recognition module automatically obtains the UWB tag number and compares it with authorized UWB tag numbers in the user authorization list synchronized in the cloud or locally. After checking, it confirms that the UWB tag number exists in the user authorization list, thus identifying the user as an authorized user. Therefore, this invention can automatically complete identity verification when the user approaches the smart locker during the package pickup process, greatly improving the user experience.

[0043] It should be noted that the first and second smart terminals, the first and second device identification information, the courier authorization list and the user authorization list, the first and second authorized device identification information, and the first and second matching results mentioned above are all used to distinguish different application scenarios of delivery and pickup. Their settings are only illustrative examples and do not constitute a limitation on the overall technical solution of this invention.

[0044] In one embodiment, such as Figure 8 As shown, step S203, which involves generating the door opening intention signal based on behavioral characteristic information and movement trajectory data, includes the following steps: S231: Based on behavioral characteristic information, verify whether the authorized user meets the preset stay duration and preset orientation conditions; S232: Based on the movement trajectory data, verify whether the authorized user meets the preset movement trajectory conditions; S233: When an authorized user meets the preset stay duration, preset orientation, and preset movement trajectory conditions, a door opening intention signal is generated.

[0045] As an example, the preset dwell time condition is used to limit the minimum dwell time that authorized users must achieve in the preset interaction area, such as 0.5 seconds, 1.0 seconds or other thresholds adapted to different scenarios, which are not limited here.

[0046] The preset orientation condition is used to ensure that the authorized user's body is facing the smart cabinet door, so as to exclude non-actual use behaviors such as passing by sideways or standing and looking into the distance.

[0047] The preset movement trajectory conditions are used to limit the movement path of authorized users within the preset interaction area to conform to the typical trajectory pattern close to the cabinet, such as moving towards the cabinet door in a basic straight line or using the central axis of the cabinet as the movement target, without making specific restrictions.

[0048] As an example, the control module can call an intent analysis model to analyze behavioral feature information, such as reasoning about dwell time and body orientation characteristics, to determine whether the authorized user meets the preset dwell time and orientation conditions. Then, the control module can analyze movement trajectory data based on the same or another intent analysis model, such as reasoning about movement direction, speed, and path shape, to determine whether the authorized user meets the preset movement trajectory conditions. When the intent analysis model's overall output indicates that the authorized user simultaneously meets all three conditions, a door-opening intent signal is generated. That is, when the intent analysis model outputs that the authorized user meets the preset dwell time, orientation, and movement trajectory conditions, a door-opening intent signal is generated. Therefore, this embodiment achieves accurate identification of user intent, reduces the false positive rate, and effectively avoids the risk of false triggering due to passersby, brief stays, or accidental directional similarities through a triple verification mechanism of preset dwell time, orientation, and movement trajectory conditions.

[0049] In one specific implementation, when the authorized user is an authorized courier, the control module can use a deployed first intent analysis model to judge the first behavioral feature information. For example, it can analyze whether the courier's first dwell time in the preset interaction area meets the first preset dwell time condition and whether the first body facing direction meets the first preset orientation condition. Simultaneously, based on the collected first movement trajectory data, it can determine whether the courier meets the first preset movement trajectory condition. When the courier meets all the above first preset conditions, a first door-opening intent signal is generated. Therefore, by setting corresponding judgment models and condition parameters for the courier, this implementation can improve the efficiency and security of the delivery process while ensuring accurate identification, enabling the courier to achieve continuous delivery without any active operation.

[0050] In another specific implementation, when the authorized user is an authorized pickup user, the control module can use a second intent analysis model to judge the second behavioral feature information. For example, it can analyze whether the pickup user's second dwell time meets the second preset dwell time condition and whether the second body orientation meets the second preset orientation condition; simultaneously, based on the second movement trajectory data, it can determine whether it meets the second preset movement trajectory condition. When the pickup user meets all the above conditions, a second door opening intent signal is generated. Therefore, by setting an independent intent analysis model and condition parameters for the pickup user, this implementation can better match the pickup user's behavioral habits, achieving higher accuracy and stronger on-site adaptability in intent judgment, thereby improving the convenience and interactive experience of the pickup process.

