Bus passenger re-identification control method and system

Abstract

The invention belongs to the technical field of passenger re-identification, and provides a bus passenger re-identification control method and system. The method comprises the following steps: capturing a first video image of a get-on lane area, determining each first passenger having a get-on intention according to the first video image, and determining and collecting a first passenger feature of each first passenger; capturing a second video image of a get-off line area, determining each second passenger having a get-off intention according to the second video image, and collecting a second passenger feature of each second passenger; passenger re-identification is carried out based on the second passenger feature and the first passenger feature so as to identify the second passenger belonging to the same passenger and the corresponding first passenger; and recording the get-on station and the get-off station of the same passenger in an associated manner. According to the invention, accurate re-identification of passengers on the bus can be realized, so that accurate basic data can be provided for bus passenger flow data analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of passenger re-identification, and in particular to a bus passenger re-identification control method and system. Background Art

[0002] With the acceleration of urbanization, public transportation systems are carrying an ever-increasing number of passengers. As a key component of urban transportation, accurate statistics and analysis of passenger flow data are crucial for bus companies' operations and management. By accurately tracking passenger boarding and alighting information at each stop, such as when someone boarded and disembarked at a certain stop, bus companies can optimize route planning, rationalize departure schedules, and improve vehicle dispatch efficiency, effectively improving service quality, meeting public travel needs, and reducing operating costs.

[0003] Currently, public transit passenger flow data primarily relies on card swipe data. Card swipe systems automatically count passengers as they swipe their cards, providing some travel information. However, this data doesn't cover all passengers (such as those who purchase tickets with cash) and can't effectively identify those who haven't swiped their cards or have unusual card swipes. This leads to missing or inaccurate data, significantly compromising the accuracy of current public transit passenger flow statistics and preventing them from providing comprehensive and reliable data support for transit companies.

[0004] Video-based bus passenger flow analysis is another new technology, centered around passenger re-identification. This involves re-identifying boarding and disembarking passengers to determine their boarding and alighting trajectories, thereby analyzing passenger flow patterns. However, existing re-identification technology relies primarily on continuous tracking of passengers throughout their entire journey, which consumes excessive computing power. Furthermore, it lacks optimal selection of passenger features, making re-identification errors prone to occur when a passenger's external features change, resulting in incorrect or unrecoverable boarding and alighting trajectories.

[0005] Therefore, a more efficient and accurate passenger re-identification solution is urgently needed to solve the above-mentioned problems of existing technologies and provide reliable data support for bus operation management. Summary of the Invention

[0006] In this regard, the present invention provides a bus passenger re-identification control method, system, electronic device, computer storage medium and computer program product to solve at least one of the above technical problems.

[0007] In a first aspect, the present invention provides a method for controlling re-identification of bus passengers, comprising the following method steps: capturing a first video image of a boarding line area, determining each first passenger who intends to board the bus based on the first video image, and determining and collecting first passenger features of each first passenger; capturing a second video image of a disembarking line area, determining each second passenger who intends to disembark based on the second video image, and collecting second passenger features of each second passenger; wherein the first passenger features are determined based on the detection success rate of the passenger features in the second video image; performing passenger re-identification based on the second passenger features and the first passenger features to identify a second passenger and a corresponding first passenger belonging to the same passenger; and recording the association between the boarding station and the disembarking station of the same passenger.

[0008] According to a second aspect of the present invention, a re-identification control system for bus passengers is provided, the system comprising a first capture unit, a second capture unit, a re-identification unit, and an association unit; wherein: the first capture unit captures a first video image of the boarding line area, determines each first passenger who intends to board the bus based on the first video image, and determines and collects the first passenger features of each first passenger; the second capture unit captures a second video image of the de-boarding line area, determines each second passenger who intends to de-board the bus based on the second video image, and collects the second passenger features of each second passenger; wherein the first passenger features are determined based on the detection success rate of the passenger features in the second video image; the re-identification unit performs passenger re-identification based on the second passenger features and the first passenger features to identify the second passenger belonging to the same passenger and the corresponding first passenger; the association unit records the association of the boarding station and the de-boarding station of the same passenger.

