A track misrouting detection method and system based on video analysis

Abstract

The present application relates to a kind of track misrouting detection methods based on video analysis, comprising the following steps: S1, first picture data is collected from railway yard, and first picture data is numbered according to track area;S2, the numbered first picture data is visually analyzed, and the occupation of each track in yard is judged clear;S3, according to the track area, the track occupation clear state obtained from first picture data is compared with the corresponding track occupation clear state sent by computer interlocking system, to judge whether there is misrouting in track;S4, if the track occupation clear state obtained by visual analysis is inconsistent with the occupation clear state sent by computer interlocking system, then the track misrouting, and the bad state is alarmed and displayed.The present application has high reliability, is more intelligent, and avoids the risk of transportation accident caused by manual misjudgment of approach.

Description

Technical Field

[0001] This invention relates to the field of railway track circuit shunting, and specifically to a method and system for detecting faulty track circuits based on video analysis. Background Technology

[0002] From the perspective of actual railway track operation, faulty track circuit shunting has always been a significant factor affecting railway safety. Based on actual conditions and historical statistics, faulty track circuit shunting occurs frequently in various railway sections, especially within stations, seriously impacting normal railway operations and potentially causing major safety accidents, threatening passenger lives and property. The probability of faulty track circuit shunting is relatively high. Once a faulty track circuit causes a train to enter a corresponding section and fail to display normal information or display incorrect messages, it can directly affect normal train scheduling and safe operation.

[0003] Currently, faulty track circuit shunting is a global problem that affects the safety and reliability of train operation. Therefore, it is essential to strengthen research and analysis of faulty track circuit shunting, identify the specific causes of the problem, and take targeted measures to solve it, reducing the occurrence of faulty circuit shunting failures and providing reliable signaling systems and data support for safe train operation.

[0004] The general solutions for the problem of faulty circuit splitters are as follows:

[0005] 1) Manual cleaning of the track or track bed is carried out in a timely manner, or track inspection vehicles are used for inspection.

[0006] 2) The microcomputer monitoring system implements monitoring by detecting and maintaining the voltage of the railway track circuit, thereby achieving a rational allocation and layout of technical parameters such as voltage, and preventing significant voltage differences between different carriers.

[0007] 3) Introduce an electronic high-voltage pulse track circuit system to improve the operating environment of the railway circuit system.

[0008] However, these solutions either fail to detect faulty track circuits in a timely and dynamic manner, some require manual intervention or extensive infrastructure construction, and others suffer from unstable detection accuracy, leading to missed detections. Summary of the Invention

[0009] To address the above problems, this invention proposes a dual-system fusion method for detecting faulty circuits, which integrates the station computer interlocking system and the station video fault detection system. This method avoids the low detection rate of a single system and reduces labor and construction costs for a large amount of infrastructure.

[0010] The detection method includes the following steps:

[0011] S1. Collect the first image data from the railway station and number the first image data according to the track area;

[0012] S2. Perform visual analysis on the first image data with the number to determine the occupancy and clearance status of each track in the station, and determine the area to which it belongs based on the number;

[0013] S3. Based on the track area, compare the track occupancy and clearance status obtained from the first image data with the corresponding track occupancy and clearance status sent by the computer interlocking system to determine whether there is a track malfunction.

[0014] S4. If the track occupancy clearance status obtained by visual analysis is inconsistent with the occupancy clearance status issued by the computer interlocking system, the track routing is faulty and an alarm will be displayed for this faulty status; if the occupancy clearance status of the two is consistent, the track routing is normal and no alarm will be issued.

[0015] Specifically, S1 includes the following steps:

[0016] S11. The station camera sends track video data to the video acquisition unit, which performs video data processing including filtering, noise reduction, and grayscale conversion.

[0017] S12. The video capture machine converts the video data into first image data and numbers the first image data.

[0018] Specifically, S2 further includes the following steps:

[0019] S21. Determine the track region to which the first image data belongs according to the numbering rules.

[0020] S22. Combining the second image data of the static station, machine learning and deep learning algorithms are used to separate the vehicles and background of the track section from the collected first image data in order to reduce the interference of background and lights;

[0021] S23. Obtain the occupancy and clearing status of each track from the first image data.

[0022] Specifically, the bad track information in S3 refers to the situation where the track section is actually occupied, but the computer interlocking system obtains that the track section is in a cleared state.

[0023] Specifically, S4 includes the following:

[0024] S41. Send the information on faulty tracks and station interlocking code information to the monitoring and early warning terminal.

[0025] S42. After obtaining information about faulty circuits, the monitoring and early warning terminal displays the station layout map and code position information in real time and issues an alarm.