[0051] It should be noted that the first intent analysis model and the second intent analysis model mentioned above can be the same model or different models, and no limitation is made here.

[0052] In one embodiment, such as Figure 9 and Figure 10 As shown, step S231, which involves verifying whether the authorized user meets the preset stay duration and preset orientation conditions based on behavioral feature information, includes the following steps: S2311: Obtain dwell time and body orientation from behavioral characteristic information; S2312: Calculate the angle of orientation based on the body's orientation direction and the normal direction of the smart cabinet door; S2313: When the dwell time is greater than or equal to the preset dwell time and the orientation angle is greater than or equal to the first preset angle threshold, it is determined that the authorized user meets the preset dwell time condition and the preset orientation condition.

[0053] As an example, the dwell time is the continuous dwell time of the authorized user in the preset interactive area set in front of the smart cabinet; the body orientation direction is the direction in which the authorized user's torso is facing, which can be represented by the direction vector in which the authorized user's torso is facing, and can be specifically calculated by millimeter-wave radar point cloud or other non-visual sensors.

[0054] As an example, besides using an intent analysis model to analyze behavioral feature information in real time to determine whether the above conditions are met, a more explicit geometric calculation method can also be used for verification. Specifically, the control module first obtains the dwell time and determines whether it is greater than or equal to a preset dwell time, where the dwell time is the duration from detecting the user approaching the preset interaction area to leaving the preset interaction area. Subsequently, the control module calculates the angle between the user's body orientation and the cabinet door normal direction. For example, as... Figure 10 As shown, a right-handed coordinate system can be used, with the center of the smart cabinet door as the origin O, the normal direction of the cabinet door as the positive X-axis, and the vertical upward direction as the Z-axis. The body's frontal direction is fitted using millimeter-wave radar point cloud data from behavioral feature information, thus obtaining a unit vector of the body's orientation direction, i.e., the body orientation vector. Next, obtain the unit vector of the cabinet door normal direction, i.e., the cabinet door normal vector. ； Finally, the angle between the two vectors is calculated using the dot product formula: ;in, The dot product of two vectors is used. When the dwell time is greater than or equal to a preset threshold, and the calculated orientation angle is greater than or equal to a first preset angle threshold, the authorized user is deemed to meet the preset dwell time and preset orientation conditions. It is evident that this embodiment achieves quantitative analysis of the body's orientation angle through precise spatial geometric calculations, providing a reliable mathematical basis for intent recognition and ensuring that the system can accurately distinguish whether the user is operating directly facing the cabinet door or passing by sideways.

[0055] In one specific implementation, when the authorized user is an authorized courier, the control module reads the first dwell time and the first body orientation direction from the first behavioral feature information collected by the perception module; calculates the first orientation angle based on the first body orientation direction and the normal direction of the smart cabinet door; when the first dwell time is greater than or equal to the first preset dwell time threshold and the first orientation angle is greater than or equal to the first preset angle threshold, it is determined that the authorized courier meets the first preset dwell time condition and the first preset orientation condition.

[0056] For example, the system detects that the courier's first dwell time in the preset interaction area in front of the locker is 0.8 seconds (the first preset dwell time is 0.5 seconds). Simultaneously, analysis of the millimeter-wave radar point cloud data in the first behavioral feature information reveals the courier's first body orientation vector to be (-1,0,0), which, when combined with the locker door normal vector (1,0,0) using the dot product formula, yields an angle of 180° (the first preset angle is 100°). Both criteria are met; therefore, the courier is deemed to meet both the first preset dwell time condition and the first preset orientation condition.

