An intelligent anti-pinch method, device, equipment and medium for a rail transit platform door

Abstract

The application relates to a rail transit platform door intelligent anti-pinch method, device, equipment and medium, the method tracks passengers getting on and off the platform through a machine vision algorithm and captures the dynamic trajectory of the passengers, realizes the capture of the behavior of the passengers through the dynamic trajectory, judges the possibility and danger of platform door pinching according to the current platform door and train door state, and thus realizes train passenger getting on and off protection. Compared with the prior art, the application has the advantages of improving the safety in a specific scene operation environment.

Description

TECHNICAL FIELD

[0001] The present application relates to a rail transit signal control system, in particular to a rail transit platform door intelligent anti-pinch method, device, equipment and medium based on machine vision and passenger trajectory analysis. BACKGROUND

[0002] In the urban rail transit signal system based on vehicle-to-vehicle communication, the control and state collection of the platform door are completed by the wayside target controller, but the wayside target controller can only perceive whether the platform door is in the closed and locked state through infrared sensors and other devices and provide related alarms. Once a passenger is pinched during getting on and off the train, the operation center can only know that the platform door cannot be closed, and cannot timely find the specific reason, which needs to be found by the station staff manually, and then the problem is solved manually according to the process. However, in the current unmanned train control system, the driver is cancelled, and once the station staff operates incorrectly, the platform door will be shielded and cut off from the system, and there is no driver to confirm the state of the platform door, which may cause danger when the platform door pinches a person.

[0003] With the development of machine vision, cameras are used to detect foreign matters between the platform door and the train door. Compared with the previous infrared sensors, the camera can identify the type of foreign matters and perform basic detection on the reason why the platform door cannot be closed. The existing visual sensor uses a 3D visual detection method, that is, only the target object in a single video frame is identified, and the same object in different video frames is not logically associated, which is helpless for some scenes that need logical judgment. It cannot accurately detect whether the object pinched by the platform door is associated with a passenger, for example, a passenger's bag is pinched, which is currently judged as pinching a person, but the existing visual perception scheme can only find that foreign matters cause the platform door to be unable to close, and the running efficiency of the rail transit system will be affected if similar situations are always judged as platform door pinching a person.

[0004] After searching, Chinese Patent Publication No. CN110091879A discloses a platform door anti-pinch detection system and control method based on image recognition, specifically discloses that a laser emitter, a projection plate and an image recognition unit are correspondingly arranged, the laser emitter emits visible laser to the projection plate, and the image recognition unit correspondingly identifies the number of detection spots, and by comparing the number of spots with the number of working laser emitters, it can be effectively determined whether a person or object is pinched between the platform door and the train. However, the existing patent cannot accurately detect whether the object pinched by the platform door is associated with a passenger, and the existing image recognition can only find that foreign matters cause the platform door to be unable to close. Therefore, how to protect the entire process of passengers getting on and off the train becomes a technical problem to be solved. SUMMARY

[0005] The present application aims to overcome the defects of the prior art and provide a rail transit platform door intelligent anti-pinch method, device, equipment and medium.

[0006] The object of the present application can be achieved by the following technical solutions:

[0007] According to a first aspect of the present application, a rail transit platform door intelligent anti-pinch method is provided, which tracks passengers boarding and alighting the platform by a machine vision algorithm and captures their dynamic trajectories, captures their behavior through dynamic trajectories, judges the possibility and danger of platform door pinching by combining the current platform door and train door state, and thus realizes train boarding and alighting protection.

[0008] As a preferred technical solution, the method uses a visual perception device to monitor the passenger boarding and alighting area, and the monitoring range needs to cover the platform door and the area near the train door inside the train at the same time in the scenario where the platform door and the train door are opened at the same time.

[0009] As a preferred technical solution, the method identifies and tracks passengers in the visual perception area, and assigns an identification code to the passengers, wherein the visual perception area is the boarding and alighting area, including the platform and the inside of the train.

