Subway train operation information pushing method and system

Abstract

The invention is suitable for the technical field of information pushing, and particularly relates to a subway train operation information pushing method and system.The method comprises the steps that video monitoring equipment deployed at a subway turnstile in advance is used for collecting inbound videos of passengers, the passengers are clustered into a plurality of categories, risk personnel are selected, and the risk personnel are sent to the subway turnstile; wherein the risk personnel at least comprise pregnant women and old people, extracting clothing characteristics of the risk personnel, reading ticket buying information of all passengers in a preset riding platform, identifying terminal equipment of all passengers through a preset timestamp, and establishing a corresponding relation between the clothing characteristics and the terminal equipment; and integrating a good exchange module into the riding platform. According to the invention, by generating the guiding task, the risk personnel can be guided to go to the seat offering position, the walking path and time are shortened, and the service experience is optimized, so that the risk personnel can go out more conveniently, the riding experience is more comfortable, and the practicability of operation message pushing is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of information push, and in particular to a method and system for pushing subway train operation information. Background Art

[0002] The push of subway train operation information mainly focuses on train arrival, congestion and transfer reminders. There are still many defects in the entry of special passengers and guiding passengers to give up their seats. For example, when special passengers such as the elderly or pregnant women enter the subway train, some passengers are busy talking, using mobile phones or are blocked, and do not pay attention to the special passengers getting on the train at the first time and are unable to give up their seats in time, causing the elderly, pregnant women and others to stand for a long time in the car and have difficulty moving, especially when the subway train enters and exits the station.

[0003] Therefore, “how to push a seat-giving reminder to the intended passengers after the special passengers enter the station” is the technical problem that the present invention needs to solve. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for pushing subway train operation information to solve the problem raised in the above background technology of "how to push seat-giving reminders to intended passengers after special passengers enter the station".

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for pushing subway train operation information, the method comprising: Using video surveillance equipment pre-deployed at subway gates, we capture passenger entry videos, cluster passengers into several categories, and identify risky individuals, including at least pregnant women and the elderly. We then extract clothing features from these risky individuals and read their ticket purchase information on a pre-set ride platform. Using pre-set timestamps, we identify all passenger terminals and establish a correspondence between clothing features and terminal devices. A goodwill exchange module is integrated into the ride platform to receive registration information uploaded by passengers, identify registrants, define the subway train that will arrive at the station where the risk person is located as a target train, locate registrants on the target train and define them as potential passengers, write their clothing characteristics into a preset template, generate a seat-giving request, and push the seat-giving request to the terminal device of the potential passenger through the correspondence relationship. Receive feedback messages uploaded by potential passengers, select the potential passengers closest to the risk person and define them as target passengers, find out the carriage where the target passenger is located based on the feedback information, obtain the target carriage, integrate the target carriage and clothing features, generate a guidance task, and send the guidance task to the preset terminal, where the preset terminal is the platform staff terminal of the station where the risk person is located.

[0006] Furthermore, the step of clustering passengers into several categories and selecting risky individuals includes: Extracting a snapshot of each passenger entering the station from the entry video, inputting the snapshot into a pre-built motion recognition model, outputting motion features of each passenger, and selecting risk features; Based on the action features, the passengers are clustered into several categories, and the categories corresponding to the risk features are defined as risky persons.

[0007] Furthermore, the step of identifying the terminal devices of all passengers via the preset timestamp and establishing a correspondence between clothing features and terminal devices includes: Establish a mapping between the boarding platform and the terminal device, and record the passenger's ticket scanning time based on the timestamp; From the inbound video, the key frames at the time of scanning the code and checking the ticket are intercepted, the clothing features are extracted, and the corresponding relationship between the clothing features and the terminal device is established.

[0008] Furthermore, the step of writing the clothing features into a preset template, generating a seat-giving request, and pushing the seat-giving request to the terminal device of the intended passenger via the corresponding relationship includes: Find the display devices in the subway train and push the seat-giving request to all the display devices in parallel; A selection button is embedded into the seat-giving request.

