Crossing guardrail control method, system, storage medium and device

Abstract

The present invention provides a method, system, storage medium and device for controlling a road crossing barrier. The method comprises: obtaining a train approaching signal from the outside, turning on a signal light to warn the approaching train; performing image detection on a road crossing area to obtain the situation of people and vehicles stranded in the road crossing area; and controlling the barrier to fall down and block the road crossing according to the situation of people and vehicles stranded in the road crossing area. The road crossing barrier control method of the present invention improves safety.

Description

Technical Field

[0001] The present invention belongs to the technical field of railway signals, and in particular relates to a method, system, storage medium and equipment for controlling a crossing barrier. Background Art

[0002] A railway crossing, also known as a railway level crossing, is a place where railway tracks intersect with roads or sidewalks on the same plane. It is an important node where the railway and road traffic systems intersect, and has a significant impact on railway transportation and road traffic safety. Railway crossings are usually equipped with safety facilities such as signal lights, fences, sirens, etc. to ensure that road users can be warned in time and take corresponding safety measures when trains pass. The conventional railway crossing train passing process is as follows: According to the train route information, the duty officer turns on the signal lights in advance and lowers the barriers to block the crossing area. After the train has completely passed, the barriers are opened and the signal lights are turned off.

[0003] Existing crossing barrier controls are mainly of two types: manual control and automatic control. Manual control closes the barriers after confirming that the crossing area is clear by observation at the crossing or remote monitoring. Automatic control generally closes the barriers after determining that there are no obstacles in the crossing area through cameras or radars. In the above two situations, all barriers will be closed at the same time.

[0004] However, when the flow of people at the crossing is large, closing all the barriers at the same time is likely to trap people in the crossing area, reduce the efficiency of train traffic, and create safety hazards. Therefore, the safety is relatively low. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a method, system, storage medium and device for controlling road crossing barriers. Based on image detection technology, the method acquires information on the detention of people and vehicles in the road crossing area in real time, and optimizes the control logic of the barriers according to the detention of people and vehicles. This can improve the traffic efficiency of trains, reduce the risks of train operations caused by detention, and improve safety without adding additional equipment.

[0006] The present invention is achieved through the following technical solutions:

[0007] Obtain the train's incoming signal from the outside and turn on the signal light to warn of the incoming train;

[0008] Perform image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area;

[0009] According to the situation of people and vehicles stranded in the crossing area, the control barriers are lowered to block the crossing.

[0010] Optional,

[0011] The control of the falling of the barrier according to the situation of people and vehicles stranded in the crossing area includes:

[0012] If there is no vehicle stranded in the crossing area, all the barriers are controlled to fall down;

[0013] If there are people or vehicles stranded in the crossing area, the movement direction of the stranded people or vehicles is detected, and the railing is controlled to fall according to the movement direction of the people or vehicles.

[0014] Optional,

[0015] The step of controlling the railing to fall according to the movement direction of the person or vehicle comprises:

[0016] If the movement directions of the person and the vehicle are the same, the barrier behind the stranded person and the vehicle is closed until no vehicle is stranded in the crossing area, and the barrier on the other side is closed;

[0017] If the movement directions of the people and the vehicles are in different directions, the oncoming vehicle will be notified to stop and observe before passing, until there are no vehicles stranded in the crossing area, and all the barriers will be controlled to fall.

[0018] Optional,

[0019] The image detection of the crossing area includes:

[0020] The YOLOv3 target detection system is used to perform image detection on the crossing area.

[0021] Optional,

[0022] The detecting the movement direction of the stranded person or vehicle comprises:

[0023] According to the image detection results, all positions of the same object in a preset number of frames are obtained to form two-dimensional data;

[0024] Linear fitting is performed based on the two-dimensional data to obtain a straight line describing the movement trend of the object, and the vector direction of the straight line is the movement direction of the object.

[0025] Optional,

[0026] According to the image detection results, objects whose position straight-line distance is less than a preset threshold are classified into one category;

[0027] Select an object in the same class for motion direction detection.

[0028] The present invention also provides a road crossing guardrail control system for implementing the above method, the system comprising:

[0029] The oncoming train warning module is used to obtain the oncoming train signal from the outside and turn on the signal light to warn the oncoming train;

[0030] An image detection module is used to perform image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area;

[0031] The barrier control module is used to control the barrier to fall and block the crossing according to the situation of people and vehicles stranded in the crossing area.