[0009] In a third aspect of the present invention, an electronic device is provided, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program implements any of the methods described above when executed by the processor.

[0010] According to a fourth aspect of the present invention, a computer storage medium is provided, wherein the computer storage medium stores a computer program executable by a processor to implement any of the methods described above.

[0011] According to a fifth aspect of the present invention, a computer program product is provided, which comprises a computer program executable by a processor to implement any of the methods described above.

[0012] This method uses a camera to capture video images of the boarding and alighting areas, applies target detection and tracking algorithms to determine passenger intent and collect features, selects key features based on the detection success rate, and combines this with a re-identification algorithm to achieve precise matching, ultimately correlating and recording boarding and alighting locations. This effectively addresses the lack of traditional passenger flow statistics, improves passenger identification accuracy, and helps bus companies understand travel patterns, optimize routes and vehicle allocation, and enhance operational efficiency and service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0014] Figure 1 The present invention is a flowchart of a bus passenger re-identification control method disclosed in an embodiment of the present invention.

[0015] Figure 2 The present invention is a schematic structural diagram of a bus passenger re-identification control system disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The following specific embodiments illustrate the implementation of this application. Those familiar with the art can easily understand the other advantages and functions of this application from the content disclosed in this specification. Obviously, the described embodiments are part of the embodiments of this application, but not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0017] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0018] like Figure 1 As shown, an embodiment of the present invention discloses a bus passenger re-identification control method, including the following method steps: S10, capturing a first video image of the boarding line area, determining each first passenger who intends to board the bus based on the first video image, and determining and collecting the first passenger features of each first passenger.

[0019] This invention utilizes a visual solution, including monocular and binocular stereo vision, and employs a dual-line approach to identify passengers' boarding and alighting intentions. Specifically, a boarding zone and corresponding cameras are located near the bus's entrances, while an alighting zone and corresponding cameras are located near the bus's exits. This eliminates the need to track passengers throughout their journey, reducing computing power consumption.

[0020] When a bus stops at a stop, a camera captures a first video image of the boarding line area. The boarding line area is a pre-defined area near the bus's entrance, used to determine whether passengers intend to board the bus. After acquiring the first video image, the system uses a target detection and tracking algorithm to analyze the people in the image and identify the first passengers who intend to board the bus. For example, if a person is detected moving toward the entrance and their movement trajectory matches the direction of boarding, the person is determined to have a clear intention to board the bus.

[0021] After identifying each first-passenger, their first-passenger features are collected, including but not limited to the passenger's hairstyle, clothing color, style, accessories, and belongings, etc., for subsequent passenger re-identification. For example, an unsupervised deep learning model can be used to train a feature extractor to extract robust passenger features (e.g., 512 dimensions).

[0022] S20, capturing a second video image of the disembarkation line area, determining each second passenger who intends to disembark based on the second video image, and collecting second passenger features of each second passenger; wherein the first passenger features are determined based on the detection success rate of the passenger features in the second video image.

[0023] The camera captures a second video image of the alighting zone, which is also pre-set near the exit door. This image is used to determine passengers' intention to alight. The second video image is analyzed, again using object detection and tracking technology, to identify any secondary passengers who intend to alight, such as those moving toward the exit door.

[0024] The first passenger feature is determined based on its detection success rate in the second video image. Specifically, features that are easily detected in the exit video image are prioritized as the first passenger feature, thereby improving the accuracy of subsequent re-identification. For example, if a passenger's distinctive hat is easily identifiable in the exit image, the hat feature in the first video image will be used as the first passenger feature.

[0025] It can be understood that the second passenger feature includes more passenger features than the first passenger feature, but only the passenger features corresponding to the first passenger feature are used during re-identification.

[0026] S30: Re-identify the passenger based on the second passenger characteristics and the first passenger characteristics to identify the second passenger and the corresponding first passenger who belong to the same passenger.

[0027] The collected features of the second passenger are compared and analyzed with those of the first passenger. Using a passenger re-identification algorithm, the second passenger and the corresponding first passenger are identified as belonging to the same passenger. Because passengers in public transportation scenarios may be obscured or have posture changes, the re-identification algorithm needs to consider multiple factors, such as feature similarity and feature stability across different images. By calculating the similarity score between features and setting an appropriate threshold, it is possible to determine which second and first passengers are the same person, thereby achieving accurate matching between passenger images at different stops and at different times.