[0009] Furthermore, the step of receiving feedback messages uploaded by potential passengers and selecting the potential passengers closest to the risk person includes: From the feedback message, the carriage of each intended passenger is read out and the expected parking location is calculated; Utilize video surveillance equipment deployed in stations to collect image data containing risk personnel, select reference objects, and locate the specific location of risk personnel; Compare the expected parking location and the specific location to find the target passenger.

[0010] Furthermore, the step of selecting the intended passenger closest to the risk person and defining them as the target passenger includes: Based on the target passengers, a multi-level reward mechanism is constructed.

[0011] Furthermore, the method further comprises: In subway trains, goodwill agents are deployed. When a risk person is detected entering a subway station, the goodwill agents in the target train are activated. Generate prompt information and push it to the display device.

[0012] Furthermore, a corresponding module is used to use video surveillance equipment pre-deployed at subway gates to collect passenger entry videos, cluster passengers into several categories, select risk persons, where the risk persons include at least pregnant women and the elderly, extract clothing features of risk persons, read ticket purchase information of risk persons on a preset boarding platform, identify the terminal devices of all passengers based on preset timestamps, and establish a correspondence between clothing features and terminal devices; A push module is used to integrate a goodwill exchange module into the ride platform, receive registration information uploaded by passengers, define registrants, define the subway train that will arrive at the station where the risk person is located as a target train, find registrants on the target train and define them as potential passengers, write their clothing characteristics into a preset template, generate a seat-giving request, and push the seat-giving request to the terminal device of the potential passenger through the corresponding relationship; The sending module is used to receive feedback messages uploaded by potential passengers, select the potential passengers closest to the risk person, and define them as target passengers. Based on the feedback information, the target passenger's carriage is found, the target carriage is obtained, the target carriage and clothing characteristics are integrated, a guidance task is generated, and the guidance task is sent to a preset terminal, where the preset terminal is the platform staff terminal of the station where the risk person is located.

[0013] Furthermore, the corresponding module includes: a selection unit, configured to extract a snapshot of each passenger entering the station from the entry video, input the snapshot into a pre-built motion recognition model, output motion features of each passenger, and select risk features; a definition unit, configured to cluster passengers into a plurality of categories based on the action characteristics, and define the categories corresponding to the risk characteristics as risk persons; A recording unit is used to establish a mapping between the boarding platform and the terminal device, and record the passenger's code scanning and ticket verification time based on the timestamp; The interception unit is used to intercept the key frames at the time of code scanning and ticket verification from the inbound video, extract clothing features, and establish a corresponding relationship between the clothing features and the terminal device.

[0014] Furthermore, the push module includes: A search unit, configured to search for display devices in the subway train and push the seat-giving request to all display devices in parallel; The embedding unit is used to embed a selection button into the seat-giving request.

[0015] Compared with the prior art, the present invention has the following beneficial effects: By selecting risky individuals, we can prioritize the travel safety of special groups, optimize service experience, improve emergency response efficiency, and provide a data basis for operational information push. By extracting clothing features, we can locate the position of risky individuals in the subway station, making it easier for target passengers to quickly identify risky individuals, shorten the walking time of risky individuals, avoid crowding or collisions of risky individuals in crowded areas, and reduce the risk of accidents. By pushing seat-giving requests to intended passengers, we can ensure the safety and comfort of vulnerable groups and cultivate passengers' caring awareness and public responsibility. By determining target passengers and target carriages, we can determine the seat-giving location in advance, further shorten the movement time of risky individuals, and simplify the seat-giving process. By generating guidance tasks, staff can guide risky individuals to the seat-giving location, shortening the walking path and time, optimizing the service experience, making travel more convenient for risky individuals and a more comfortable riding experience, and greatly improving the practicality of operational message push. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a method for pushing subway train operation information provided by an embodiment of the present invention; Figure 2 A flowchart of the first sub-process of the method for pushing subway train operation information provided by an embodiment of the present invention; Figure 3 A flowchart of the second sub-process of the method for pushing subway train operation information provided by an embodiment of the present invention; Figure 4 A flowchart of the third sub-process of the method for pushing subway train operation information provided by an embodiment of the present invention; Figure 5 A block diagram of the subway train operation information push system provided by an embodiment of the present invention; Figure 6 A block diagram of the components of the corresponding modules in the subway train operation information push system provided by an embodiment of the present invention; Figure 7 A block diagram of the push module in the subway train operation information push system provided by an embodiment of the present invention; Figure 8 This is a block diagram of the composition of the sending module in the subway train operation information push system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] In Example 1, Figure 1The implementation process of the method for pushing subway train operation information provided by an embodiment of the present invention is shown and described in detail below: S100: Utilize the video surveillance equipment pre-deployed at the subway gate to collect the passenger entry video, cluster the passengers into several categories, select the risk persons, where the risk persons include at least: pregnant women and the elderly, extract the clothing features of the risk persons, read the ticket purchase information of the risk persons in the preset boarding platform, identify the terminal devices of all passengers through the preset timestamp, and establish the corresponding relationship between the clothing features and the terminal devices.