[0032] Optional,

[0033] The fence control module is also configured to:

[0034] If there is no vehicle stranded in the crossing area, all the barriers are controlled to fall down;

[0035] If there are people or vehicles stranded in the crossing area, the movement direction of the stranded people or vehicles is detected, and the railing is controlled to fall according to the movement direction of the people or vehicles.

[0036] Optional,

[0037] The step of controlling the railing to fall according to the movement direction of the person or vehicle comprises:

[0038] If the movement directions of the person and the vehicle are the same, the barrier behind the stranded person and the vehicle is closed until no vehicle is stranded in the crossing area, and the barrier on the other side is closed;

[0039] If the movement directions of the people and the vehicles are in different directions, the oncoming vehicle will be notified to stop and observe before passing, until there are no vehicles stranded in the crossing area, and all the barriers will be controlled to fall.

[0040] Optional,

[0041] The image detection module is further configured to:

[0042] The YOLOv3 target detection system is used to perform image detection on the crossing area.

[0043] Optional,

[0044] The detecting the movement direction of the stranded person or vehicle comprises:

[0045] According to the image detection results, all positions of the same object in a preset number of frames are obtained to form two-dimensional data;

[0046] Linear fitting is performed based on the two-dimensional data to obtain a straight line describing the movement trend of the object, and the vector direction of the straight line is the movement direction of the object.

[0047] The present invention also provides a computer-readable storage medium storing one or more programs, which, when executed, can implement the aforementioned road crossing barrier control method.

[0048] The present invention also provides a device, comprising a processor, a communication interface, a computer-readable storage medium and a communication bus; wherein the processor, the communication interface and the computer-readable storage medium communicate with each other via the communication bus;

[0049] The processor is configured to execute a program stored in a computer-readable storage medium.

[0050] Compared with the prior art, the present invention has the following advantages:

[0051] 1. The crossing barrier control method proposed in the present invention obtains the situation of people and vehicles stranded in the crossing area in real time, optimizes the barrier control logic according to the situation of people and vehicles stranded, can adapt to more complex railway crossings with higher passenger flow, and improves the travel efficiency and safety of trains.

[0052] 2. Use image detection technology to automatically identify various objects in the crossing area. Using the YOLOv3 target detection system, only one forward propagation of the network is required to complete the entire detection process. It can run faster and better understand the objects in the scene while considering the entire image, thereby improving the real-time performance and flexibility of image detection.

[0053] 3. It can be achieved through existing equipment without adding new sensing equipment and control equipment. It can reduce the risk of train operations caused by detention and improve safety without adding additional equipment.

[0054] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0056] Figure 1 A schematic flow chart of a method for controlling a road crossing guardrail is shown;

[0057] Figure 2 The structure schematic block diagram of the crossing guardrail control system is shown;

[0058] Figure 3 A control flow diagram of an embodiment of the present invention is shown;

[0059] Figure 4 It is a structural schematic diagram of a device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] See attached Figure 1 , the method of the present invention comprises:

[0062] S1. Obtain the train approaching signal from the outside and turn on the signal light to warn of the approaching train.

[0063] S2, performing image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area;

[0064] Among them, the YOLOv3 target detection system is used to perform image detection on the crossing area.

[0065] S3. According to the situation of people and vehicles stranded in the crossing area, control the railings to fall and block the crossing.

[0066] If an unmanned vehicle is stranded in the crossing area, all the barriers are controlled to fall down;

[0067] If there are people or vehicles stranded in the crossing area, the moving direction of the stranded people or vehicles is detected, and the railing is controlled to fall according to the moving direction of the people or vehicles.

[0068] If the movement directions of the person and the vehicle are the same, the barrier behind the stranded person and vehicle is closed until no vehicle is stranded in the crossing area, and the barrier on the other side is closed;

[0069] If the movement directions of the people and the vehicles are in different directions, the oncoming vehicle will be notified to stop and observe before passing, until there are no vehicles stranded in the crossing area, and all the barriers will be controlled to fall.

[0070] Among them, according to the image detection results, all positions of the same object in a preset number of frames are obtained to form two-dimensional data;

[0071] Performing linear fitting based on the two-dimensional data to obtain a straight line describing the movement trend of the object, wherein the vector direction of the straight line is the movement direction of the object,

[0072] Among them, according to the image detection results, objects whose position straight-line distance is less than a preset threshold are classified into one category;

[0073] Select an object in the same class for motion direction detection.