[0057] In another specific implementation, when the authorized user is an authorized pickup user, the control module reads the second dwell time and the second body orientation direction from the second behavioral feature information collected by the sensing module; and calculates the second orientation angle based on the second body orientation direction and the cabinet door normal direction. When the second dwell time is greater than or equal to the second preset dwell time and the second orientation angle is greater than or equal to the second preset angle threshold, it is determined that the authorized pickup user meets the second preset dwell time condition and the second preset orientation condition.

[0058] For example, the system detects that the second dwell time of the user picking up the package in the preset interaction area in front of the locker is 0.6 seconds (the second preset dwell time is 0.5 seconds); simultaneously, through analysis of millimeter-wave radar point cloud data in the second behavioral feature information, the user's second body orientation vector is obtained as (-1,0,0), and the angle between this vector and the locker door normal vector (1,0,0) is calculated using the dot product formula to be equal to 180° (the second preset angle is 100°). Both criteria are met; therefore, the courier is determined to meet the second preset dwell time condition and the second preset orientation condition.

[0059] Note: The body's orientation vector (-1,0,0) and the cabinet door's normal vector (1,0,0) are parallel to each other in opposite directions, with an angle of 180°, which conforms to the actual scenario logic of "facing the cabinet". It should be noted that the above example is only used to illustrate the determination mechanism of this invention and does not constitute a limitation on the scope of protection of this invention.

[0060] In one embodiment, such as Figure 11 and Figure 12 As shown, step S232, which involves verifying whether the authorized user meets the preset movement trajectory conditions based on the movement trajectory data, includes the following steps: S2321: Construct the movement trajectory of authorized users based on movement trajectory data; S2322: Extract the trajectory shape and direction of motion from the movement trajectory; S2323: Calculate trajectory similarity based on trajectory shape and preset ideal trajectory shape; S2324: Calculate the angle of motion based on the direction of motion and the normal direction of the smart cabinet door; S2325: When the trajectory similarity is greater than or equal to the preset trajectory similarity threshold, and the motion angle is greater than or equal to the second preset angle threshold, the movement trajectory is determined to meet the preset movement trajectory conditions.

[0061] As an example, besides real-time analysis of motion trajectory data through an intent analysis model, this invention can also construct the motion trajectory of authorized users based on the motion trajectory data. For example, the trajectory points in the collected motion trajectory data can first undergo coordinate transformation and filtering to enhance trajectory continuity and reduce measurement noise. Then, the trajectory points are connected in chronological order to form a complete temporal path, and a smooth and continuous motion trajectory can be generated through interpolation algorithms. Subsequently, the trajectory shape and direction of motion are extracted from the motion trajectory. For example, the main shape direction and motion trend of the motion trajectory can be determined through statistical analysis methods to obtain the trajectory shape and direction of motion. Next, the trajectory similarity is calculated based on the trajectory shape and a preset ideal trajectory shape. For example, the actual trajectory can be compared with the preset ideal trajectory, and the trajectory similarity is determined by evaluating the matching degree and deviation degree of the trajectory shape. Further, the motion angle is calculated based on the direction of motion and the normal direction of the smart cabinet door. For example, a right-handed coordinate system is established with the center of the cabinet door as the origin O, the vertical upward direction as the Z-axis, and the normal direction of the cabinet door as the positive X-axis direction; the unit vector of the direction of motion, i.e., the direction of motion vector. The unit vector of the cabinet door normal direction obtained is denoted as , which is the cabinet door normal vector. Next, the orientation angle is calculated using the vector dot product formula: ;in This is the dot product of two vectors. Finally, it is determined whether the trajectory similarity is greater than or equal to a preset trajectory similarity threshold, and whether the motion angle is greater than or equal to a second preset angle threshold. When the trajectory similarity is greater than or equal to the preset trajectory similarity threshold, and the motion angle is greater than or equal to the second preset angle threshold, the movement trajectory is determined to meet the preset movement trajectory conditions.

