Crossing control method and system based on video analysis and magnetic steel sensor

Abstract

The invention provides a crossing control method and system based on video analysis and a magnetic steel sensor, and belongs to the technical field of crossing control. By obtaining a train route state in front of a crossing, when a train triggers a first route position, a first position signal is obtained; tracking the train according to the first position signal video to obtain the position of the train, and giving an alarm through an audible and visual alarm; when the train triggers the second position of the access road, a second position signal is obtained, and the barrier machine is controlled to close the crossing according to the second position signal; and judging whether the current train passes through the crossing area or not according to a target identification result of video tracking, and determining to start a barrier machine to open the crossing and turn off an audible and visual alarm according to a judgment result. Therefore, the automation degree and the accuracy of barrier control are improved through video tracking and identification of the train route by judging the network, and the intelligent degree and convenience are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of road crossing control, and in particular relates to a road crossing control method and system based on video analysis and magnetic steel sensors. Background Art

[0002] Level crossings are an important part of railway operations. At present, railway level crossings are mainly monitored manually, but manual level crossings are prone to unsafe factors such as misoperation and missed operations. The Mulan machine is an important device in level crossing control. In the prior art, the control of the railing machine is determined by the dispatcher of the monitoring center based on the remote level crossing video. It is completed through human control, which is essentially manual management. The existing level crossing monitoring system is not highly intelligent, and lacks intelligent methods that combine video analysis and sensor applications. Summary of the invention

[0003] In view of the above problems, the present invention proposes a crossing control method and system based on video analysis and magnetic steel sensors, which integrate the magnetic steel sensors and video analysis results to automatically control the lifting and lowering of the barrier machine, thereby improving the degree of automation and the accuracy of barrier control.

[0004] An embodiment of the present invention provides a crossing control method based on video analysis and magnetic steel sensors, comprising:

[0005] Acquire the train route status in front of the crossing, wherein the route status includes triggering the first route position and triggering the second route position;

[0006] When the train triggers the first position of the route, a first position signal is obtained, the train position is obtained by video tracking of the train according to the first position signal, and an alarm is issued through an audible and visual alarm;

[0007] When the train triggers the second position of the route, the second position signal is obtained, and the barrier machine is controlled to close the crossing according to the second position signal;

[0008] The target recognition results of video tracking are used to determine whether the current train is passing through the crossing area. Based on the determination results, the barrier machine is turned on to open the crossing and the sound and light alarm is turned off.

[0009] Furthermore, the approach status is obtained by a magnetic steel sensor arranged beside the rail in front of the crossing;

[0010] The first position of the triggered route in the route state is obtained by a magnetic steel sensor set 500 meters in front of the crossing;

[0011] The second position of the triggered route in the route state is obtained by a magnetic steel sensor arranged 300 meters in front of the crossing.

[0012] Furthermore, the method of tracking the train according to the first position signal video to obtain the train position and issuing an alarm through an audible and visual alarm includes:

[0013] Obtaining train route video information, inputting the train route video information into the YOLOV5 network, and identifying the train in the video information through the YOLOV5 network to determine that the train is passing the first position;

[0014] When the train passes the first position, the sound and light alarm sounds an alarm.

[0015] Further, the identifying the train in the video information by using the YOLOV5 network to determine that the train is passing the first position includes:

[0016] Input the video image in the video information into the YOLOV5 network, extract the point coordinates of the target box in the video image, and extract the width and height of the target box in the video image;

[0017] It is determined that the train is passing through the first position according to the point coordinates, width and height of the target frame.

[0018] Further, determining that the train is passing through the first position according to the point coordinates, width and height of the target frame includes:

[0019] The train existence probability in the target frame is calculated according to the point coordinates, width and height of the target frame, and a column existence probability set is generated, and the maximum value in the train existence probability set is determined as the train identification position.

[0020] Furthermore, the target recognition result of the video tracking is used to determine whether the current train is passing through the crossing area, including:

[0021] The target recognition result is normalized, and the normalized result is input into a train departure judgment network, and the train departure area is determined according to the judgment network.

[0022] Furthermore, the judgment network is a convolutional neural network based on one-dimensional convolution and average pooling layers to realize the judgment network of train departure.