[0028] S40: Record the association between the boarding station and the alighting station of the same passenger.

[0029] After successfully identifying the second and first passengers as belonging to the same group, the passenger's boarding and alighting stops are associated and recorded. This recorded data can be used by bus companies to analyze passenger travel patterns, optimize bus routes, and rationally allocate vehicle resources. For example, by recording the boarding and alighting stops of a large number of passengers, bus companies can identify areas with concentrated passenger flow on certain routes during specific time periods, thereby increasing the number of departures during these periods and improving bus service quality.

[0030] This method uses a camera to capture video images of the boarding and alighting areas, applies target detection and tracking algorithms to determine passenger intent and collect features, selects key features based on the detection success rate, and combines this with a re-identification algorithm to achieve precise matching, ultimately correlating and recording boarding and alighting locations. This effectively addresses the lack of traditional passenger flow statistics, improves passenger identification accuracy, and helps bus companies understand travel patterns, optimize routes and vehicle allocation, and enhance operational efficiency and service quality.

[0031] As an example, the first passenger feature is determined based on the success rate of passenger feature detection in the second video image, specifically including: determining the first congestion of real-time passengers in the bus, and the median of the historical number of people getting off at subsequent bus stops, and predicting the second congestion of the get-off line area when the bus stops at subsequent bus stops based on the first congestion and the median; when the second congestion is higher than the congestion threshold, determining a height mark based on the shooting angle height of the second video image, and determining each passenger feature on the passenger's body that is higher than the height mark as the first passenger feature.

[0032] Traditional passenger re-identification technologies typically use a fixed feature collection strategy, for example, using a passenger's hairstyle, clothing color, style, accessories, and belongings as features for re-identification. This fixed feature collection strategy, which doesn't account for the dynamic nature of public transportation scenarios, can easily obscure certain passenger features (such as trouser color and messenger bag style) during peak hours when the alighting area is crowded (i.e., when many people alight). This significantly reduces the detection success rate in the second video image. Furthermore, unfiltered collection of all features increases computational costs, and low-stability features can interfere with matching accuracy.

[0033] To this end, the present invention designs a dynamic feature collection strategy to predict the congestion level of the alighting line area of ​​subsequent bus stops by combining dual data. That is, the second congestion level is predicted based on the real-time first congestion level of passengers on the bus and the median value of the historical number of passengers getting off the bus at the subsequent bus stops.

[0034] In public transportation scenarios, relying solely on the real-time initial congestion level of a bus fails to fully account for the specific characteristics of boarding and alighting at each station. For example, even if the bus is not currently crowded at a transfer station, the number of people getting off the bus later (i.e., the number of people in the alighting area) may be high. Simply relying on the median number of alighting passengers at a particular station at subsequent bus stops is also inadequate for unexpected situations such as sudden surges in passenger flow. Therefore, this invention combines these two key data points to more accurately predict the congestion level in the alighting area at subsequent stations.

[0035] The second congestion degree is calculated, for example, as follows: ;in, is the second congestion degree, is the first congestion degree, The median number of people getting off the bus in history. Over time The dynamically changing weight coefficient is determined by the following reinforcement learning algorithm: The first congestion , the median number of people getting off the bus in history , and the error between the second congestion degree predicted at the previous moment and the actual second congestion degree, together constitute the state .

[0036] At every moment , the agent can choose to adjust and The action space is defined as the weight combination within a certain range.

[0037] Setting the reward function If the error between the predicted second congestion degree and the actual value is less than the set threshold, a positive reward is given; otherwise, a negative reward is given. .

[0038] The intelligent agent continuously interacts with the environment, optimizes its strategy based on reward feedback, and uses algorithms such as the Deep Q Network (DQN) to update weights to maximize long-term cumulative rewards.