[0019] Using video surveillance equipment pre-deployed at subway gates, video images of passengers entering the station, i.e., entry videos, are collected. Using algorithms such as human posture recognition, motion feature analysis, and facial recognition, the entry videos are analyzed and passengers are divided into several categories, such as adults, children, pregnant women, passengers with children, and the elderly, from which risky individuals are selected. Risky individuals are passengers who need to accept seats being offered to them, including but not limited to the aforementioned categories of pregnant women, the elderly, and passengers with children.

[0020] Using computer vision algorithms, we extract the clothing features of risky individuals, such as red dresses, black trousers, white shirts, hats, and crutches. When risky individuals use the ride-hailing platform to scan the code to enter the station, we find the clothing features of the corresponding individuals from the entry video, and establish a correspondence between the ride-hailing platform and the clothing features. We also use the binding information of the order on the ride-hailing platform (such as mobile phone number, device ID, or APP login information) to identify the terminal device used by the risky individual, thereby establishing a correspondence between the clothing features and the terminal device. In this process, the timestamp is mainly used to align the code-scanning entry data and the clothing features.

[0021] For example, the backend of the ride-hailing platform shows that a passenger entered the station by scanning the code at gate No. 32 at 15:31. At this time, the video surveillance equipment at the corresponding gate is called to confirm that a passenger has entered the station. The passenger's binding information (such as mobile phone number and device ID) is read out through the order information in the ride-hailing platform; by reading the binding information, the passenger's communication method and receiving channel can be determined, thereby realizing the accurate distribution and targeted push of the seat request message.

[0022] S200: Integrate a goodwill exchange module into the ride platform, receive registration information uploaded by passengers, define registrants, define the subway train that will arrive at the station where the risk person is located as the target train, find out the registrants on the target train, and define them as intended passengers, write clothing features into a preset template, generate a seat-giving request, and push the seat-giving request to the terminal device of the intended passenger through the corresponding relationship.

[0023] A goodwill exchange module is integrated into the ride platform. The goodwill exchange module is used to receive registration information actively uploaded by passengers. The registration information includes: the passenger's age information, the passenger's willingness to give up his seat, and the travel time period. The registered passenger is defined as the registrant. It should be noted that the passenger has completed registration before boarding the ride. According to the timestamp, the entry time of the risk person is determined, the operation schedule of the subway train is queried, and the subway train that first arrives at the station where the risk person is located is defined as the target train.

[0024] Based on each registrant's location data and ticket purchase information, registrants on the target train are defined as potential passengers, meaning they are willing to offer their seats to those at risk. The high-risk individual's clothing characteristics are then incorporated into a pre-set seat-offering request template, generating a seat-offering request message. For example, a seat-offering request might read: "Dear passenger, your carriage is about to welcome a pregnant passenger (wearing a pink dress). If you are willing to offer your seat, we sincerely invite you to participate in this act of kindness." The seat-offering request is then sent to the potential passenger's mobile device.

[0025] S300: Receive feedback messages uploaded by potential passengers, select the potential passengers closest to the risk person, and define them as target passengers. Based on the feedback information, find out the carriage where the target passenger is located, obtain the target carriage, integrate the target carriage and clothing features, generate a guidance task, and send the guidance task to a preset terminal, where the preset terminal is the platform staff terminal of the station where the risk person is located.

