Bus umbrella stand intelligent management method and system based on passenger getting-off intention recognition

Abstract

This invention relates to an intelligent management method and system for bus umbrella racks based on passenger disembarkation intention recognition. By binding the identity information of passengers storing umbrellas with the assigned umbrella slots, and using a visual tracking system to track and identify target passengers with disembarkation intentions in real time, the dynamic behavior of passengers is linked with the static physical storage of umbrellas. This proactively identifies and reminds passengers of the risk of forgetting their umbrellas at critical moments, fundamentally solving the problems of misplaced umbrellas and forgotten umbrellas, reducing disputes between drivers and passengers caused by lost items, and greatly improving management efficiency and service quality. At the same time, it expands the function of umbrella racks from simple storage to an intelligent service terminal with proactive care and risk warning, breaking through the limitations of the single function of umbrella racks.

Description

Technical Field

[0001] This invention relates to the field of public transportation facility management technology, and more specifically, to a method and system for intelligent management of bus umbrella racks based on passenger disembarkation intention recognition. Background Technology

[0002] Umbrellas are essential for travel in rainy weather. Bus drivers and passengers alike carry umbrellas and place them inside the bus upon boarding. Public transportation such as buses usually have designated umbrella racks to maintain a clean environment. However, most existing bus umbrella racks are traditional mechanical structures, such as hook-type, bucket-type, or wall-mounted racks. While these racks provide basic storage space, passengers are prone to forgetting their belongings when rushing to their stops, leading to property loss and inconvenience.

[0003] Existing technologies include several smart locker solutions, which typically manage the storage and retrieval of items through passwords, QR codes, or biometrics (such as fingerprints) to ensure the security of belongings. However, these solutions have significant shortcomings: 1) They are passive management systems, responding only when passengers actively perform storage or retrieval operations. When passengers forget to take their umbrellas due to rushing off the bus, the system cannot proactively remind them at critical moments, resulting in weak anti-forgetting capabilities; 2) They fail to correlate with passengers' dynamic behaviors (such as their intention to disembark), limiting their level of intelligence; 3) In crowded and dynamic public transportation environments, requiring passengers to perform additional QR code or fingerprint verification is cumbersome and provides a poor user experience. Summary of the Invention

[0004] The technical problem this invention aims to solve is how to proactively prevent umbrellas from being forgotten, and to achieve a convenient, reliable, and intelligent umbrella rack management method.

[0005] This invention provides a method for intelligent management of bus umbrella racks based on passenger disembarkation intention recognition, comprising: S1, when a passenger carries an umbrella onto the bus and arrives at the umbrella rack to store the umbrella, the umbrella rack collects the passenger's identification information, allocates an empty umbrella slot for the passenger to store the umbrella, and binds the identification information to the allocated umbrella slot. S2 uses a visual tracking system installed inside the bus to track and analyze passengers in real time in order to identify target passengers who intend to get off the bus. S3, when a target passenger is identified, obtain the target passenger's identity information; match the target passenger's identity information with the identity identifier information bound in step S1. If the match is successful, proceed to S4; if the match is unsuccessful, return to S2. S4. Determine whether the umbrella compartment linked to the identity information is in an unclaimed state. If yes, determine that the target passenger is at risk of forgetting their umbrella and trigger an umbrella retrieval reminder for the target passenger. If no, return to S2.

[0006] Compared with existing technologies, the method of this application has the following advantages: By binding the identity information of passengers storing umbrellas with the allocated umbrella slots, and by using a visual tracking system to track and identify target passengers with the intention to get off the bus in real time, the dynamic behavior of passengers is linked with the static physical storage of umbrellas. This enables proactive identification and reminders at the critical point when passengers forget their umbrellas (before getting off the bus), fundamentally solving the problems of mistaking umbrellas and forgetting them after getting off the bus. It also reduces disputes between drivers and passengers caused by lost items, greatly improving management efficiency and service quality. At the same time, it expands the function of the umbrella rack from a simple storage device to an intelligent service terminal with proactive care and risk warning, breaking through the limitations of the single function of the umbrella rack.