[0039] When the predicted second congestion level exceeds the pre-set congestion threshold, indicating a high probability of severe occlusion in the alighting area, an appropriate height marker is calculated based on the camera's viewing angle height in the alighting area. This height marker is virtually projected onto the passengers in the boarding area. The area below this height marker is more likely to be occluded in the second video image. Therefore, passenger features above this height marker, such as hairstyles, distinctive hats, and high-mounted backpacks (e.g., backpacks), are identified as primary passenger features because these features have a higher detection success rate in the second video image.

[0040] Through such a dynamic feature screening mechanism, it is possible to actively adapt to complex public transportation scenarios, give priority to collecting and identifying features that are less affected by occlusion and have high stability, and avoid wasting computing resources on low-value features that are easily occluded.

[0041] It is understood that the real-time primary congestion level of passengers on a bus can be calculated by counting the number of passengers getting on and off the bus using cameras installed at boarding and alighting locations, or by installing a panoramic camera inside the bus to capture the number of passengers on board and further assess the congestion level, without specific limitations. The historical number of passengers getting off the bus at a stop can be stored in the bus's memory or retrieved from a bus management system, without specific limitations.

[0042] As an example, the method further includes: determining several third passengers who are traveling companions with the first passenger based on the first video image; then performing passenger re-identification based on the second passenger characteristics and the first passenger characteristics to identify the second passenger and the corresponding first passenger who are the same passenger, including: performing passenger re-identification based on the second passenger characteristics and the first passenger characteristics to obtain a first probability that the second passenger and the corresponding first passenger are the same passenger; if the first probability is lower than a specified threshold and the second probability that the fourth passenger in the second video image and the third passenger are the same passenger is higher than a specified threshold, then: extracting conversation features and / or carried object assistance features of the second passenger and the fourth passenger, and evaluating the probability of traveling companions based on this; and when the probability of traveling companions is higher than the probability threshold, the second passenger and the corresponding first passenger are identified as belonging to the same passenger.

[0043] During the process of bus passenger re-identification, it is common for passengers to change their external features after getting on the bus, such as taking off their hats or changing their coats. This makes it easy for feature matching to fail when relying solely on the external features of a single individual for re-identification, resulting in the inability to accurately associate the boarding and alighting information of the same passenger.

[0044] In response to the above technical problems, the present invention has designed a companion relationship inference mechanism, which is as follows: First, during the boarding stage of the bus, based on the first video image, based on the conversation characteristics with the first passenger (such as face-to-face communication, gesture interaction, etc.) and the assistance characteristics of carrying items (such as carrying and guarding items together), at least one third passenger who is a companion of the first passenger is determined, and the third passenger characteristics of each third passenger are collected.

[0045] During passenger re-identification, according to the re-identification method of the aforementioned embodiment, the second passenger's features are first matched with the recorded first passenger's features to determine a first probability that the second passenger and the corresponding first passenger are the same. If the first probability is below a specified threshold (but still above a lower threshold, indicating that the first probability is in a gray area and cannot be directly excluded), and the second probability that the fourth passenger in the second video image is the same passenger as the recorded third passenger is above the specified threshold, the conversation features and / or the accompanying features of the second and fourth passengers are extracted to assess the probability that the second and fourth passengers are traveling together.

[0046] When the probability of traveling together is higher than the threshold, the second and fourth passengers can be determined to be the same passengers as the first and third passengers identified as traveling together upon boarding. At this point, the second passenger, whose first probability is lower than the specified threshold, is considered the same passenger as the first passenger. This way, even if the first passenger's appearance changes, the second passenger can still be identified as the same passenger based on their relationship with the traveling companions.

[0047] As an example, the probability threshold is determined in the following manner: the number of third passengers is calculated and used as the number of companions, and a preliminary probability threshold is obtained by matching the number of companions; the preliminary probability threshold is negatively correlated with the number of companions; a correction coefficient is obtained by matching the second congestion level, and the preliminary probability threshold is corrected using the correction coefficient to obtain the probability threshold; the correction coefficient is negatively correlated with the second congestion level.