[0026] When a passenger receives a request to give up his seat, he edits feedback information according to his own status and uploads it to the goodwill exchange module. The feedback message mainly includes: whether he is willing to give up his seat and the car where the passenger is located; according to the feedback message, all passengers who are willing to give up their seats are found, and it is determined whether the number of passengers who are willing to give up their seats is 1. If so, the passenger who is willing to give up his seat is directly defined as the target passenger; if there are multiple passengers who are willing to give up their seats, the expected parking position of each car of the subway train is determined, and the video surveillance equipment in the station is used to compare the preset reference objects to determine the location of the risk person, and combined with the expected parking position, the local After the train stops, the nearest intended passenger to the risk person is identified and defined as the target passenger; based on the feedback information, the carriage where the target passenger is located is determined and identified as the target carriage, and the target carriage and clothing features are integrated to generate a guidance task containing "guidance object features + precise carriage number"; for example, a guidance task is: "Staff, please note that a pregnant passenger (wearing a blue loose dress and carrying a pink handbag) will board the train at this station. Please guide her to the confirmed passenger who has given up her seat in the middle of carriage No. 3 and assist her to sit down safely"; the guidance task is sent to the preset terminal, where the preset terminal is the staff terminal at the platform where the risk person is located.

[0027] In Example 2, Figure 2 The implementation process of the method for pushing subway train operation information provided by an embodiment of the present invention is shown. The steps of clustering passengers into several categories and selecting risk personnel are described in detail below: S101: From the entry video, a snapshot of each passenger's entry is captured, and the snapshot is input into a pre-built motion recognition model to output motion features of each passenger, and select risk features.

[0028] The key image frames of passengers during the entry process are captured as entry snapshots, and each passenger's entry snapshot is input into a pre-built motion recognition model. The motion recognition model is built based on a deep neural network and can extract dynamic behavioral features such as passengers' body movements and posture changes from the image; it is determined whether the output motion features include: slow walking and crutches, etc. If so, the corresponding motion features are defined as risk features.

[0029] S102: Based on the action features, the passengers are clustered into several categories, and the categories corresponding to the risk features are defined as risky persons.

[0030] According to the movement characteristics of each passenger, the passengers are divided into several categories, such as adults, children, pregnant women, passengers with children and the elderly, and the passengers in the categories corresponding to the risk characteristics are defined as risk persons.

[0031] In Example 3, Figure 2 The implementation process of the method for pushing subway train operation information provided by an embodiment of the present invention is shown. The following details the steps of identifying all passengers' terminal devices by presetting timestamps and establishing a correspondence between clothing features and terminal devices. S103: Establish a mapping between the boarding platform and the terminal device, and record the passenger's code scanning and ticket verification time based on the timestamp.

[0032] When a passenger enters the station and scans the code to check the ticket through the terminal device, the corresponding timestamp information is extracted based on the data fed back by the terminal device and recorded as the passenger's code scanning and ticket checking time.

[0033] S104: From the inbound video, extract the key frame at the time of scanning the code and checking the ticket, extract the clothing features, and establish a corresponding relationship between the clothing features and the terminal device.

[0034] From the entry video, the key frames at the time of code scanning and ticket verification are captured, and the clothing features of the incoming passengers in the key frames are extracted. A mapping relationship is established between the extracted clothing features and the corresponding terminal devices; thereby establishing an association and matching between the visual appearance information and the device data, providing a data basis for subsequent passenger tracking, identity verification or risk identification.

[0035] In Example 4, Figure 3 The implementation process of the method for pushing subway train operation information provided by an embodiment of the present invention is shown. The following details the steps of writing clothing features into a preset template, generating a seat-giving request, and pushing the seat-giving request to the terminal device of the intended passenger via the corresponding relationship. S201: Find out the display devices in the subway train, and push the seat-giving request to all the display devices in parallel.

[0036] Find out the display equipment in the subway train, where the display equipment mainly includes various display screens in the subway train, such as electronic signs on the seat heads and LCD information bars in the carriages, etc. Use the display screens to display requests to give up seats.

[0037] S202: Embed a selection button into the seat-giving request.