[0007] In one possible implementation, step S1 specifically includes: S101, when a passenger carries an umbrella onto the bus and stores it at the umbrella stand, the first image acquisition device installed at the umbrella stand captures the passenger's facial image and extracts the biometric features from the facial image to generate the passenger's identity information. S102, allocate vacant umbrella compartments for passengers to store their umbrellas, and control the locking mechanism of the allocated umbrella compartments to lock them. S103, Establish the binding relationship between the identity information and the allocated umbrella compartment; S104 generates an umbrella retrieval identification code based on the identity information and the coding information of the umbrella compartment, and presents the umbrella retrieval identification code to the passenger through the local output device of the umbrella frame.

[0008] Compared with existing technologies, using biometrics as identification information is highly unique and difficult to impersonate, thus improving security. At the same time, generating an umbrella retrieval code provides passengers with another convenient and reliable way to verify their umbrella retrieval, in addition to biometric identification. Furthermore, through the locking mechanism and output of the umbrella retrieval code, passengers are given clear feedback on successful umbrella storage and proof of retrieval, which improves user experience and system credibility.

[0009] In one possible implementation, step S2, which uses a visual tracking system installed inside the bus to detect and identify target passengers on the bus who intend to get off, specifically includes: S201, based on a two-dimensional grid model built inside the bus compartment, continuously tracks passengers through multiple time-synchronized cameras and generates a location trajectory for each tracked passenger; S202, for each tracked passenger, extract multi-dimensional behavioral features that reflect the passenger's intention to get off the vehicle based on their location trajectory; S203, Combining the current vehicle status and the information of the passenger compartment environment, perform a fusion analysis on the multi-dimensional behavioral characteristics to calculate the confidence level of the passenger's intention to get off the vehicle; S204, if the confidence level of the intention to get off the bus exceeds a preset threshold, then the passenger is determined to be the target passenger with the intention to get off the bus.

[0010] Compared with existing technologies, this method establishes a two-dimensional grid model inside the bus, continuously tracks passengers as they board to generate their location trajectories, extracts multi-dimensional behavioral features, and combines these features with the current vehicle status and the information of the carriage environment for feature fusion analysis. This results in a highly robust and accurate method for recognizing passengers' intention to disembark, solving the problem of accurately identifying passengers' intention to disembark in crowded carriages.

[0011] In one possible implementation, generating a location trajectory for each tracked passenger in step S201 specifically includes: When a passenger passes through the boarding door, a unique tracking identifier is assigned to them, and their first grid position after entering the carriage is recorded; At each synchronization point, based on images captured by multiple cameras at different grid locations, the current grid location of the passenger is determined through feature matching and location verification. Connect the passenger's grid location sequence in chronological order to form their location trajectory.

[0012] Compared with existing technologies, the tracking method based on two-dimensional grids and multi-camera collaboration effectively overcomes the limitations of single-camera occlusion and field of view, ensuring the continuity and accuracy of passenger position trajectories, and laying a reliable data foundation for intent analysis.

[0013] In one possible implementation, the multi-dimensional behavioral characteristics include at least two of spatial directional characteristics, intentional dwelling characteristics, and posture and attention characteristics; wherein, the spatial directional characteristics are used to characterize the consistency between the passenger's movement direction and the direction pointing towards the exit door; the intentional dwelling characteristics are used to characterize the duration of the passenger's stay in the preset drop-off area and the regularity of their movement during the stay; the posture and attention characteristics are used to characterize the degree of alignment between the passenger's body orientation and the direction towards the door and whether they initiate a drop-off interaction action.

[0014] Compared with existing technologies, the comprehensive analysis of multi-dimensional features, including spatial directional characteristics, intentional dwell characteristics, and posture and attention characteristics, has changed the defect of easy misjudgment due to single-location triggering, making intention recognition more in line with human behavioral logic and significantly improving the accuracy of judgment.