[0023] Further, the determining that the train leaves the crossing area according to the judgment network includes:

[0024] The normalized train recognition results of 32 consecutive frames are input into the judgment network as feature vectors, and the input feature vectors are reduced in dimension through one-dimensional convolution and average pooling layers;

[0025] After reducing the dimension of the input feature vector, the recognition results of different sizes are output through the fully connected layer, which are the probability of the train leaving and the probability of the train not leaving. The largest probability between the probability of the train leaving and the probability of the train not leaving is determined as the judgment result.

[0026] Furthermore, the judgment result is that the train leaves the crossing area, and it is determined to open the crossing by turning on the barrier machine and turning off the sound and light alarm.

[0027] Based on the same inventive concept, the present invention also provides a crossing control system based on video analysis and magnetic steel sensors, comprising:

[0028] An acquisition unit, used for acquiring a train route status in front of the crossing, wherein the route status includes a first route triggering position and a second route triggering position;

[0029] An alarm unit is used to obtain a first position signal when the train triggers the first position of the route, track the train according to the first position signal video to obtain the train position and issue an alarm through an audible and visual alarm;

[0030] A control unit, used for obtaining a second position signal when the train triggers the second approach position, and controlling the barrier machine to close the crossing according to the second position signal;

[0031] The judgment unit is used to judge whether the current train is passing through the crossing area through the target recognition result of video tracking, and determine to open the barrier machine to open the crossing and turn off the sound and light alarm according to the judgment result.

[0032] Based on the same inventive concept, the present invention also provides an electronic device on the other hand, comprising: 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;

[0033] a memory storing a computer program;

[0034] The processor implements the crossing control method based on video analysis and magnetic steel sensor when executing the program stored in the memory.

[0035] Based on the same inventive concept, the present invention also provides a computer-readable storage medium on the other hand, storing a computer program, and when the computer program is executed by a processor, the crossing control method based on video analysis and magnetic steel sensors is implemented.

[0036] Beneficial effects of the present invention:

[0037] The present invention obtains the train route status in front of the crossing, and when the train triggers the first route position, obtains the first position signal, tracks the train according to the first position signal, obtains the train position, and issues an alarm through the sound and light alarm; when the train triggers the second route position, obtains the second position signal, and controls the wood barrier to close the crossing according to the second position signal; determines whether the current train is passing through the crossing area through the target recognition result of video tracking, and determines to open the wood barrier to open the crossing and close the sound and light alarm according to the judgment result. It can be seen that the degree of automation and the accuracy of wood barrier control are improved through video tracking and the identification of the train route through the judgment network, and the degree of intelligence and convenience are improved.

[0038] Other features and advantages of the present invention will be described in the following description, and partly become obvious from the description, or be understood by implementing 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 and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. 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.

[0040] Figure 1 A flow chart of a crossing control method based on video analysis and magnetic steel sensors of the present invention;

[0041] Figure 2 Schematic diagram of the crossing control process in an embodiment of the present invention;

[0042] Figure 3 Schematic diagram of the network structure for judging train departure in an embodiment of the present invention;

[0043] Figure 4 A schematic diagram of a crossing control system based on video analysis and magnetic steel sensors according to the present invention;

[0044] Figure 5 The figure is a structural diagram of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] 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.

[0046] It should be noted that the terms "first", "second" etc. in the present application are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged in appropriate circumstances, so that the embodiments of the present application described here. In the present application, the orientation or positional relationship indicated by the terms "upper", "lower", "left", "right", "front", "back", "top", "bottom", "inside", "outside", "middle", "vertical", "horizontal", "lateral", "longitudinal" etc. are based on the orientation or positional relationship shown in the accompanying drawings.

[0047] The embodiment of the present invention provides a crossing control method based on video analysis and magnetic steel sensor, see Figure 1 ,include:

[0048] S101: Acquire the train route status in front of the crossing, where the route status includes triggering the first route position and triggering the second route position;

[0049] S102: when the train triggers the first position of the route, a first position signal is obtained, the train position is obtained by video tracking of the train according to the first position signal, and an alarm is issued through an audible and visual alarm;

[0050] Specifically, two magnetic steel sensors are installed beside the rails in front of the crossing, 500 meters and 300 meters in front of the crossing respectively. When the train approaches the crossing, it will press over the magnetic steel sensors in turn. When the magnetic steel sensor at 500 meters is triggered, the system automatically activates the sound and light alarm to warn pedestrians and social vehicles; when the magnetic steel sensor at 300 meters is triggered, the system automatically controls the barrier machine to fall and close the crossing area.