[0062] In one specific implementation, when the authorized user is an authorized courier, firstly, a first movement trajectory is constructed based on the first movement trajectory data, and the first trajectory shape and first movement direction are extracted from the first movement trajectory; then, based on the first trajectory shape and the first preset ideal trajectory shape (such as a linear trajectory that "directly approaches the cabinet"), the first trajectory similarity is calculated. For example, by evaluating the matching degree and deviation degree between the first trajectory shape and the first preset ideal trajectory shape, the first trajectory similarity is obtained as 0.92 (the first preset trajectory similarity threshold is 0.8); according to the first movement direction and the cabinet door normal direction, the first movement angle is calculated. For example, based on the dot product calculation of the first movement direction vector (-1,0,0) and the cabinet door normal direction vector (1,0,0), the first movement angle is obtained as 180° (the first preset movement angle threshold is 100°); since the first trajectory similarity is greater than or equal to the first preset trajectory similarity threshold, and the first movement angle is greater than or equal to the first preset movement angle threshold, it is determined that the courier's first movement trajectory meets the first preset movement trajectory condition.

[0063] In another specific implementation, when the authorized user is an authorized pickup user, firstly, a second movement trajectory is constructed based on the second movement trajectory data, and the second trajectory shape and second movement direction are extracted from the second movement trajectory; then, based on the second trajectory shape and the second preset ideal trajectory shape (such as a linear trajectory that "short-distance direct approach to the cabinet"), the similarity of the second trajectory is calculated. For example, by evaluating the matching degree and deviation degree between the second trajectory shape and the second preset ideal trajectory shape, the similarity of the second trajectory is obtained as 0.89 (the second preset trajectory similarity threshold is 0.75); according to the second movement direction and the cabinet door normal direction, the second movement angle is calculated. For example, based on the dot product calculation of the second movement direction vector (-1,0,0) and the cabinet door normal direction vector (1,0,0), the second movement angle is obtained as 180° (the second preset movement angle threshold is 90°); since the second trajectory similarity is greater than or equal to the second preset trajectory similarity threshold, and the second movement angle is greater than or equal to the second preset movement angle threshold, it is determined that the second movement trajectory of the pickup user meets the second preset movement trajectory condition.

[0064] Note: The motion direction vector (-1,0,0) and the cabinet door normal vector (1,0,0) are parallel to each other in opposite directions, with a motion angle of 180°, which perfectly matches the ideal interaction scenario of "moving towards the cabinet". The trajectory similarity calculation ensures that the trajectory shape is an effective trajectory of "actively approaching the cabinet", thereby improving the accuracy of the access intention determination.

[0065] It should be noted that the above embodiments are merely examples and do not constitute a limitation of the present invention.

[0066] In one embodiment, such as Figure 13As shown, in step S205, when the authorized user is an authorized courier, that is, based on the current business status and the door opening intention signal, the courier controls the opening and closing of the corresponding target compartment door on the smart locker to complete the delivery or pickup process, including the following steps: S251A: When the current business status is idle, based on the door opening intention signal, control the first cabinet door of the corresponding first target compartment on the smart cabinet to open, and switch the idle status to the delivery status; S252A: When the first weight increase in the first target compartment is detected and reaches a stable state, the delivery is determined to be complete; S253A: Start the first timer after delivery is completed; S254A: When the first timer reaches the first preset delay duration, control the first cabinet door of the first target compartment to close and switch the delivery status to the delivery status to complete one delivery process.

[0067] As an example, the control module first reads the current business status to determine whether the smart locker is ready to perform a delivery operation. If the current business status is "idle", the control module, based on the first target compartment number carried in the door opening intention signal, drives the first door of the corresponding first target compartment to open through the execution module, and switches the current business status from "idle" to "delivery in progress".