[0015] In one possible implementation, the spatial directional feature The calculation formula is: ; In the formula, Table of actual displacement vectors The vector pointing from the starting point of the trajectory to the center of the exit door The angle between them; Represents the actual displacement vector The vector pointing from the starting point of the trajectory to the center of the exit door Cosine similarity between them; The time length representing the location trajectory. The preset maximum reasonable speed of passenger movement; Represents the magnitude of the actual displacement vector; The intentional dwell feature The calculation formula is: ; In the formula, This indicates the total time passengers spend within the designated drop-off area. It represents the positional distribution entropy of passengers moving within a bus, used to characterize the degree of disorder in movement; This represents the normalization function, which maps the location distribution entropy to the interval [0, 1]. The calculation of the posture and attention features includes: First, based on the passenger's shoulder and head key points, calculating the cosine of the angle between the passenger's frontal body and facial orientation and the direction of exiting the vehicle door, and then weighted averaging to obtain the posture attention component. Secondly, the action recognition model is used to determine whether the passenger has performed the disembarkation request action within a preset time window, and the interaction action component is assigned accordingly. Finally, using the weighted formula By fusing the features, we can obtain the pose and attention characteristics. ,in, These are the preset weighting coefficients.

[0016] In one possible implementation, step S203 specifically includes: Operational status information representing the vehicle's travel, arrival, or departure status is obtained through the vehicle bus. ; Passenger density information in the carriage is estimated based on visual detection results. , ,in, This indicates the total number of passengers detected. This indicates the approved passenger capacity of the bus. Based on the operational status information With the passenger density information The weights of each of the multi-dimensional behavioral features in the fusion analysis are dynamically adjusted. Then, the confidence level of the intention to get off the vehicle is calculated using the following formula: .

[0017] Compared with existing technologies, by basing information on operational status... With the passenger density information By dynamically adjusting feature weights, the system can intelligently distinguish between behaviors such as "standing by the door due to crowding while the bus is in motion" and "actively intending to get off at the stop," which greatly reduces false alarms and missed alarms, making the system reliable even in real and complex bus operation environments.

[0018] In one possible implementation, the intelligent management method for bus umbrella stands further includes: S5: After the passenger submits the umbrella retrieval identification code or verifies their identity via the image capture device at the umbrella stand, the corresponding umbrella compartment is unlocked, and the passenger retrieves the umbrella.

[0019] A smart management system for bus umbrella stands, used to implement the aforementioned smart management method for bus umbrella stands based on passenger disembarkation intention recognition, includes: The umbrella frame unit has multiple independent umbrella compartments with locking mechanisms and indicator lights, and a first image acquisition device for collecting passenger identity information to perform steps S1 and S4, and manage the locking and unlocking of the locking mechanisms on the umbrella compartments. The passenger disembarkation intention recognition unit is used to perform step S2; The central processing unit is communicatively connected to the umbrella frame unit and the passenger disembarkation intention recognition unit, respectively, and is used to execute the above step S3 and send control commands to the umbrella frame unit according to the comparison results.

[0020] Compared with the prior art, the method of this application has the following advantages: by setting up an umbrella frame unit, a passenger disembarkation intention recognition unit and a central processing unit, the complex task is professionally decomposed, and the functional boundaries of each unit are clear, making the system easy to be modularly installed or upgraded on existing buses.

[0021] In one possible implementation, the umbrella frame unit further includes a display screen for outputting an umbrella identification code; the central processing unit is connected to an in-vehicle voice system. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a smart management method for bus umbrella racks based on passenger disembarkation intention recognition. Detailed Implementation

[0023] First, those skilled in the art should understand that these embodiments are merely used to explain the technical principles of the embodiments of this application and are not intended to limit the scope of protection of the embodiments of this application. Those skilled in the art can make adjustments as needed to adapt to specific application scenarios.

[0024] In the description of the embodiments of this application, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of this application based on the specific circumstances.