[0048] In public transit passenger re-identification scenarios, a single, fixed probability threshold is difficult to adapt to varying sizes of passenger groups and complex crowded environments, leading to an increased risk of misidentification. When there are a large number of passengers, the second passenger may need to interact with more fourth passengers (i.e., third passengers) within a limited timeframe (i.e., the short period of time required to disembark). This results in more dispersed interactions between the second passenger and their passengers, and a high probability of reduced interaction intensity with each fourth passenger. In this scenario, using a higher probability threshold can easily miss true passenger relationships. Furthermore, in highly crowded scenarios, the probability of accidental interactions between the second passenger and other non-companion passengers increases, further reducing the intensity of interaction between the second passenger and each fourth passenger.

[0049] To address this practical situation, the present invention employs a dynamic probability threshold determination mechanism. Specifically, the number of third passengers is calculated and used as the number of accompanying passengers. The greater the number of accompanying passengers, the greater the probability that the interaction intensity between the second passenger and each fourth passenger will significantly decrease during the disembarkation period. Therefore, a negative correlation is established between the number of accompanying passengers and the initial probability threshold: the greater the number of accompanying passengers, the lower the initial probability threshold.

[0050] For example, when the number of companions is 2, the initial probability threshold is set to 0.9; if the number increases to 5 people, the initial probability threshold is correspondingly lowered to 0.8, thereby lowering the judgment standard in large group scenarios and avoiding missing real companion relationships.

[0051] Next, the initial probability threshold is modified based on the second degree of crowding. The second degree of crowding reflects the density of people in the alighting area. Higher crowding increases the probability of accidental interactions between the second passenger and other non-companion passengers, further reducing the intensity of interaction between the second passenger and each fourth passenger. Therefore, a correction coefficient is set to be negatively correlated with the second degree of crowding. When the second degree of crowding is higher, the correction coefficient becomes smaller, thus revising the initial probability threshold to a lower value.

[0052] For example, in a high-crowding scenario, the correction coefficient is set to 0.9, and the initial probability threshold of 0.8 is further lowered to 0.72 to strictly distinguish between real peer relationships and accidental behaviors; in a low-crowding scenario, the correction coefficient is 1.0, 0.95, 0.96, etc., that is, the degree of lowering the initial probability threshold is reduced or the initial probability threshold is not adjusted to ensure the accuracy and inclusiveness of recognition.

[0053] The above-mentioned dynamic threshold determination method, which combines the number of companions and the congestion of the alighting area, can adapt to different bus operation scenarios. In complex situations such as large groups of passengers and high congestion, it can effectively balance the accuracy and completeness of recognition, reduce the probability of misjudgment and missed judgment, and significantly improve the reliability and effectiveness of passenger re-identification based on companion relationships.

[0054] As an example, the method of determining the first passengers who intend to get on the bus based on the first video image includes: obtaining the body posture of each passenger in the first video image, and extracting the passengers whose inclination angle toward the door is greater than an angle threshold and whose footstep movement trajectory extends toward the door for N consecutive frames based on the body posture, and determining these passengers as the first passengers who intend to get on the bus; wherein N is dynamically determined according to the peak period characteristics of bus operation.

[0055] In public transportation scenarios, traditional methods for determining boarding intention often rely on single visual features (such as a person facing the door), which can be easily affected by behaviors such as passengers stopping or wandering, leading to misjudgment. For example, a passenger asking for directions near a door might be mistakenly identified as boarding.

[0056] To this end, the present invention uses human posture estimation technology to perform three-dimensional posture modeling on the passenger in the first video image, extracts the angle between the torso's central axis and the direction of the door as the tilt angle, and simultaneously tracks the passenger's footsteps in consecutive video frames. The passenger's intention to board the vehicle is determined when the following two conditions are met: (1) Angle threshold constraint: The tilt angle is greater than the angle threshold (e.g., 30°), indicating that the passenger is clearly facing the door in spatial orientation; (2) Temporal continuity constraint: The footsteps' trajectory stably points to the door area in N consecutive frames of images, avoiding misjudging the accidental turning toward the door as an intention to board the vehicle.

[0057] In addition, the N value can be dynamically determined based on the peak period characteristics of bus operations. For example, during peak hours in the morning and evening, the N value can be appropriately increased (e.g., set to 150 frames) to improve judgment accuracy; during off-peak hours, the N value can be reduced (e.g., set to 100 or 50 frames) to improve response speed.