[0051] S103: When the train triggers the second approach position, a second position signal is obtained, and the barrier machine is controlled to close the crossing according to the second position signal;

[0052] S104: judging whether the current train is passing through the crossing area through the target recognition result of the video tracking, and determining to open the crossing and turn off the sound and light alarm according to the judgment result.

[0053] Specifically, the approach status is obtained by a magnetic steel sensor arranged beside the rail in front of the crossing;

[0054] The first position of the triggered route in the route state is obtained by a magnetic steel sensor set 500 meters in front of the crossing;

[0055] The second position of the triggered route in the route state is obtained by a magnetic steel sensor arranged 300 meters in front of the crossing.

[0056] It should be noted that see Figure 2 The present invention sets magnetic steel sensors on both sides of the train approach in front of the crossing to obtain train passing signals, which are installed 500 meters and 300 meters in front of the crossing respectively, wherein the magnetic steel sensor at the first position is set at 500 meters in front of the crossing, and the magnetic steel sensor at the second position is set at 300 meters in front of the crossing. Video tracking is achieved by a high-definition camera, which is installed on the trackside pole and has an infrared night vision function, and can achieve 24-hour monitoring. Trackside hosts are set on both sides of the rails, with built-in IO control and 4G communication modules, and have the functions of driving sound and light alarms, sending alarm information to wireless train modulation, central monitoring equipment, and uploading alarm pictures and videos, which can realize laser radar point cloud detection, image processing and detection functions. The public network cloud platform has a public network IP address, which can realize the forwarding function of alarm information, pictures, and videos. The central monitoring equipment is deployed in the duty room and has voice alarms, alarm information, pictures, and video viewing and playback functions.

[0057] In some optional embodiments, the step of tracking the train according to the first position signal video to obtain the train position and issuing an alarm through an audible and visual alarm includes:

[0058] Acquire train route video information, input the train route video information into the YOLOV5 network, and identify the train in the video information through the YOLOV5 network to determine that the train is passing the first position;

[0059] When the train passes the first position, the sound and light alarm sounds an alarm.

[0060] Further, the identifying the train in the video information by using the YOLOV5 network to determine that the train is passing the first position includes:

[0061] Input the video image in the video information into the YOLOV5 network, extract the point coordinates of the target box in the video image, and extract the width and height of the target box in the video image;

[0062] It is determined that the train is passing through the first position according to the point coordinates, width and height of the target frame.

[0063] It should be noted that the present invention uses the YOLOV5 network to identify trains in video images, and uses the train target recognition results of each image to determine whether the current train is passing through the crossing area. Assume that the train recognition results in each image are (pi ,x i ,y i ,w i ,h i ), p i is the probability that a train exists in the target box, (x i ,y i ) is the coordinate of the upper left corner of the target box, (w i ,h i ) are the width and height of the target box. The YOLOV5 network can be used to identify trains in video images, which can improve the accuracy of the system and ensure the safe operation of trains.

[0064] Further, determining that the train is passing through the first position according to the point coordinates, width and height of the target frame includes:

[0065] The train existence probability in the target frame is calculated according to the point coordinates, width and height of the target frame, and a column existence probability set is generated, and the maximum value in the train existence probability set is determined as the train identification position.

[0066] Furthermore, the target recognition result through video tracking determines whether the current train is passing through the crossing area, including:

[0067] The target recognition result is normalized, and the normalized result is input into a train departure judgment network, and the train departure area is determined according to the judgment network.

[0068] Specifically, the train target recognition result is input into the following formula:

[0069] p=max(p1,…,p n )

[0070]

[0071] Among them, p i is the probability that a train exists in the target box, (x i ,y i ) is the coordinate of the upper left corner of the target box, (w i ,h i ) are the width and height of the target frame. The train leaving the crossing area is determined by the above judgment network.