[0025] In the embodiments of this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0026] The present application will now be described in further detail with reference to the accompanying drawings and specific embodiments. Specific Implementation Example 1: See Figure 1 As shown in the figure, this application discloses a method for intelligent management of bus umbrella racks based on passenger disembarkation intention recognition, including: S1, when a passenger boards the bus carrying an umbrella and stores it at the umbrella rack, the umbrella rack collects the passenger's identification information, allocates an available umbrella storage space for the passenger, and binds the identification information to the allocated umbrella storage space; specifically including: S101 When a passenger carries an umbrella onto the bus and stores it at the umbrella stand, the first image acquisition device installed at the umbrella stand captures the passenger's facial image and extracts the biometric features from the facial image to generate the passenger's identification information.

[0028] S102, allocate vacant umbrella slots for passengers to store their umbrellas, and control the locking mechanism of the allocated umbrella slots to lock them. In this embodiment, the locking mechanism is an electromagnetic lock. After the passenger puts the umbrella into the allocated umbrella slot, the locking mechanism of the umbrella slot is controlled to perform a locking operation to prevent accidental removal. At the same time, the indicator light of the umbrella slot flashes yellow to indicate "umbrella being stored". After the locking mechanism completes locking, it turns to a solid red light and prompts the passenger that the umbrella has been successfully stored via voice or display screen.

[0029] S103, establish the binding relationship between the identity information and the allocated umbrella compartment.

[0030] S104, an umbrella retrieval identification code is generated based on the identity information and the coding information of the umbrella compartment, and presented to the passenger through the local output device of the umbrella frame. This embodiment of the application uses biometrics as identity information, which is highly unique and difficult to impersonate, thus improving security; at the same time, the generation of the umbrella retrieval code provides passengers with another convenient and reliable way to verify umbrella retrieval besides biometric identification.

[0031] S2, through a visual tracking system installed inside the bus, tracks and analyzes passengers in real time to identify target passengers with the intention to get off; specifically including: S201, based on a two-dimensional mesh model established inside the bus compartment, multiple time-synchronized cameras continuously track passengers and generate a position trajectory for each tracked passenger; in this embodiment, the top-view plane inside the bus compartment is divided into a regular two-dimensional mesh. Specifically, the origin is the center of the bus's rear axle, the length direction of the bus is the X-axis, and the width direction of the bus is the Y-axis. Each mesh has a unique position code. To overcome the limitations of single-camera obstruction and viewing angle, and to ensure the continuity and accuracy of passenger location trajectories, this application embodiment deploys multiple cameras on the roof of the carriage, with overlapping coverage areas between adjacent cameras; specifically including: When a passenger passes through the boarding door, they are detected by a camera at the boarding door and assigned a unique tracking identifier. Their physical characteristics are extracted, and their first grid position after entering the carriage is recorded. At each synchronization point, each camera independently captures images at different grid locations. Through feature matching and location verification, specifically, when multiple cameras detect a target with high feature matching at the same timestamp, in the same grid, or in adjacent grids, the system identifies it as the same tracking identifier, achieving seamless tracking and location fusion across perspectives, and determining the current grid location of the passenger. Each tracking identifier is continuously recorded and connected in chronological order to form the passenger's location trajectory.