[0068] After the first cabinet door is opened, the system monitors the weight output by the status monitoring module inside the first target compartment in real time. When the weight increases compared to before the door opened and remains stable within a preset stabilization window (to filter out noise caused by short-term vibrations, user hand operations, or environmental interference, thus effectively avoiding misjudgments due to brief touches, accidental placement, or environmental interference), delivery is considered complete. After delivery, the control module starts a first timer to allow the courier time to organize the items and complete the necessary actions. When the first timer reaches a preset delay, the control module sends a door-closing command to the execution module, causing the execution module to automatically close the first cabinet door and switch the current business status from "delivering" to "delivered."

[0069] For example, the control module confirms the current service status is "idle," then opens the first cabinet door of the first target compartment A01 and switches the current service status to "delivery in progress." During the monitoring process after the first cabinet door opens, the weight sensor detects that the weight inside the compartment increases from 0.1 kg to 2.3 kg, and this weight value remains unchanged within a first preset stable window of 3 consecutive seconds. This indicates that delivery is complete and the first timer is started (e.g., 5 seconds). After the timer expires, the first cabinet door of the first target compartment A01 automatically closes, the service status is switched from "delivery in progress" to "delivered," and a pickup notification is pushed to the bound recipient's smart terminal via the communication module (e.g., "Your package has been safely delivered to the cabinet by the courier; please pick it up promptly.").

[0070] It should be noted that if an abnormal situation is detected (e.g., the first cabinet door remains open for a long time, or the first weight output by the cabin status monitoring module is abnormal), a preset fault handling strategy will be executed. For example, the alarm device will be controlled to issue an audible and visual alarm, a fault alarm will be pushed to the operation and maintenance personnel, or the business status will be rolled back to a safe state so that maintenance and handling can be carried out in a timely manner in abnormal situations.

[0071] In one embodiment, such as Figure 14 As shown, in step S205, when the authorized user is an authorized pickup user, that is, based on the current business status and the door opening intention signal, the door of the corresponding target compartment on the smart cabinet is controlled to open and close to complete the delivery or pickup process, including the following steps: S251B: When the current business status is "delivered", based on the door opening intention signal, control the second cabinet door of the corresponding second target compartment on the smart cabinet to open and switch the "delivered" status to "pick-up" status. S252B: When the weight in the second target compartment decreases and reaches or falls below a preset zero value, the item retrieval is deemed complete. S253B: Start the second timer after the item is retrieved; S254B: When the second timer reaches the second preset delay duration, control the second cabinet door of the second target compartment to close and switch the item retrieval status to the idle status to complete one item retrieval process.

[0072] As an example, the control module first reads the current business status to determine whether the smart locker allows the pickup operation. If the current business status is "delivered", the control module, based on the target compartment number carried in the door opening intention signal, drives the second door of the corresponding second target compartment to open through the execution module, and switches the current business status from "delivered" to "pickup in progress".

[0073] After the second cabinet door is opened, the second weight output by the cabin status monitoring module inside the second target compartment is monitored in real time. When the second weight is detected to be less than or equal to the preset zero value (to eliminate weighing noise) compared to before opening, and remains within the second preset steady-state window, the item retrieval is determined to be complete. After the item retrieval is completed, the control module starts a second timer to allow the retrieval user to complete the departure and necessary operations. When the second timer reaches the second preset delay duration, the control module sends a door closing command to the execution module, causing the execution module to control the second cabinet door to close automatically and switch the current business status from "item retrieval in progress" to "idle state".

[0074] For example, the control module confirms the current service status as "delivered," then opens the second cabinet door of the second target compartment B01 and switches the "delivered status" to "pickup status." During the monitoring process after the door opens, the weight sensor detects that the weight inside the compartment has decreased from 2.3 kg to 0.1 kg (e.g., the preset zero value is 0.2 kg), and remains stable within a second preset steady-state window for 3 consecutive seconds. This indicates that the pickup is complete and a second timer is started (e.g., 5 seconds). After the timer expires, the second cabinet door of the second target compartment B01 automatically closes, switches the "pickup status" to "idle status," and pushes a completion notification (e.g., "You have successfully picked up your package, please confirm.") to the smart terminal of the bound pickup user via the communication module.