[0010] As a preferred technical solution, the identification code is divided into multiple levels.

[0011] If the passenger in the visual perception area changes from an overlapping individual to an independent individual, the previous identification code is retained and a next level identification code is added.

[0012] If the passenger in the visual perception area changes from an independent individual to an overlapping individual, the previous identification code is retained and a next level identification code is added.

[0013] If the passengers in the visual perception area do not separate, the overlapping target is determined to be the same target object, which has one identification code.

[0014] As a preferred technical solution, the method records the trajectory of each passenger and associates it with the identification code, wherein the starting point of the trajectory is when it enters the visual perception area and the ending point is when it leaves the visual perception area.

[0015] As a preferred technical solution, for an object with multiple identification codes, the trajectory is spliced according to the order of the identification codes to ensure the continuity of the trajectory until it leaves the perception area.

[0016] For each level of identification code, the uniqueness of the identification code needs to be ensured, and finally when a target object leaves the perception area, multiple trajectories are allowed to be formed.

[0017] A unique identification code is set for each trajectory.

[0018] As a preferred technical solution, when the trajectory display passenger leaves the visual perception area, the identification code and the historical trajectory are emptied, and when it reenters, a new identification code is assigned.

[0019] As a preferred technical solution, the method forms a complete behavior logic scene of the passenger getting on and off the vehicle through passenger trajectory analysis.

[0020] As a preferred technical solution, when the real-time trajectory displays that the passenger stops for a set time, it is identified as a passenger stop point;

[0021] The area that may cause the platform door to pinch a person is divided in the visual perception area, and the passenger trajectory and the divided area that may cause the platform door to pinch a person are unified into the same pixel coordinate system.

[0022] As a preferred technical solution, the method confirms whether a single passenger is getting on or off the vehicle by obtaining the time point at which the platform door fails to execute the train control system closing instruction and the historical trajectory of the passenger before the time point.

[0023] As a preferred technical solution, the method forms a passenger logic scene by comprehensively considering the passenger historical behavior, the real-time trajectory of the passenger after the closing failure of the screen door, and the passenger stop point, so as to determine the risk of being pinched by the passenger.

[0024] As a preferred technical solution, the final result of the method is the risk of being pinched by all passengers appearing in the visual perception area at the time of the closing failure of the platform door, so as to determine the risk and danger of the platform door pinching a person.

[0025] As a preferred technical solution, the method unifies the possibility and danger of the platform door pinching a person into three levels, namely low, ordinary, and high, and configures corresponding processing schemes for them.

[0026] According to a second aspect of the present application, a device for the intelligent anti-pinch method of the rail transit platform door is provided, which comprises a visual perception device, a trackside target controller, and a trackside edge computing device.

[0027] The arrangement of the visual perception device satisfies that when the train door and the platform door are opened, the area inside the train and near the train door and the area near the platform outside the platform door can be covered.

[0028] The trackside target controller collects the current state of the platform door and the train door and whether the platform door and the train door successfully execute the closing instruction issued by the train control system.

[0029] The trackside edge computing device is responsible for structuring video data to track the same object in different video frames and obtain its dynamic trajectory; meanwhile, the trackside edge computing device pre-stores the position of the platform door train door in the pixel coordinate system, and performs logical operation on the current platform door state and the target object trajectory captured by the visual perception device to identify the dangerous scene and issue an alarm.

[0030] According to a third aspect of the present application, an electronic device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the method when executing the program.

[0031] According to a fourth aspect of the present application, a computer readable storage medium is provided, which stores a computer program, and the program is executed by a processor to implement the method.

[0032] Compared with the prior art, the present application has the following advantages:

[0033] 1) The present application improves the identification accuracy of the cause of the failure scenario of the platform door closing failure through the trajectory and behavior logic of the passenger during the boarding and alighting process.