[0032] S202, for each tracked passenger, extract multi-dimensional behavioral features reflecting the passenger's intention to get off the vehicle based on their location trajectory; the multi-dimensional behavioral features include at least two of the following: spatial directional characteristics, intentional dwell characteristics, and posture and attention characteristics. The spatial orientation feature is used to characterize the consistency between the passenger's movement direction and the direction pointing towards the exit door; the spatial orientation feature The calculation formula is: ; In the formula, Table of actual displacement vectors The vector pointing from the starting point of the trajectory to the center of the exit door The angle between them; Represents the actual displacement vector The vector pointing from the starting point of the trajectory to the center of the exit door Cosine similarity between them; The time length representing the location trajectory. The preset maximum reasonable speed of passenger movement; Represents the magnitude of the actual displacement vector; The intentional dwell feature is used to characterize the duration of a passenger's stay within a pre-defined drop-off area and the regularity of their movement during that stay; the intentional dwell feature The calculation formula is: ; In the formula, This indicates the total time passengers spend within the designated drop-off area. It represents the positional distribution entropy of passengers moving within a bus, used to characterize the degree of disorder in movement; This represents the normalization function, which maps the location distribution entropy to the interval [0, 1]. The posture and attention features are used to characterize the degree of alignment between the passenger's body orientation and the direction of the vehicle door, as well as whether the passenger initiates an interaction action to get off the vehicle. The calculation of the posture and attention features includes: First, based on the passenger's shoulder and head key points, calculating the cosine of the angle between the passenger's frontal body and facial orientation and the direction of exiting the vehicle door, and then weighted averaging to obtain the posture attention component. Secondly, the action recognition model is used to determine whether the passenger has performed the disembarkation request action within a preset time window, and the interaction action component is assigned accordingly. Finally, using the weighted formula By fusing the features, we can obtain the pose and attention characteristics. ,in, These are the preset weighting coefficients.

[0033] S203, combining the current vehicle status and cabin environment information, perform fusion analysis on the multi-dimensional behavioral characteristics to calculate the confidence level of the passenger's intention to disembark; specifically including: Operational status information representing the vehicle's travel, arrival, or departure status is obtained through the vehicle bus. It is quantified as discrete values ​​{0: in motion, 1: decelerating to enter the station, 2: stopped and opening the door, 3: about to close the door}; Passenger density information in the carriage is estimated based on visual detection results. , ,in, This indicates the total number of passengers detected. This indicates the approved passenger capacity of the bus. Based on the operational status information With the passenger density information The weights of each of the multi-dimensional behavioral features in the fusion analysis are dynamically adjusted. Then, the confidence level of the intention to get off the vehicle is calculated using the following formula: .

[0034] S204, if the confidence level of the intention to get off the bus exceeds a preset threshold, then the passenger is determined to be the target passenger with the intention to get off the bus.

[0035] S3, when a target passenger is identified, obtain the target passenger's identity information; match the target passenger's identity information with the identity identifier information bound in step S1. If the match is successful, proceed to S4; if the match is unsuccessful, return to S2. S4. Determine whether the umbrella compartment linked to the identity information is in an unclaimed state. If yes, determine that the target passenger is at risk of forgetting their umbrella and trigger an umbrella retrieval reminder for the target passenger. If no, return to S2.

[0036] S5: After the passenger submits the umbrella retrieval identification code or verifies their identity via the image capture device at the umbrella stand, the corresponding umbrella compartment is unlocked, and the passenger retrieves the umbrella. Specific Implementation Example 2: A smart management system for bus umbrella stands, used to implement the intelligent management method for bus umbrella stands based on passenger disembarkation intention recognition as described above, includes: The umbrella frame unit has multiple independent umbrella compartments with locking mechanisms and indicator lights, and a first image acquisition device for collecting passenger identity information to execute steps S1 and S4, and manage the locking and unlocking of the locking mechanisms on the umbrella compartments. In this embodiment, the umbrella frame unit is first initialized, and the status of each umbrella compartment is detected. If it is idle, the corresponding indicator light is controlled to be constantly green; if an umbrella is stored and locked, the indicator light is controlled to be constantly red. The passenger disembarkation intention recognition unit is used to perform step S2; The central processing unit is communicatively connected to the umbrella frame unit and the passenger disembarkation intention recognition unit, respectively, and is used to execute the above step S3 and send control commands to the umbrella frame unit according to the comparison results.

[0038] Compared with the prior art, the method of this application has the following advantages: by setting up an umbrella frame unit, a passenger disembarkation intention recognition unit and a central processing unit, the complex task is professionally decomposed, and the functional boundaries of each unit are clear, making the system easy to be modularly installed or upgraded on existing buses. The umbrella frame unit also includes a display screen for outputting an umbrella identification code; the central processing unit is connected to the vehicle-mounted voice system.