[0038] During the display process, selection buttons are embedded in the seat-giving request, where the selection buttons include "Yes" and "No". When a passenger on the subway train clicks "Yes", it means that he or she is willing to give up his or her seat for the risk person. At this time, the passenger's carriage number is sent to the platform staff, who will guide the risk person to that carriage.

[0039] In Example 5, Figure 4The implementation process of the method for pushing subway train operation information provided by an embodiment of the present invention is shown. The following details the steps of receiving feedback messages uploaded by prospective passengers and selecting the prospective passengers closest to the risk person. S301: Read the carriage where each intended passenger is located from the feedback message and calculate the expected parking location.

[0040] From the feedback information, the car of each passenger who is willing to give up his seat is read out, and the expected parking position of the car on the platform when it is about to arrive at the station is calculated, that is, the precise positioning area of ​​the car on the platform; by determining the expected parking position, risk personnel can be guided to the corresponding area in advance before the subway train arrives.

[0041] S302: Utilize the video surveillance equipment deployed in the station to collect image data containing risk persons, select reference objects, and locate the specific location of the risk persons.

[0042] At the station, several reference points are selected and, using video surveillance equipment, the closest reference point to the risk person is found and the specific location is determined. For example, the specific location of a risk person is a rest seat in the middle of the platform.

[0043] S303: Compare the expected parking location and the specific location to find the target passenger.

[0044] Determine the expected parking position of each intended passenger in the carriage, and define the intended passenger closest to the risk person in the expected parking position as the target passenger.

[0045] In Example 6, Figure 4 The implementation process of the method for pushing subway train operation information provided by an embodiment of the present invention is shown. The steps of selecting the intended passenger closest to the risk person and defining them as the target passenger are described in detail below: S304: Building a multi-level reward mechanism based on the target passengers.

[0046] After the target passenger is identified, corresponding rewards are given, where the multi-level reward mechanism consists of multiple reward rules. The specific reward rules may be: the next time the target passenger takes the subway train, he or she can be defined as a priority person. Through the camera in the subway train, the car with less crowd density is found, and the car number of this car is pushed to the priority person to improve the riding experience of the priority person.

[0047] In real life, if the carriage numbers with lower population density are pushed to all passengers, it will not only increase the data processing burden, but may also lead to a new round of congestion. However, pushing the carriage numbers with lower population density only to priority personnel can divert the crowd, alleviate congestion, and improve the riding experience of priority personnel.

[0048] In Example 7, different from Example 1, in this embodiment of the present invention, the method further includes: In subway trains, goodwill agents are deployed. When a risk person is detected entering a subway station, the goodwill agents in the target train are activated. Generate prompt information and push it to the display device.

[0049] "Goodwill seats" are deployed in subway trains to reserve seats for high-risk individuals. The goodwill seats are in regular use during the normal operation of the train and can be used normally by ordinary passengers. The video surveillance equipment at the local iron gate will activate the goodwill seats in the corresponding car of the target train after detecting that the high-risk individuals have entered the station. The activation operation includes setting the corresponding seat to the "goodwill reserved" state and pushing prompt information to the display equipment near the seat or in the car (such as electronic screens, electronic signs on the seat heads, car LCD information bars, etc.). For example, a prompt message is: "This seat will give priority to serving special passengers, please understand and cooperate."

[0050] Figure 5 The following is a structural block diagram of a subway train operation information push system provided by an embodiment of the present invention. The subway train operation information push system 1 includes: Corresponding module 11 is used to use video surveillance equipment pre-deployed at subway gates to collect passenger entry videos, cluster passengers into several categories, select risk individuals, where the risk individuals include at least pregnant women and the elderly, extract clothing features of risk individuals, read all passengers' ticket purchase information on a preset boarding platform, identify all passengers' terminal devices based on preset timestamps, and establish a correspondence between clothing features and terminal devices; The push module 12 is configured to integrate a goodwill exchange module into the ride platform, receive registration information uploaded by passengers, identify registrants, define a subway train that will arrive at the station where the risk person is located as a target train, locate registrants on the target train and define them as potential passengers, write their clothing characteristics into a preset template, generate a seat-giving request, and push the seat-giving request to the terminal device of the potential passenger via the corresponding relationship. The sending module 13 is used to receive feedback messages uploaded by potential passengers, select the potential passengers closest to the risk person, and define them as target passengers. Based on the feedback information, the carriage where the target passenger is located is found, the target carriage is obtained, the target carriage and clothing characteristics are integrated, a guidance task is generated, and the guidance task is sent to a preset terminal, where the preset terminal is the platform staff terminal of the station where the risk person is located.