[0075] It should be noted that if an abnormal situation is detected (for example, the second cabinet door remains open for a long time, or the second weight output by the cabin status monitoring module is abnormal), a preset fault handling strategy will be executed. This may include controlling the alarm device to issue an audible and visual alarm, pushing an alarm to the maintenance personnel, or rolling back the business status to a safe state, so as to maintain and handle the abnormal situation in a timely manner.

[0076] It should be noted that the first timer and the second timer mentioned above can be the same timer or different timers, and no limitation is made here.

[0077] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0078] Example B Secondly, a smart cabinet is provided, such as... Figure 1 As shown, the smart cabinet includes an identity recognition module 101, a control module 102, a sensing module 103, a status management module 104, and an execution module 105; The identity recognition module 101 is used to identify whether a user approaching the smart cabinet within its sensing range is an authorized user; The control module 102 is further configured to, if the approaching user is the authorized user, acquire the behavioral feature information and movement trajectory data of the authorized user collected by the perception module 103; The control module 102 is also used to generate a door opening intention signal based on the behavioral feature information and the movement trajectory data; The control module 102 is also used to obtain the current business status of the smart cabinet maintained by the status management module 104; The control module 102 is also used to control the execution module 105 to open and close the cabinet door of the corresponding target compartment on the smart cabinet based on the current business status and the door opening intention signal, so as to complete the delivery process or the pickup process.

[0079] In one embodiment, the identity recognition module 101 is further configured to: Obtain the device identification information of the smart terminal that is close to the user; The device identification information is matched with the authorized device identification information in the target authorization list to obtain the matching result; When the matching result is a successful match, the nearest user is determined to be the authorized user.

[0080] In one embodiment, the control module 102 is further configured to: Based on the behavioral characteristic information, verify whether the authorized user meets the preset stay time condition and preset orientation condition; Based on the movement trajectory data, verify whether the authorized user meets the preset movement trajectory conditions; When the authorized user meets the preset stay duration condition, the preset orientation condition, and the preset movement trajectory condition, the door opening intention signal is generated.

[0081] In one embodiment, the control module 102 is further configured to: The dwell time and body orientation are obtained from the behavioral feature information; wherein, the dwell time is the dwell time of the authorized user in the preset interaction area in front of the smart cabinet, and the body orientation is the direction in which the authorized user's torso is facing. Calculate the angle of orientation based on the body's orientation direction and the cabinet door normal direction of the smart cabinet; When the dwell time is greater than or equal to the preset dwell time and the orientation angle is greater than or equal to the first preset angle threshold, it is determined that the authorized user meets the preset dwell time condition and the preset orientation condition.

[0082] In one embodiment, the control module 102 is further configured to: Based on the movement trajectory data, the movement trajectory of the authorized user is constructed; Extract the trajectory shape and direction of motion from the movement trajectory; Based on the trajectory shape and the preset ideal trajectory shape, the trajectory similarity is calculated; Calculate the angle of motion based on the direction of motion and the normal direction of the cabinet door of the smart cabinet; When the trajectory similarity is greater than or equal to a preset trajectory similarity threshold, and the motion angle is greater than or equal to a second preset angle threshold, the movement trajectory is determined to satisfy the preset movement trajectory condition.

[0083] In one embodiment, the authorized user includes an authorized courier; When the authorized user is the authorized courier, the control module 102 is further configured to: When the authorized user is the authorized courier, the process of controlling the opening and closing of the locker door of the corresponding target compartment on the smart locker, based on the current business status and the door opening intention signal, to complete the delivery or pickup process, includes: When the current business state is idle, based on the door opening intention signal, the control execution module 105 opens the first cabinet door of the corresponding first target compartment on the smart cabinet and switches the idle state to the delivery state. When the first weight in the first target compartment increases and reaches a stable state, the delivery is determined to be complete. After delivery is completed, start the first timer; When the first timer reaches the first preset delay duration, the control execution module 105 closes the first cabinet door of the first target compartment and switches the delivery status to the delivered status to complete a delivery process.