[0039] This application deeply integrates and links two originally isolated technical links: static umbrella storage information management and dynamic passenger behavior recognition. By using the key trigger of disembarkation intention recognition, it activates the risk assessment of the umbrella storage status of a specific passenger, thereby realizing a fundamental shift from "passively responding to storage requests" to "actively predicting and preventing the risk of forgetting".

[0040] In the description of the embodiments of this application, it should be noted that the terms "inner" and "outer" and other terms indicating direction or positional relationship are based on the direction or positional relationship shown in the drawings. This is only for the convenience of description and does not indicate or imply that the device or component must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this application.

[0041] In the description of this application, the references to terms such as "an embodiment," "some embodiments," "in this embodiment," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0042] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for intelligent management of bus umbrella racks based on passenger disembarkation intention recognition, characterized in that, include: S1, when a passenger carries an umbrella onto the bus and arrives at the umbrella rack to store the umbrella, the umbrella rack collects the passenger's identification information, allocates an empty umbrella slot for the passenger to store the umbrella, and binds the identification information to the allocated umbrella slot. S2 uses a visual tracking system installed inside the bus to track and analyze passengers in real time in order to identify target passengers who intend to get off the bus. S3, when a target passenger is identified, obtain the identity information of the target passenger; The identity information of the target passenger is matched with the identity identifier information bound in step S1. If the match is successful, proceed to S4; if the match is unsuccessful, return to S2. S4. Determine whether the umbrella compartment linked to the identity information is in an unclaimed state. If yes, determine that the target passenger is at risk of forgetting their umbrella and trigger an umbrella retrieval reminder for the target passenger. If no, return to S2.

2. The intelligent management method for bus umbrella racks based on passenger disembarkation intention recognition as described in claim 1, characterized in that, Step S1 specifically includes: S101, when a passenger carries an umbrella onto the bus and stores it at the umbrella stand, the first image acquisition device installed at the umbrella stand captures the passenger's facial image and extracts the biometric features from the facial image to generate the passenger's identity information. S102, allocate vacant umbrella compartments for passengers to store their umbrellas, and control the locking mechanism of the allocated umbrella compartments to lock them. S103, Establish the binding relationship between the identity information and the allocated umbrella compartment; S104 generates an umbrella retrieval identification code based on the identity information and the coding information of the umbrella compartment, and presents the umbrella retrieval identification code to the passenger through the local output device of the umbrella frame.

3. The intelligent management method for bus umbrella racks based on passenger disembarkation intention recognition as described in claim 1, characterized in that, In step S2, the visual tracking system installed inside the bus is used to detect and identify target passengers on the bus who intend to get off in real time. Specifically, this includes: S201, based on a two-dimensional grid model built inside the bus compartment, continuously tracks passengers through multiple time-synchronized cameras and generates a location trajectory for each tracked passenger; S202, for each tracked passenger, extract multi-dimensional behavioral features that reflect the passenger's intention to get off the vehicle based on their location trajectory; S203, Combining the current vehicle status and the information of the passenger compartment environment, perform a fusion analysis on the multi-dimensional behavioral characteristics to calculate the confidence level of the passenger's intention to get off the vehicle; S204, if the confidence level of the intention to get off the bus exceeds a preset threshold, then the passenger is determined to be the target passenger with the intention to get off the bus.

4. The intelligent management method for bus umbrella racks based on passenger disembarkation intention recognition according to claim 3, characterized in that, The specific steps in step S201 for generating a location trajectory for each tracked passenger include: When a passenger passes through the boarding door, a unique tracking identifier is assigned to them, and their first grid position after entering the carriage is recorded; At each synchronization point, based on images captured by multiple cameras at different grid locations, the current grid location of the passenger is determined through feature matching and location verification. Connect the passenger's grid location sequence in chronological order to form their location trajectory.