[0034] 2) The present application unifies the platform door pinning possibility and its risk level into one metric, reducing the complexity of the system.

[0035] 3) The present application increases the safety of the system through the identification of the platform door pinning risk probability.

[0036] 4) The present application can accurately distinguish the platform door pinning risk from the platform door closing failure caused by other reasons, enhancing the operation efficiency of the system.

[0037] 5) The present application has low coupling with the existing unmanned train control system, and has high system redundancy. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 Fig. 1 is a schematic diagram of the intelligent anti-pinning device of the track transit platform door of the present application;

[0039] Figure 2 Fig. 4 is a flowchart of the overall passenger identification and tracking algorithm of the present application;

[0040] Figure 3 Fig. 6 is a flowchart of the platform door pinning possibility algorithm judgment of the present application;

[0041] Figure 4 Fig. 8 is a schematic diagram of the passenger alighting pinning high possibility scenario under the condition of the platform door closing failure and the train door closing success;

[0042] Figure 5A passenger getting off the train under the scenario of the platform door closing failure and the train door closing success;

[0043] Figure 6 A passenger getting on the train under the scenario of the platform door closing failure and the train door closing success;

[0044] Figure 7 A passenger getting on the train under the scenario of the platform door and the train door closing failure;

[0045] Figure 8 A passenger getting off the train under the scenario of the platform door and the train door closing failure;

[0046] Figure 9 A passenger getting on and off the train under the scenario of the platform door and the train door closing failure; DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work should belong to the protection scope of the present application.

[0048] The rail transit intelligent platform door anti-pinch method based on machine vision passenger trajectory analysis provided in the present application uses a machine vision algorithm to identify and track passengers in a visual perception area and record the unique trajectories of the passengers. The behavior logic of the passengers is judged by combining the trajectories of the passengers with the states of platform doors and train doors, so that the possibility and danger of platform door pinching from the perspective of passenger behavior are determined. The possibility and danger of the related hazards are unified as low, ordinary and high, and different processing solutions are provided for the passengers. The present application can achieve overall protection for the whole process of passenger getting on and off the train. By judging the trajectories and behaviors of the passengers, passenger behavior logic scenarios are formed, and the possibility and danger of platform door pinching are determined. The present application is beneficial to the safety protection of passengers getting on and off the train. After being linked with the existing signal system, the present application can be used as a supplement to the existing system to improve the safety in specific scene operation environments. The present application can be adapted to platform operations in existing rail transit unmanned and manned systems, and realizes intelligent protection of platform doors.

[0049] To realize the identification and tracking of passengers, in the implementation, a target capture deep learning algorithm and a multi-target tracking algorithm are selected. The neural network algorithm based on Draknet 53 can identify small target objects, realize detailed detection of the target objects, and has excellent real-time performance. The multi-target tracking algorithm is evolved from the Sort algorithm. The Sort algorithm uses simple Kalman filtering to process the relevance of frame data and uses the Hungarian algorithm for relevance measurement. This simple algorithm has good performance at high frame rates. However, since the Sort algorithm ignores the surface features of the detected objects, it is only accurate when the object state estimation uncertainty is low. In the implementation, a more reliable measurement is used instead of the relevance measurement, and a CNN network is trained on a large-scale pedestrian dataset to extract features and increase the robustness of the network to loss and obstacles. The overall algorithm is equivalent to a second-order structure in target detection, and target recognition + tracking are adopted. The overall algorithm implementation is as shown in Figure 2 The algorithm is used to obtain a unique ID for each passenger in the video monitoring area, and the moving track of the passenger in the pixel coordinate system is obtained. Each unique ID corresponds to a pedestrian track for subsequent logical judgment.