[0074] Specifically,

[0075] 1. Obtain the train's oncoming signal from the outside and turn on the signal light to warn of the oncoming train.

[0076] When the crossing equipment learns from the outside that a train is about to pass through the crossing, it turns on the signal light through a relay to warn the oncoming vehicle, reminding people and vehicles about to pass through the crossing to stop entering the crossing area.

[0077] 2. Perform image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area.

[0078] Conduct real-time image detection of the crossing area and continuously identify the stranded personnel and vehicles in the crossing area.

[0079] In this embodiment, the YOLOv3 (You Only Look Once version 3) target detection system can be used for image detection. The YOLOv3 target detection system is an advanced computer vision technology that can automatically identify various objects in an image. The system consists of four main components: an input layer, a feature extraction network based on Darknet-53, a feature fusion network, and a prediction layer.

[0080] When using the YOLOv3 target detection system for image detection, the input image is resized to a fixed size, such as 416x416 pixels, to ensure that all images have the same size for subsequent processing; the input image is sent to the Darknet-53 convolutional neural network to extract features at various levels; the extracted features gradually transition from simple lines and textures to more complex shapes and category information, and the features at different levels are combined through the feature fusion network to enhance the details and contextual information in the input image; predictions are made on three feature maps of different scales selected from the original feature map obtained in the feature extraction stage to accurately locate and classify targets of different sizes, and non-maximum suppression is used to remove overlapping detection frames, retaining only the most likely detection results.

[0081] Compared with traditional two-stage detectors, YOLOv3 only requires one forward propagation of the network to complete the entire detection process, can run faster, and can better understand objects in the scene while considering the entire image, thus having higher real-time performance and flexibility.

[0082] In addition to using the YOLOv3 target detection system, there are some commonly used detection schemes in the field of machine vision, such as SSD, Faster R-CNN, Mask R-CNN, and YOLOv4, the successor version of YOLO. Among them, SSD (Single Shot MultiBox Detector) is a fast object detection system that predicts object categories and bounding box positions in a single deep neural network, similar to YOLO; Faster R-CNN uses a region proposal network (RPN) to generate candidate regions, and then classifies and locates them through a convolutional neural network; Mask R-CNN is an extension of Faster R-CNN, which can achieve pixel-level instance segmentation, that is, it can not only identify objects, but also accurately segment the contours of objects. YOLOv4 is a successor version of the YOLO series, which improves speed and accuracy, especially in real-time applications. The advantages and disadvantages of the above technologies are compared in Table 1. After comprehensive comparison, YOLOv3 is used in this embodiment.

[0083] Table 1

[0084]

[0085] 3. Based on the situation of people and vehicles stranded in the crossing area, control the railings to fall and block the crossing.

[0086] According to the situation of people and vehicles stranded in the crossing area, control the lowering of the barriers, including:

[0087] If there is no vehicle stranded in the crossing area, all the barriers are controlled to fall down;

[0088] If there are people or vehicles stranded in the crossing area, the movement direction of the stranded people or vehicles is detected, and the railing is controlled to fall according to the movement direction of the people or vehicles.

[0089] The step of controlling the railing to fall according to the movement direction of the person or vehicle comprises:

[0090] If the movement directions of the person and the vehicle are the same, the barrier behind the stranded person and the vehicle is closed until no vehicle is stranded in the crossing area, and the barrier on the other side is closed;

[0091] If the movement directions of the people and the vehicles are in different directions, the oncoming vehicle will be notified to stop and observe before passing, until there are no vehicles stranded in the crossing area, and all the barriers will be controlled to fall.

[0092] In this embodiment, if no pedestrians or vehicles are identified in the crossing area, all barriers are closed to block the crossing;

[0093] If there are stranded pedestrians or vehicles in the crossing area, the forward direction of the stranded pedestrians or vehicles is further detected:

[0094] If the pedestrians or vehicles stranded in the crossing area are moving in the same direction, the barrier behind the pedestrian or vehicle will be closed to prevent other pedestrians or vehicles from continuing to enter the crossing. When it is determined that there are no pedestrians or vehicles stranded in the crossing area, the barrier on the other side will be closed, and the train can pass through the crossing.

[0095] If there are pedestrians or vehicles entering the crossing area on both sides of the crossing, the crossing barriers will not move. At this time, a notification will be sent to the train that is about to pass through the crossing, reminding it to stop and observe before passing, until image detection recognizes that there are no pedestrians or vehicles in the crossing area. At this time, the crossing equipment controls the barriers to fall, closing the crossing before the train can pass.