[0051] Figure 6 The following is a structural block diagram of a subway train operation information push system provided by an embodiment of the present invention. The corresponding module 11 includes: The selection unit 111 is configured to extract a snapshot of each passenger entering the station from the entry video, input the snapshot into a pre-built motion recognition model, output motion features of each passenger, and select risk features; a defining unit 112, configured to cluster passengers into a plurality of categories based on the action characteristics, and define the categories corresponding to the risk characteristics as risk persons; The recording unit 113 is used to establish a mapping between the boarding platform and the terminal device, and record the passenger's code scanning and ticket verification time based on the timestamp; The interception unit 114 is used to intercept the key frames at the time of scanning the code and checking the ticket from the inbound video, extract the clothing features, and establish a corresponding relationship between the clothing features and the terminal device.

[0052] Figure 7 The following is a structural block diagram of a subway train operation information push system according to an embodiment of the present invention. The push module 12 includes: A search unit 121 is configured to search for display devices in the subway train and push the seat-giving request to all display devices in parallel; The embedding unit 122 is configured to embed a selection button into the seat-giving request.

[0053] Figure 8 The following is a structural block diagram of a subway train operation information push system provided by an embodiment of the present invention. The sending module 13 includes: The calculation unit 131 is used to read the carriage of each intended passenger from the feedback message and calculate the expected parking location; The positioning unit 132 is used to use the video surveillance equipment deployed in the station to collect image data containing risk personnel, select reference objects, and locate the specific location of the risk personnel; A comparison unit 133 is used to compare the expected parking location with the specific location to find the target passenger; The construction unit 134 is configured to construct a multi-level reward mechanism based on the target passenger.

[0054] The corresponding module 11 is mainly used to complete step S100, the pushing module 12 is mainly used to complete step S200, and the sending module 13 is mainly used to complete step S200; The selection unit 111 is mainly used to complete step S101, the definition unit 112 is mainly used to complete step S102, the recording unit 113 is mainly used to complete step S103, and the interception unit 114 is mainly used to complete step S104; The searching unit 121 is mainly used to complete step S201, and the embedding unit 122 is mainly used to complete step S202; The inference unit 131 is mainly used to complete step S301 , the positioning unit 132 is mainly used to complete step S302 , the comparison unit 133 is mainly used to complete step S303 , and the construction unit 134 is mainly used to complete step S304 .

[0055] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0056] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0057] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for pushing subway train operation information, characterized in that: The method comprises: Using video surveillance equipment pre-deployed at subway gates, we capture passenger entry videos, cluster passengers into several categories, and identify risk individuals, including at least pregnant women and the elderly. We then extract clothing features from these risk individuals, read all passenger ticket purchase information from a pre-set ride platform, identify all passenger terminals using pre-set timestamps, and establish a correspondence between clothing features and terminal devices. A goodwill exchange module is integrated into the ride platform to receive registration information uploaded by passengers, identify registrants, define the subway train that will arrive at the station where the risk person is located as a target train, locate registrants on the target train and define them as potential passengers, write their clothing characteristics into a preset template, generate a seat-giving request, and push the seat-giving request to the terminal device of the potential passenger through the correspondence relationship. Receive feedback messages uploaded by potential passengers, select the potential passengers closest to the risk person and define them as target passengers, find out the carriage where the target passenger is located based on the feedback information, obtain the target carriage, integrate the target carriage and clothing features, generate a guidance task, and send the guidance task to the preset terminal, where the preset terminal is the platform staff terminal of the station where the risk person is located.