5. The intelligent management method for bus umbrella racks based on passenger disembarkation intention recognition according to claim 3, characterized in that, The multi-dimensional behavioral characteristics include at least two of the following: spatial orientation characteristics, intentional dwelling characteristics, and posture and attention characteristics; wherein, the spatial orientation characteristics are used to characterize the consistency between the passenger's movement direction and the direction pointing towards the exit door; the intentional dwelling characteristics are used to characterize the duration of the passenger's stay in the preset exit area and the regularity of the movement during the stay; the posture and attention characteristics are used to characterize the degree of alignment between the passenger's body orientation and the direction of the door and whether the passenger initiates an exit interaction action.

6. The intelligent management method for bus umbrella racks based on passenger disembarkation intention recognition according to claim 5, characterized in that, The spatial directional characteristics The calculation formula is: ； In the formula, Table of actual displacement vectors The vector pointing from the starting point of the trajectory to the center of the exit door The angle between them; Represents the actual displacement vector The vector pointing from the starting point of the trajectory to the center of the exit door Cosine similarity between them; The time length representing the location trajectory. The preset maximum reasonable speed of passenger movement; Represents the magnitude of the actual displacement vector; The intentional dwell feature The calculation formula is: ； In the formula, This indicates the total time passengers spend within the designated drop-off area. It represents the positional distribution entropy of passengers moving within a bus, used to characterize the degree of disorder in movement; This represents the normalization function, which maps the location distribution entropy to the interval [0, 1]. The calculation of the posture and attention features includes: First, based on the passenger's shoulder and head key points, calculating the cosine of the angle between the passenger's frontal body and facial orientation and the direction of exiting the vehicle door, and then weighted averaging to obtain the posture attention component. Secondly, the action recognition model is used to determine whether the passenger has performed the disembarkation request action within a preset time window, and the interaction action component is assigned accordingly. Finally, using the weighted formula By fusing the features, we can obtain the pose and attention characteristics. ,in, These are the preset weighting coefficients.

7. The intelligent management method for bus umbrella racks based on passenger disembarkation intention recognition according to claim 6, characterized in that, Step S203 specifically includes: Operational status information representing the vehicle's travel, arrival, or departure status is obtained through the vehicle bus. ; Passenger density information in the carriage is estimated based on visual detection results. , ,in, This indicates the total number of passengers detected. This indicates the approved passenger capacity of the bus. Based on the operational status information With the passenger density information The weights of each of the multi-dimensional behavioral features in the fusion analysis are dynamically adjusted. Then, the confidence level of the intention to get off the vehicle is calculated using the following formula: 。 8. The intelligent management method for bus umbrella racks based on passenger disembarkation intention recognition according to claim 1, characterized in that, The intelligent management method for bus umbrella holders also includes: S5: After the passenger submits the umbrella retrieval identification code or verifies their identity via the image capture device at the umbrella stand, the corresponding umbrella compartment is unlocked, and the passenger retrieves the umbrella.

9. A smart management system for bus umbrella stands, used to implement the smart management method for bus umbrella stands based on passenger disembarkation intention recognition as described in any one of claims 2-8, characterized in that, include: The umbrella frame unit has multiple independent umbrella compartments with locking mechanisms and indicator lights, and a first image acquisition device for collecting passenger identity information to perform steps S1 and S4, and manage the locking and unlocking of the locking mechanisms on the umbrella compartments. The passenger disembarkation intention recognition unit is used to perform step S2; The central processing unit is communicatively connected to the umbrella frame unit and the passenger disembarkation intention recognition unit, respectively, and is used to execute the above step S3 and send control commands to the umbrella frame unit according to the comparison results.

10. The intelligent management system for bus umbrella stands according to claim 9, characterized in that, The umbrella frame unit also includes a display screen for outputting an umbrella identification code; the central processing unit is connected to the vehicle-mounted voice system.

Images