[0096] In this embodiment, detecting the direction of travel of a person or vehicle includes:

[0097] According to the image detection results, all positions of the same object in a preset number of frames are obtained to form two-dimensional data;

[0098] Linear fitting is performed based on the two-dimensional data to obtain a straight line describing the movement trend of the object, and the vector direction of the straight line is the movement direction of the object.

[0099] Taking the YOLOv3 algorithm as an example, the video is passed into the YOLOv3 model at a speed of 25 frames per second for calculation. Each frame of the picture will return a set of detection results, including object category, confidence, position and size. The object categories include trains, cars, bicycles, people, etc. The confidence is a percentage data ranging from 0-100%, which is used to characterize the probability of detecting the object. The position is the center point of the size box identified, and the size is the coordinates of the four vertices of the size box. When the confidence is greater than 90%, it can be considered that the detection result is accurate.

[0100] Based on the accurate detection results of 10 frames, all positions of the same object in 10 frames are obtained, and then a set of two-dimensional data can be obtained. The least squares problem in linear regression is used to find a straight line that minimizes the sum of the squares of the distances from all these points to the straight line. The vector direction represented by this straight line is the moving direction of the object.

[0101] In actual detection, objects whose straight-line distance is less than a preset threshold can be classified into one category, and the moving direction of this category is determined to be consistent. Selecting an object in the same category to detect the moving direction can improve detection efficiency.

[0102] Different barrier control schemes are executed according to whether the movement directions of the objects detected in the current crossing area are in the same direction.

[0103] At present, the common method for controlling fences is to connect the control level in the fence machine to the relay of the crossing equipment. When the crossing equipment receives the fence drop command, it sends a command to the relay control box to set all relay levels high, and control all fences to move simultaneously. In this embodiment, the communication content between the crossing host and the relay control box can be modified as follows: setting some relay levels high can control a single fence, and there is no need to modify the existing equipment.

[0104] In this embodiment, the camera is connected to the switch in the road crossing equipment, and the barrier machine and the signal light are connected to the relay. The camera can obtain the video information of the road crossing in real time, the switch connects the industrial computer and the camera, the industrial computer calls the camera through the network port to obtain the video content for processing, and the relay controls the power switch of the barrier machine and the signal light, and sends control instructions to drive the barrier machine and the signal light to work.

[0105] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0106] See attached Figure 2 , shows the structure of a road crossing barrier control system for implementing the above method, including an incoming vehicle warning module, an image detection module, and a barrier control module.

[0107] The oncoming train warning module is used to obtain the oncoming train signal from the outside and turn on the signal light to warn the oncoming train;

[0108] The image detection module is used to perform image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area;

[0109] The barrier control module is used to control the barrier to fall and block the crossing according to the situation of people and vehicles stranded in the crossing area.

[0110] Figure 3 This is a schematic diagram of the control flow of an embodiment of the present invention. After obtaining a signal from the outside that a train is about to pass through a crossing, the signal light is turned on; image detection is started to confirm whether there are people or vehicles stranded in the crossing area: if there are no people or vehicles stranded, all barriers are controlled to fall at the same time; if there are people or vehicles stranded, the forward direction of the stranded people or vehicles is identified: if the forward directions are the same, the barriers behind the people and vehicles are closed; if the forward directions are different, the oncoming vehicle is notified to stop and observe before passing; wait for the stranded people and vehicles to leave the crossing area and close all barriers.

[0111] In addition, an embodiment of the present invention further provides a road crossing guardrail control device, comprising:

[0112] The oncoming train warning module obtains the oncoming train signal from the outside and turns on the signal light to warn of the oncoming train.

[0113] The image detection module performs image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area; wherein the YOLOv3 target detection system is used to perform image detection on the road crossing area.

[0114] The barrier control module controls the barriers to fall and blocks the crossing according to the situation of stranded people and vehicles in the crossing area; if no vehicle is stranded in the crossing area, all the barriers are controlled to fall; if there are stranded people and vehicles in the crossing area, the movement direction of the stranded people and vehicles is detected, and the barriers are controlled to fall according to the movement direction of the people and vehicles; if the movement directions of the people and vehicles are in the same direction, the barriers behind the stranded people and vehicles are closed until no vehicle is stranded in the crossing area, and the barriers on the other side are closed; if the movement directions of the people and vehicles are in different directions, the oncoming vehicle is notified to stop and observe before passing, until no vehicle is stranded in the crossing area, and all the barriers are controlled to fall.