2. The method for pushing subway train operation information according to claim 1, characterized in that: The steps of clustering passengers into several categories and selecting risky individuals include: Extracting a snapshot of each passenger entering the station from the entry video, inputting the snapshot into a pre-built motion recognition model, outputting motion features of each passenger, and selecting risk features; Based on the action features, the passengers are clustered into several categories, and the categories corresponding to the risk features are defined as risky persons.

3. The method for pushing subway train operation information according to claim 1, characterized in that: The steps of identifying the terminal devices of all passengers via the preset timestamp and establishing a correspondence between clothing features and the terminal devices include: Establish a mapping between the boarding platform and the terminal device, and record the passenger's ticket scanning time based on the timestamp; From the inbound video, the key frames at the time of scanning the code and checking the ticket are intercepted, the clothing features are extracted, and the corresponding relationship between the clothing features and the terminal device is established.

4. The method for pushing subway train operation information according to claim 1, characterized in that: The steps of writing the clothing features into a preset template, generating a seat-giving request, and pushing the seat-giving request to a terminal device of the intended passenger via the corresponding relationship include: Find the display devices in the subway train and push the seat-giving request to all the display devices in parallel; A selection button is embedded into the seat-giving request.

5. The method for pushing subway train operation information according to claim 1, characterized in that: The step of receiving feedback messages uploaded by potential passengers and selecting the potential passengers closest to the risk person includes: From the feedback message, the carriage of each intended passenger is read out and the expected parking location is calculated; Utilize video surveillance equipment deployed in stations to collect image data containing risk personnel, select reference objects, and locate the specific location of risk personnel; Compare the expected parking location and the specific location to find the target passenger.

6. The method for pushing subway train operation information according to claim 5, characterized in that: The step of selecting the intended passenger closest to the risk person and defining them as the target passenger includes: Based on the target passengers, a multi-level reward mechanism is constructed.

7. The method for pushing subway train operation information according to claim 4, characterized in that: The method further comprises: In subway trains, goodwill agents are deployed. When a risk person is detected entering a subway station, the goodwill agents in the target train are activated. Generate prompt information and push it to the display device.

8. A subway train operation information push system, characterized in that: The system comprises: The corresponding module is used to use video surveillance equipment pre-deployed at subway gates to collect passenger entry videos, cluster passengers into several categories, select risk persons, where the risk persons include at least pregnant women and the elderly, extract the clothing characteristics of the risk persons, read the ticket purchase information of the risk persons on the preset boarding platform, identify the terminal devices of all passengers based on the preset timestamp, and establish a correspondence between the clothing characteristics and the terminal devices; A push module is used to integrate a goodwill exchange module into the ride platform, receive registration information uploaded by passengers, define registrants, define the subway train that will arrive at the station where the risk person is located as a target train, find registrants on the target train and define them as potential passengers, write their clothing characteristics into a preset template, generate a seat-giving request, and push the seat-giving request to the terminal device of the potential passenger through the corresponding relationship; The sending module is used to receive feedback messages uploaded by potential passengers, select the potential passengers closest to the risk person, and define them as target passengers. Based on the feedback information, the target passenger's carriage is found, the target carriage is obtained, the target carriage and clothing characteristics are integrated, a guidance task is generated, and the guidance task is sent to a preset terminal, where the preset terminal is the platform staff terminal of the station where the risk person is located.

9. The subway train operation information push system according to claim 8, characterized in that: The corresponding modules include: a selection unit, configured to extract a snapshot of each passenger entering the station from the entry video, input the snapshot into a pre-built motion recognition model, output motion features of each passenger, and select risk features; a definition unit, configured to cluster passengers into a plurality of categories based on the action characteristics, and define the categories corresponding to the risk characteristics as risk persons; A recording unit is used to establish a mapping between the boarding platform and the terminal device, and record the passenger's code scanning and ticket verification time based on the timestamp; The interception unit is used to intercept the key frames at the time of code scanning and ticket verification from the inbound video, extract clothing features, and establish a corresponding relationship between the clothing features and the terminal device.

10. The subway train operation information push system according to claim 8, characterized in that: The push module includes: A search unit, configured to search for display devices in the subway train and push the seat-giving request to all display devices in parallel; The embedding unit is used to embed a selection button into the seat-giving request.

Images