[0050] The possibility of the occurrence of the passenger trapping scene is judged according to the action track of the passenger in the visual perception area and the behavior logic after the door closing failure. The possibility is divided into low, ordinary, and high as shown in Figure 3 Different possibilities correspond to different processing procedures, which reduce the passenger trapping risk while taking into account the operation efficiency. For example, if the passenger trapping risk is low, a remote secondary door closing attempt is performed. If the passenger trapping risk is ordinary, the operation center links the CCTV to perform secondary confirmation to determine whether passenger trapping occurs. If the passenger trapping risk is high, a train remote braking instruction is directly sent to ensure that the train does not move and no safety problem occurs.

[0051] When the train door and the platform door are successfully closed at the same time, it is a normal state.

[0052] When the train door is successfully closed but the platform door fails to close, the current time is recorded as t1. At this time, all passengers in the video monitoring area are identified, and the track of the passenger in the pixel coordinate system is queried.

[0053] As shown in Figure 4 When the passenger behavior track shows that the passenger behavior intention is to get off, and the passenger track has no obvious movement within 7 seconds from the time t1, it is determined that the track stops, and the track stop point is in the dangerous area (within one meter from the longitudinal and transverse direction of the platform door). If the passenger track in the monitoring area meets the above description, it is determined that the passenger is highly likely to be trapped when getting off the train. Therefore, the train door is successfully closed. If the platform staff handle it improperly, the automatic driving train has a movement risk, and it is determined that the platform door trapping risk level is high.

[0054] When the passenger behavior trajectory shows that the passenger behavior intention is to get off as shown in FIG. 6, the passenger trajectory does not form a stopping point or the stopping point is outside the monitoring area from the t1 moment, if the passenger trajectory in the monitoring area meets the above description, it is determined that the above passenger is less likely to be sandwiched, if all the passenger trajectories do not have ordinary or high risk levels, it is determined that the platform door sandwiching risk is low, if all the passenger trajectories contain higher level sandwiching risks, the highest possibility is alarmed and the next step is processed. Figure 5 When the passenger behavior trajectory shows that the passenger behavior intention is to get off as shown in FIG. 6, the passenger trajectory does not form a stopping point or the stopping point is outside the monitoring area from the t1 moment, if the passenger trajectory in the monitoring area meets the above description, it is determined that the above passenger is less likely to be sandwiched, if all the passenger trajectories do not have ordinary or high risk levels, it is determined that the platform door sandwiching risk is low, if all the passenger trajectories contain higher level sandwiching risks, the highest possibility is alarmed and the next step is processed.

[0055] Figure 6 When the passenger behavior trajectory shows that the passenger behavior intention is to get off as shown in FIG. 6, the passenger trajectory does not form a stopping point or the stopping point is outside the monitoring area from the t1 moment, if the passenger trajectory in the monitoring area meets the above description, it is determined that the above passenger is less likely to be sandwiched, if all the passenger trajectories do not have ordinary or high risk levels, it is determined that the platform door sandwiching risk is low, if all the passenger trajectories contain higher level sandwiching risks, the highest possibility is alarmed and the next step is processed.

[0056] When the train door closing fails and the platform door closing fails, the current time is recorded as t1, at this time, all passengers in the video monitoring area are identified, and the trajectory of the passenger in the pixel coordinate system is queried.

[0057] When the passenger behavior trajectory shows that the passenger behavior intention is to get off as shown in FIG. 6, the passenger trajectory does not form a stopping point or the stopping point is outside the monitoring area from the t1 moment, if the passenger trajectory in the monitoring area meets the above description, it is determined that the above passenger is less likely to be sandwiched, if all the passenger trajectories do not have ordinary or high risk levels, it is determined that the platform door sandwiching risk is low, if all the passenger trajectories contain higher level sandwiching risks, the highest possibility is alarmed and the next step is processed. Figure 7 Figure 8 When the passenger behavior trajectory shows that the passenger behavior intention is to get off as shown in FIG. 6, the passenger trajectory does not form a stopping point or the stopping point is outside the monitoring area from the t1 moment, if the passenger trajectory in the monitoring area meets the above description, it is determined that the above passenger is less likely to be sandwiched, if all the passenger trajectories do not have ordinary or high risk levels, it is determined that the platform door sandwiching risk is low, if all the passenger trajectories contain higher level sandwiching risks, the highest possibility is alarmed and the next step is processed.