[0058] like Figure 2 As shown, an embodiment of the present invention further provides a bus passenger re-identification control system 100, the system comprising a first capture unit 11, a second capture unit 12, a re-identification unit 13, and an association unit 14; wherein: the first capture unit 11 captures a first video image of the boarding line area, determines each first passenger who intends to board the bus based on the first video image, and determines and collects the first passenger features of each first passenger; the second capture unit 12 captures a second video image of the alighting line area, determines each second passenger who intends to alight the bus based on the second video image, and collects the second passenger features of each second passenger; wherein the first passenger features are determined based on the detection success rate of the passenger features in the second video image; the re-identification unit 13 performs passenger re-identification based on the second passenger features and the first passenger features to identify the second passenger and the corresponding first passenger who belong to the same passenger; the association unit 14 records the association between the boarding station and the alighting station of the same passenger.

[0059] As an example, the first passenger feature is determined based on the success rate of passenger feature detection in the second video image, specifically including: determining the first congestion of real-time passengers in the bus, and the median of the historical number of people getting off at subsequent bus stops, and predicting the second congestion of the get-off line area when the bus stops at subsequent bus stops based on the first congestion and the median; when the second congestion is higher than the congestion threshold, determining a height mark based on the shooting angle height of the second video image, and determining each passenger feature on the passenger's body that is higher than the height mark as the first passenger feature.

[0060] As an example, the re-identification unit 13 is specifically used to: determine several third passengers who are traveling with the first passenger based on the first video image; then perform passenger re-identification based on the second passenger characteristics and the first passenger characteristics to identify the second passenger and the corresponding first passenger who are the same passenger, including: performing passenger re-identification based on the second passenger characteristics and the first passenger characteristics to obtain a first probability that the second passenger and the corresponding first passenger are the same passenger; if the first probability is lower than a specified threshold and the second probability that the fourth passenger in the second video image and the third passenger are the same passenger is higher than a specified threshold, then: extracting conversation features and / or carried object assistance features of the second passenger and the fourth passenger, and evaluating the probability of traveling together based on this; when the probability of traveling together is higher than the probability threshold, the second passenger and the corresponding first passenger are identified as belonging to the same passenger.

[0061] As an example, the probability threshold is determined in the following manner: the number of third passengers is calculated and used as the number of companions, and a preliminary probability threshold is obtained by matching the number of companions; the preliminary probability threshold is negatively correlated with the number of companions; a correction coefficient is obtained by matching the second congestion level, and the preliminary probability threshold is corrected using the correction coefficient to obtain the probability threshold; the correction coefficient is negatively correlated with the second congestion level.

[0062] As an example, the first capture unit 11 is specifically used to: obtain the body posture of each passenger in the first video image, and extract the passengers whose inclination angle toward the door is greater than an angle threshold and whose footstep movement trajectory extends toward the door for N consecutive frames based on the body posture, and determine these passengers as the first passengers who intend to get on the bus; wherein N is dynamically determined according to the peak period characteristics of bus operation.

[0063] An embodiment of the present invention further provides an electronic device comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program implements any of the aforementioned methods when executed by the processor.

[0064] An embodiment of the present invention further provides a computer storage medium storing a computer program that can be executed by a processor to implement any of the methods described above.

[0065] An embodiment of the present invention further provides a computer program product, which includes a computer program that can be executed by a processor to implement any of the methods described above.

[0066] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0067] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and such modifications or substitutions are intended to be within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. A bus passenger re-identification control method, characterized by: The method comprises the following steps: capturing a first video image of a boarding line area, determining first passengers intending to board the bus based on the first video image, and determining and collecting first passenger features of the first passengers; Capture a second video image of the disembarkation line area, determine each second passenger who intends to disembark based on the second video image, and collect second passenger features of each second passenger; wherein the first passenger features are determined based on the detection success rate of the passenger features in the second video image; perform passenger re-identification based on the second passenger features and the first passenger features to identify the second passenger and the corresponding first passenger belonging to the same passenger; and record the association between the boarding station and the alighting station of the same passenger.