[0072] See also Figure 3, the judgment network is a convolutional neural network based on one-dimensional convolution and average pooling layer to realize the judgment network of train departure. The input dimension of the judgment network is 32×5, and the convolution layer 1 includes: 8 convolution kernels with a length of 5 and a width of 5 and 8 convolution kernels with a length of 3 and a width of 5, and the step length is 1; the pooling layer 1 is an average pooling layer. The convolution layer 2 includes: 8 convolution kernels with a length of 3 and a width of 1, and the step length is 1; the pooling layer 2 is an average pooling layer. The convolution layer 3 includes: 4 convolution kernels with a length of 3 and a width of 1, and the step length is 1; the pooling layer 3 is an average pooling layer, and the output dimension is 2×1.

[0073] In some optional embodiments, determining that the train leaves the crossing area according to the judgment network includes:

[0074] The normalized train recognition results of 32 consecutive frames are input into the judgment network as feature vectors, and the input feature vectors are reduced in dimension through one-dimensional convolution and average pooling layers;

[0075] After reducing the dimension of the input feature vector, the recognition results of different sizes are output through the fully connected layer, which are the probability of the train leaving and the probability of the train not leaving. The largest probability between the probability of the train leaving and the probability of the train not leaving is determined as the judgment result.

[0076] Specifically, the input of the train departure judgment network is the normalized train recognition results of 32 consecutive frames, so the input dimension is 32×5. The network continuously reduces the dimension of the input feature vector through one-dimensional convolution and average pooling layers, and finally outputs a 2×1 recognition result (P1, P2) through a fully connected layer, which is the probability of the train leaving and the probability of the train not leaving. The result with a larger probability is taken as the current train status.

[0077] The judgment result is that the train leaves the crossing area, and it is determined to open the crossing by turning on the barrier machine and turning off the sound and light alarm.

[0078] The present invention obtains the train route status in front of the crossing, and when the train triggers the first route position, obtains the first position signal, tracks the train according to the first position signal, obtains the train position, and issues an alarm through the sound and light alarm; when the train triggers the second route position, obtains the second position signal, and controls the wood barrier to close the crossing according to the second position signal; determines whether the current train is passing through the crossing area through the target recognition result of video tracking, and determines to open the wood barrier to open the crossing and close the sound and light alarm according to the judgment result. It can be seen that the degree of automation and the accuracy of wood barrier control are improved through video tracking and the identification of the train route through the judgment network, and the degree of intelligence and convenience are improved.

[0079] Based on the same inventive concept, the present invention also provides a crossing control system based on video analysis and magnetic steel sensors, see Figure 4 ,include:

[0080] The acquisition sheet, 201, is used to acquire the train route status in front of the crossing, wherein the route status includes a first position of triggering the route and a second position of triggering the route;

[0081] An alarm unit 202 is used to obtain a first position signal when the train triggers the first position of the route, track the train according to the first position signal video to obtain the train position and issue an alarm through an audible and visual alarm;

[0082] The control unit 203 is used to obtain a second position signal when the train triggers the second approach position, and control the barrier machine to close the crossing according to the second position signal;

[0083] The judgment unit 204 is used to judge whether the current train is passing through the crossing area through the target recognition result of video tracking, and determine to open the barrier machine to open the crossing and close the sound and light alarm according to the judgment result.

[0084] Based on the same inventive concept, the present invention also provides an electronic device 161, see Figure 5 , including a processor 164, a communication interface 165, a memory 162 and a communication bus, wherein the processor 164, the communication interface 165 and the memory 162 communicate with each other through the communication bus;

[0085] A memory 162 storing a computer program 163;

[0086] When the processor 164 executes the program stored in the memory 162, a crossing control method based on video analysis and magnetic steel sensors is implemented.

[0087] The above communication bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc.

[0088] The communication interface 165 is used for communication between the electronic device 161 and other devices.

[0089] The memory 162 may include a random access memory 162 (RAM), or may include a non-volatile memory 162 (non-volatile memory), such as at least one disk memory 162. Optionally, the memory 162 may also be at least one storage device located away from the processor 164.

[0090] The above-mentioned processor 164 can be a general-purpose processor 164, including a central processing unit 164 (CPU), a network processor 164 (NP), etc.; it can also be a digital signal processor 164 (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0091] Based on the same inventive concept, the present invention further provides a computer-readable storage medium storing a computer program 163 , which, when executed by a processor 164 , implements a crossing control method based on video analysis and magnetic steel sensors.

[0092] The computer-readable storage medium may be included in the device / apparatus described in the above embodiment; or it may exist independently without being assembled into the device / apparatus. The above computer-readable storage medium carries one or more programs. When the above one or more programs are executed, the crossing control method based on video analysis and magnetic steel sensor according to the embodiment of the present disclosure is implemented.