[0058] When the passenger behavior trajectory shows that the passenger behavior intention is to get off as shown in FIG. 6, the passenger trajectory does not form a stopping point or the stopping point is outside the monitoring area from the t1 moment, if the passenger trajectory in the monitoring area meets the above description, it is determined that the above passenger is less likely to be sandwiched, if all the passenger trajectories do not have ordinary or high risk levels, it is determined that the platform door sandwiching risk is low, if all the passenger trajectories contain higher level sandwiching risks, the highest possibility is alarmed and the next step is processed. Figure 9 The above is the introduction of the method embodiment, and the scheme described in the application is further described through the device embodiment.

[0059] When the passenger behavior trajectory shows that the passenger behavior intention is to get off as shown in FIG. 6, the passenger trajectory does not form a stopping point or the stopping point is outside the monitoring area from the t1 moment, if the passenger trajectory in the monitoring area meets the above description, it is determined that the above passenger is less likely to be sandwiched, if all the passenger trajectories do not have ordinary or high risk levels, it is determined that the platform door sandwiching risk is low, if all the passenger trajectories contain higher level sandwiching risks, the highest possibility is alarmed and the next step is processed.

[0060] Figure 1 ​​​As shown, the platform door intelligent anti-pinch device based on machine vision passenger trajectory capture of the application comprises a camera, a trackside target controller and a trackside edge computing device. The camera is arranged to cover the area near the train door and the area near the platform outside the platform door when the train door and the platform door are opened. The trackside target controller is responsible for collecting the current state of the platform door and the train door and whether the platform door and the train door successfully execute the door closing instruction issued by the train control system. The trackside edge computing device is responsible for structuring the video data to track the same object in different video frames and obtain the dynamic trajectory of the object. At the same time, the trackside edge computing device pre-stores the position of the platform door and the train door in the pixel coordinate system, and performs logical operation on the current platform door state and the target object trajectory captured by the visual perception device to identify the dangerous scene and issue an alarm.

[0061] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described modules can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0062] The electronic device of the application includes a central processing unit (CPU) that can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM, and RAM are connected to each other through a bus. An input / output (I / O) interface is also connected to the bus.

[0063] A plurality of components in the device are connected to the I / O interface, including: an input unit such as a keyboard, a mouse, etc.; an output unit such as various types of displays, a speaker, etc.; a storage unit such as a magnetic disk, an optical disk, etc.; and a communication unit such as a network card, a modem, a wireless communication transceiver, etc. The communication unit allows the device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0064] The processing unit performs various methods and processes described above, such as the method of the application. For example, in some embodiments, the method of the application can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program can be loaded and / or installed on the device via the ROM and / or the communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of the method of the application described above can be performed. Alternatively, in other embodiments, the CPU can be configured to perform the method of the application by any other appropriate means (e.g., by means of firmware).

[0065] The functionality described above in this document can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-programmable Gate Arrays (FPGAs), Program-specific Integrated Circuits (ASICs), Program- specific Standard Products (ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), etc.

[0066] Program code for carrying out methods of the present application can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces a means for implementing the functions / acts specified in the flowcharts and / or block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0067] In the context of the present application, a machine-readable medium can be a tangible medium that can contain or store program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable storage medium can include, without limitation, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing.