2. A bus passenger re-identification control method according to claim 1, characterized in that: The first passenger feature is determined based on the success rate of detecting passenger features in the second video image, specifically including: determining a first congestion degree of passengers in real time in the bus, and a median value of the number of historical passengers getting off at subsequent bus stops, and predicting a second congestion degree of the get-off line area when the bus stops at subsequent bus stops based on the first congestion degree and the median value; when the second congestion degree is higher than a congestion degree threshold, determining a height mark based on the shooting angle height of the second video image, and determining each passenger feature on the passenger's body that is higher than the height mark as the first passenger feature.

3. A bus passenger re-identification control method according to claim 2, characterized in that: The method further includes: determining a number of third passengers who are traveling companions with the first passenger based on the first video image; then performing passenger re-identification based on the second passenger characteristics and the first passenger characteristics to identify a second passenger and a corresponding first passenger who are the same passenger, including: performing passenger re-identification based on the second passenger characteristics and the first passenger characteristics to obtain a first probability that the second passenger and the corresponding first passenger are the same passenger; if the first probability is lower than a specified threshold and the second probability that the fourth passenger in the second video image and the third passenger are the same passenger is higher than a specified threshold, then: extracting conversation features and / or carried object assistance features of the second passenger and the fourth passenger, evaluating and obtaining a probability of traveling companions based on the features, and determining that the second passenger and the corresponding first passenger are the same passenger when the probability of traveling companions is higher than the probability threshold.

4. A bus passenger re-identification control method according to claim 3, characterized in that: The probability threshold is determined by: calculating the number of third passengers and using it as the number of companions, and obtaining a preliminary probability threshold based on the number of companions; The preliminary probability threshold is negatively correlated with the number of said co-travelers; A correction coefficient is obtained according to the second congestion degree matching, and the initial probability threshold is corrected using the correction coefficient to obtain the probability threshold; The correction coefficient is negatively correlated with the second congestion level.

5. The bus passenger re-identification control method according to claim 1, characterized in that: Determining first passengers intending to board the bus based on the first video image includes: obtaining the body posture of each passenger in the first video image, extracting passengers whose inclination angle toward the bus door is greater than an angle threshold and whose footstep movement trajectories extend toward the bus door for N consecutive frames based on the body posture, and determining these passengers as first passengers intending to board the bus; wherein N is dynamically determined based on peak period characteristics of bus operation.

6. A bus passenger re-identification control system, characterized in that: The system includes a first capture unit, a second capture unit, a re-identification unit, and an association unit; wherein: the first capture unit captures a first video image of the boarding line area, determines each first passenger who intends to board the bus based on the first video image, and determines and collects the first passenger features of each first passenger; the second capture unit captures a second video image of the deboarding line area, determines each second passenger who intends to deboard the bus based on the second video image, and collects the second passenger features of each second passenger; wherein the first passenger features are determined based on the detection success rate of the passenger features in the second video image; the re-identification unit performs passenger re-identification based on the second passenger features and the first passenger features to identify the second passenger and the corresponding first passenger who belong to the same passenger; the association unit records the association of the boarding station and the deboarding station of the same passenger.

7. A bus passenger re-identification control system according to claim 6, characterized in that: The first passenger feature is determined based on the success rate of detecting passenger features in the second video image, specifically including: determining a first congestion degree of passengers in real time in the bus, and a median value of the number of historical passengers getting off at subsequent bus stops, and predicting a second congestion degree of the get-off line area when the bus stops at subsequent bus stops based on the first congestion degree and the median value; when the second congestion degree is higher than a congestion degree threshold, determining a height mark based on the shooting angle height of the second video image, and determining each passenger feature on the passenger's body that is higher than the height mark as the first passenger feature.

8. An electronic device, characterized in that: The electronic device includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program implements the method according to any one of claims 1 to 5 when executed by the processor.

9. A computer storage medium, characterized in that: The computer storage medium stores a computer program that can be executed by a processor to implement the method according to any one of claims 1 to 5.

10. A computer program product, characterized in that: The computer program product comprises a computer program that can be executed by a processor to implement the method according to any one of claims 1 to 5.

Images