[0093] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application 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 replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A crossing control method based on video analysis and magnetic steel sensor, characterized in that: include: Acquire the train route status in front of the crossing, wherein the route status includes triggering the first route position and triggering the second route position; When the train triggers the first position of the route, a first position signal is obtained, the train position is obtained by video tracking of the train according to the first position signal, and an alarm is issued through an audible and visual alarm; When the train triggers the second position of the route, the second position signal is obtained, and the barrier machine is controlled to close the crossing according to the second position signal; The target recognition results of video tracking are used to determine whether the current train is passing through the crossing area. Based on the determination results, the barrier machine is turned on to open the crossing and the sound and light alarm is turned off.

2. The method according to claim 1, characterized in that The approach status is obtained by a magnetic steel sensor arranged beside the rail in front of the crossing; The first position of the triggered route in the route state is obtained by a magnetic steel sensor arranged 500 meters in front of the crossing; The second position of the triggered route in the route state is obtained by a magnetic steel sensor arranged 300 meters in front of the crossing.

3. The method according to claim 1, characterized in that The method of tracking the train according to the first position signal video to obtain the train position and issuing an alarm through an audible and visual alarm includes: The train route video information is obtained, and the train route video information is input into the YOLOV5 network. The train in the video information is identified by the YOLOV5 network to determine that the train is passing through the first position; when the train passes through the first position, an audible and visual alarm is used to send out an alarm.

4. The method according to claim 3, characterized in that The step of identifying the train in the video information by using the YOLOV5 network to determine that the train is passing through the first position includes: Input the video image in the video information into the YOLOV5 network, extract the point coordinates of the target box in the video image, and extract the width and height of the target box in the video image; It is determined that the train is passing through the first position according to the point coordinates, width and height of the target frame.

5. The method according to claim 4, characterized in that Determining that the train is passing through the first position according to the point coordinates, width and height of the target frame includes: The train existence probability in the target frame is calculated according to the point coordinates, width and height of the target frame, and a column existence probability set is generated, and the maximum value in the train existence probability set is determined as the train identification position.

6. The method according to claim 1, characterized in that The target recognition result of the video tracking is used to determine whether the current train is passing through the crossing area, including: The target recognition result is normalized, and the normalized result is input into a train departure judgment network, and the train departure area is determined according to the judgment network.

7. The method according to claim 6, characterized in that The judgment network is a convolutional neural network based on one-dimensional convolution and average pooling layers to realize the judgment network of train departure.

8. The method according to claim 6, characterized in that The step of determining the train leaving the crossing area according to the judgment network includes: The normalized train recognition results of 32 consecutive frames are input into the judgment network as feature vectors, and the input feature vectors are reduced in dimension through one-dimensional convolution and average pooling layers; After reducing the dimension of the input feature vector, the recognition results of different sizes are output through the fully connected layer, which are the probability of the train leaving and the probability of the train not leaving. The largest probability between the probability of the train leaving and the probability of the train not leaving is determined as the judgment result.

9. The method according to claim 1, characterized in that: The judgment result is that the train leaves the crossing area, and it is determined to open the crossing by turning on the barrier machine and turning off the sound and light alarm.

10. A crossing control system based on video analysis and magnetic steel sensors, characterized in that: include: An acquisition unit, used for acquiring a train route status in front of the crossing, wherein the route status includes a first route triggering position and a second route triggering position; An alarm unit is used to obtain a first position signal when the train triggers the first position of the route, track the train according to the first position signal video to obtain the train position and issue an alarm through an audible and visual alarm; A control unit, used for obtaining a second position signal when the train triggers the second approach position, and controlling the barrier machine to close the crossing according to the second position signal; The judgment unit is used to judge whether the current train is passing through the crossing area through the target recognition result of video tracking, and determine to open the barrier machine to open the crossing and turn off the sound and light alarm according to the judgment result.

11. An electronic device, characterized in that: include: 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; a memory storing a computer program; The processor, when executing the program stored in the memory, implements the crossing control method based on video analysis and magnetic steel sensor as described in any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the crossing control method based on video analysis and magnetic steel sensor as described in any one of claims 1 to 9 is implemented.

Images