[0084] In one embodiment, the authorized user includes an authorized pickup user; When the authorized user is an authorized pickup user, the control module 102 is further configured to: When the current business status is "delivered", based on the door opening intention signal, the control execution module 105 opens the second cabinet door of the corresponding second target compartment on the smart cabinet and switches the "delivered" status to "retrieving". When the weight in the second target compartment decreases and reaches or falls below a preset zero value, the item retrieval is deemed complete. After the item is retrieved, start the second timer; When the second timer reaches the second preset delay duration, the control execution module 105 closes the second cabinet door of the second target compartment and switches the item retrieval status to the idle status to complete one item retrieval process.

[0085] For specific limitations regarding smart cabinets, please refer to the limitations of the intent recognition-based smart cabinet control method mentioned above, which will not be repeated here.

[0086] Example C Thirdly, a smart cabinet is provided. In one embodiment, a computer device is provided, the internal structure of which can be shown in the diagram below. Figure 15 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. When the computer program is executed by the processor, it implements the steps of an intent-based intelligent cabinet control method.

[0087] In one embodiment, a smart cabinet is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the intention-based smart cabinet control method described in the above embodiments, for example... Figure 2 As shown in S201-S205, or Figures 2 to 14 As shown, to avoid repetition, it will not be described again here. Alternatively, the processor executes the computer program to implement the functions of each module / unit in this embodiment of the intelligent cabinet, for example, Figure 1 The functions of the intent recognition-based smart cabinet control method shown are not described again here to avoid repetition.

[0088] Example D Fourthly, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the intelligent cabinet control method based on intent recognition described in the above embodiments, for example, Figure 2 As shown in S201-S205, or Figures 2 to 14 As shown, to avoid repetition, it will not be described again here. Alternatively, when the computer program is executed by the processor, it implements the functions of each module / unit in this embodiment of the intelligent cabinet, for example, Figure 1 The functions of the intelligent cabinet control based on intent recognition, as shown, will not be described again here to avoid repetition. The computer-readable storage medium can be non-volatile or volatile.

[0089] Users skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0090] Those skilled in the art will understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0091] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, users of ordinary skills in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A smart cabinet control method based on intent recognition, characterized in that, include: Identify whether a user approaching the smart locker within its sensing range is an authorized user; If the user being approached is the authorized user, then the behavioral characteristics information and movement trajectory data of the authorized user are obtained; Based on the behavioral characteristic information and the movement trajectory data, a door opening intention signal is generated; Obtain the current business status of the smart cabinet; Based on the current business status and the door opening intention signal, the cabinet door of the corresponding target compartment on the smart cabinet is controlled to open and close to complete the delivery process or the pickup process.

2. The intelligent cabinet control method based on intent recognition as described in claim 1, characterized in that, The process of identifying whether a user approaching the smart cabinet within its sensing range is an authorized user includes: Obtain the device identification information of the smart terminal that is close to the user; The device identification information is matched with the authorized device identification information in the target authorization list to obtain the matching result; When the matching result is a successful match, the nearest user is determined to be the authorized user.

3. The intelligent cabinet control method based on intent recognition as described in claim 1, characterized in that, The step of generating a door-opening intention signal based on the behavioral feature information and the movement trajectory data includes: Based on the behavioral characteristic information, verify whether the authorized user meets the preset stay time condition and preset orientation condition; Based on the movement trajectory data, verify whether the authorized user meets the preset movement trajectory conditions; When the authorized user meets the preset stay duration condition, the preset orientation condition, and the preset movement trajectory condition, the door opening intention signal is generated.