[0068] The above description is only specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An intelligent anti-pinch method for rail transit platform doors, characterized in that: This method uses a machine vision algorithm to track passengers getting on and off the platform and capture their dynamic trajectories. The method captures their behavior based on the dynamic trajectories and combines the current status of the platform door and train door to determine the possibility and danger of platform door pinching, thereby achieving passenger boarding and alighting protection. The method identifies and tracks passengers within a visual perception area, and assigns an identification code to each passenger, wherein the visual perception area is the boarding and alighting area, including the platform and the interior of the train; The identification codes are divided into multiple levels; if a passenger changes from an overlapping individual to an independent individual within the visual perception area, the previous identification code is retained and a lower-level identification code is added; if a passenger changes from an independent individual to an overlapping individual within the visual perception area, the previous identification code is retained and a lower-level identification code is added; if the passengers do not separate within the visual perception area, the overlapping objects are determined to be the same object and have one identification code; This method records the trajectory of each passenger and associates it with an identification code, where the starting point of the trajectory is when the passenger enters the visual perception area and the end point is when the passenger leaves the visual perception area. For objects with multiple identification codes, their trajectories are spliced ​​according to the order of the identification codes to ensure the continuity of the trajectory until they leave the perception area. The uniqueness of the identification code must be guaranteed for each level, and ultimately, multiple trajectories are allowed to form when a target object leaves the perception area. A unique identification code is set for each trajectory.

2. The intelligent anti-pinch method for rail transit platform doors according to claim 1, characterized in that: This method uses visual perception equipment to monitor the area where passengers get on and off the train. Its monitoring range needs to cover both the platform door and the area inside the train near the train door when the platform door and the train door are open at the same time.

3. The intelligent anti-pinch method for rail transit platform doors according to claim 1, characterized in that: When the trajectory shows that the passenger leaves the visual perception area, the identification code and historical trajectory are cleared, and a new identification code is assigned when the passenger re-enters.

4. The intelligent anti-pinch method for rail transit platform doors according to claim 1, characterized in that: This method forms a complete behavioral logic scenario of passengers when getting on and off the bus through passenger trajectory analysis.

5. The intelligent anti-pinch method for rail transit platform doors according to claim 4, characterized in that: When the real-time trajectory shows that the passenger stops for a set time, it is identified as a passenger stop point; The area where the platform door may pinch people is divided within the visual perception area, and the passenger trajectory and the divided area where the platform door may pinch people are unified into the same pixel coordinate system.

6. The intelligent anti-pinch method for rail transit platform doors according to claim 4, characterized in that: This method obtains the time point when the platform door fails to execute the closing command of the train control system, and confirms whether a single passenger is boarding or getting off the train based on his or her historical trajectory before this time point.

7. The intelligent anti-pinch method for rail transit platform doors according to claim 4, characterized in that: This method integrates the passenger's historical behavior, the passenger's real-time trajectory after the screen door fails to close, and the passenger's stopping point to form a passenger logic scenario to determine the risk of passenger being trapped.

8. The intelligent anti-pinch method for rail transit platform doors according to claim 4, characterized in that: The final result of this method is to judge the risk of being pinched by the platform door and its dangerousness by comprehensively considering the risk of all passengers appearing in the visual perception area when the platform door fails to close.

9. The intelligent anti-pinch method for rail transit platform doors according to claim 4, characterized in that: This method unifies the possibility and danger of platform door pinching people into three levels: low, normal, and high, and configures corresponding treatment plans for them.

10. A device for the intelligent anti-pinch method for rail transit platform doors according to claim 1, characterized in that: Including visual perception equipment, trackside target controller and trackside edge computing equipment; The visual perception equipment is arranged so as to cover the area inside the train and near the train doors, and the area outside the platform doors and near the platform when the train doors and platform doors are open; The trackside target controller collects the current status of the platform door and the train door and whether the platform door and the train door have successfully executed the door closing command issued by the train control system; The trackside edge computing device is responsible for structuring the video data to track the same object in different video frames and obtain its dynamic trajectory; at the same time, the trackside edge computing device will pre-store the position of the platform door and train door in the pixel coordinate system, and integrate the current platform door status and the target object trajectory captured by the visual perception device to perform logical operations to identify dangerous scenes and issue alarms.

11. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images