[0115] Based on the same inventive concept, the present invention also provides a computer-readable storage medium storing one or more programs, which, when executed, can implement the aforementioned crossing barrier control method.

[0116] like Figure 4 As shown, an embodiment of the present invention further provides a device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus.

[0117] The memory is a computer-readable storage medium for storing one or more programs.

[0118] The processor is used to execute the program stored in the computer-readable storage medium.

[0119] The computer-readable storage medium may be included in the device / apparatus described in the above embodiments; or may exist independently without being assembled into the device / apparatus.

[0120] Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent substitutions for some of the technical features therein; and these modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for controlling a road crossing guardrail, characterized in that: include: Obtain the train's incoming signal from the outside and turn on the signal light to warn of the incoming train; Perform image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area; According to the situation of people and vehicles stranded in the crossing area, control the barriers to fall and block the crossing; if no vehicles are stranded in the crossing area, control all barriers to fall; if there are people and vehicles stranded in the crossing area, detect the movement direction of the stranded people and vehicles, and control the barriers to fall according to the movement direction of the people and vehicles; Wherein, the control of the barrier falling down according to the moving direction of the person and the vehicle includes: if the moving directions of the person and the vehicle are in the same direction, closing the barrier behind the stranded person and the vehicle until no vehicle is stranded in the crossing area, and closing the barrier on the other side; if the moving directions of the person and the vehicle are in different directions, notifying the oncoming vehicle to stop and observe before passing, until no vehicle is stranded in the crossing area, and controlling all the barriers to fall down; Among them, the detecting of the movement direction of the stranded person or vehicle includes: obtaining all positions of the same object in a preset number of frames according to the image detection results to form two-dimensional data; performing linear fitting based on the two-dimensional data to obtain a straight line describing the movement trend of the object, and the vector direction of the straight line is the movement direction of the object.

2. The method according to claim 1, characterized in that The image detection of the crossing area includes: The YOLOv3 target detection system is used to perform image detection on the crossing area.

3. The method according to any one of claims 1 to 2, characterized in that: The method further comprises: According to the image detection results, objects whose position straight-line distance is less than a preset threshold are classified into one category; Select an object in the same class for motion direction detection.

4. A road crossing guardrail control system, characterized in that: The system comprises: The oncoming train warning module is used to obtain the oncoming train signal from the outside and turn on the signal light to warn the oncoming train; An image detection module is used to perform image detection on the road crossing area to obtain the situation of people and vehicles stranded in the road crossing area; The barrier control module is used to control the barriers to fall and block the road crossing according to the situation of people and vehicles stranded in the road crossing area; if no vehicles are stranded in the road crossing area, all barriers are controlled to fall; if there are people and vehicles stranded in the road crossing area, the movement direction of the stranded people and vehicles is detected, and the barriers are controlled to fall according to the movement direction of the people and vehicles; Wherein, the control of the barrier falling down according to the moving direction of the person and the vehicle includes: if the moving directions of the person and the vehicle are in the same direction, closing the barrier behind the stranded person and the vehicle until no vehicle is stranded in the crossing area, and closing the barrier on the other side; if the moving directions of the person and the vehicle are in different directions, notifying the oncoming vehicle to stop and observe before passing, until no vehicle is stranded in the crossing area, and controlling all the barriers to fall down; Among them, the detecting of the movement direction of the stranded person or vehicle includes: obtaining all positions of the same object in a preset number of frames according to the image detection results to form two-dimensional data; performing linear fitting based on the two-dimensional data to obtain a straight line describing the movement trend of the object, and the vector direction of the straight line is the movement direction of the object.

5. The system according to claim 4, characterized in that The image detection module is further configured to: The YOLOv3 target detection system is used to perform image detection on the crossing area.

6. A computer-readable storage medium storing one or more programs, characterized in that: When the one or more programs are executed, the crossing guardrail control method described in any one of claims 1 to 3 can be implemented.

7. An electronic device comprising a processor, a communication interface, the computer-readable storage medium of claim 6, and a communication bus; wherein: The processor, the communication interface, and the computer-readable storage medium communicate electronically with each other via a communication bus; characterized in that: The processor is configured to execute a program stored in a computer-readable storage medium.

Images