4. The intelligent cabinet control method based on intent recognition as described in claim 3, characterized in that, The step of verifying whether the authorized user meets the preset stay duration and preset orientation conditions based on the behavioral feature information includes: The dwell time and body orientation are obtained from the behavioral feature information; wherein, the dwell time is the dwell time of the authorized user in the preset interaction area in front of the smart cabinet, and the body orientation is the direction in which the authorized user's torso is facing. Calculate the angle of orientation based on the body's orientation direction and the cabinet door normal direction of the smart cabinet; When the dwell time is greater than or equal to the preset dwell time and the orientation angle is greater than or equal to the first preset angle threshold, it is determined that the authorized user meets the preset dwell time condition and the preset orientation condition.

5. The intelligent cabinet control method based on intent recognition as described in claim 3, characterized in that, The step of verifying whether the authorized user meets the preset movement trajectory conditions based on the movement trajectory data includes: Based on the movement trajectory data, the movement trajectory of the authorized user is constructed; Extract the trajectory shape and direction of motion from the movement trajectory; Based on the trajectory shape and the preset ideal trajectory shape, the trajectory similarity is calculated; Calculate the angle of motion based on the direction of motion and the normal direction of the smart cabinet door; When the trajectory similarity is greater than or equal to a preset trajectory similarity threshold, and the motion angle is greater than or equal to a second preset angle threshold, the movement trajectory is determined to satisfy the preset movement trajectory condition.

6. The intelligent cabinet control method based on intent recognition as described in any one of claims 1-5, characterized in that, The authorized users include authorized delivery personnel; When the authorized user is the authorized courier, the process of controlling the opening and closing of the locker door of the corresponding target compartment on the smart locker, based on the current business status and the door opening intention signal, to complete the delivery or pickup process, includes: When the current business status is idle, based on the door opening intention signal, the first cabinet door of the corresponding first target compartment on the smart cabinet is opened, and the idle status is switched to the delivery state; When the first weight in the first target compartment increases and reaches a stable state, the delivery is determined to be complete. After delivery is completed, start the first timer; When the first timer reaches the first preset delay duration, the first cabinet door of the first target compartment is closed, and the delivery status is switched to the delivered status to complete one delivery process.

7. The intelligent cabinet control method based on intent recognition as described in any one of claims 1-5, characterized in that, The authorized users include authorized pickup users; When the authorized user is an authorized pickup user, the process of controlling the opening and closing of the cabinet door of the corresponding target compartment on the smart cabinet based on the current business status and the door opening intention signal to complete the delivery or pickup process includes: When the current business status is "delivered", based on the opening intention signal, the second cabinet door of the corresponding second target compartment on the smart cabinet is opened, and the "delivered" status is switched to "retrieving". When the weight in the second target compartment decreases and reaches or falls below a preset zero value, the item retrieval is deemed complete. After the item is retrieved, start the second timer; When the second timer reaches the second preset delay duration, the second cabinet door of the second target compartment is closed, and the item retrieval status is switched to the idle status to complete one item retrieval process.

8. A smart cabinet, characterized in that, The smart cabinet includes an identity recognition module, a control module, a sensing module, a status management module, and an execution module; The identity recognition module is used to identify whether a user approaching the smart cabinet within its sensing range is an authorized user; The control module is further configured to, if the approaching user is the authorized user, acquire the behavioral characteristic information and movement trajectory data of the authorized user collected by the perception module; The control module is also used to generate a door opening intention signal based on the behavioral feature information and the movement trajectory data; The control module is also used to obtain the current business status of the smart cabinet maintained by the status management module; The control module is also used to control the execution module to open and close the cabinet door of the corresponding target compartment on the smart cabinet based on the current business status and the door opening intention signal, so as to complete the delivery process or the pickup process.

9. A smart cabinet, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent cabinet control method based on intent recognition as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent cabinet control method based on intent recognition as described in any one of claims 1-7.

Images