[0026] Furthermore, the detection method also includes: S5, sending alarm information of the faulty track to a microcomputer monitoring system and a CTC system.

[0027] To implement the above detection method, this invention also proposes a track shunting defect detection system based on video analysis, which includes:

[0028] Cameras are installed at both ends of the station track to monitor the track.

[0029] A video capture device, connected to a camera, is used to initially process the video data transmitted from the camera and convert the video data into first image data;

[0030] The analysis server, connected to the video capture machine, is used to analyze and process the first image data to obtain the occupancy and clearing status of each track;

[0031] The diagnostic server is connected to both the analysis server and the computer interlocking system. It receives the track occupancy and clearance status sent by the analysis server and compares it with the track occupancy and clearance status sent by the computer interlocking system to analyze the bad conditions of the track.

[0032] The monitoring and early warning terminal is connected to the diagnostic server. It is used to receive adverse statuses issued by the diagnostic server and display the station layout map in real time. If an adverse status is received, an alarm will be issued.

[0033] Furthermore, the cameras are numbered according to the area where they are located.

[0034] Furthermore, the video acquisition machine assigns a number to each of the first image data converted from the video data according to the camera number.

[0035] Furthermore, the analysis server is connected to the computer interlocking system via an interlocking interface machine.

[0036] Furthermore, the detection system is also connected to the microcomputer monitoring system and the CTC system; the monitoring and early warning terminals are all connected to the microcomputer monitoring system and the CTC system, and are used to send alarm information of adverse conditions to the microcomputer monitoring system and the CTC system.

[0037] In summary, the monitoring system and method proposed in this invention combine a video intelligent analysis system and a computer interlocking system. By combining traditional and artificial intelligence technologies, the accuracy of fault detection is improved. A monitoring and early warning terminal for power supply and train operation personnel is proposed, enabling real-time detection of faulty circuits and sending alarms to relevant maintenance and train route handling personnel. Faulty circuit warning information can also be transmitted to the microcomputer monitoring system and CTC system via an interface device, avoiding the risk of transportation accidents caused by human error in route judgment. Furthermore, a video data sharing strategy based on video acquisition equipment is proposed, allowing a single station video acquisition system to meet various application needs and reduce the construction cost of a single system. Attached Figure Description

[0038] Figure 1 This is a block diagram of the detection system of the present invention;

[0039] Figure 2 This is a diagram showing the layout of the station cameras according to the present invention.

[0040] Figure 3 This is a flowchart of the detection method of the present invention. Detailed Implementation

[0041] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a further detailed explanation of the video analysis-based method and system for detecting track shunting defects proposed in this invention.

[0042] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0043] This method proposes a track alignment failure detection method based on video analysis, which can realize real-time and accurate detection of track occupancy and clearance status within the station. Combined with the station's computer interlocking system, it can promptly detect track alignment failure sections in the interlocking system and send track alignment failure alarms to relevant personnel in the electrical and train operations departments to avoid train operation accidents such as route misoperation caused by track alignment failure.

[0044] like Figure 1As shown, this detection method is based on a detection system, which includes a camera 1, a video acquisition unit 2, an analysis server 3, a diagnostic server 4, and a monitoring and early warning terminal 5. Figure 2 As shown, the cameras 1 are installed at both ends of each track in the station yard. Each camera 1 is numbered according to its corresponding track area and is used to monitor each track. The video acquisition unit 2 is connected to the cameras 1 and performs preliminary processing and analysis on the video signals acquired by the cameras 1, such as filtering, noise reduction, and grayscale conversion. Then, it is converted into first image data in a fixed format and numbered according to the acquisition source. The numbered first image data is then sent to the analysis server 3. After receiving the transmitted first image data, the analysis server 3 determines the track area to which the first image data belongs according to the numbering rules and time. At the same time, it combines a large amount of static second image data of the station yard as background data and uses machine learning and deep learning algorithms to separate the vehicles and background in the track section, remove background interference and light interference in the first image data, and thus accurately identify the occupancy and clearance status of the track section in a fixed area from the first image data. Then, the output track section occupancy and clearance status is forwarded to the diagnostic server 4. The diagnostic server 4 is connected to the computer interlocking system 6 through the interlocking interface machine. The system obtains station interlocking information from the computer interlocking system 6, and combines it with the second image data of the station track section static station yard, the track occupancy clearance status in the interlocking information, and the track section occupancy clearance information obtained from the analysis server 3 to determine whether the actual track section is occupied. If it is found that the track section is actually occupied by a vehicle or locomotive, but the track section obtained from the computer interlocking system 6 is in a cleared state, then the track section is judged to be in a poor routing state. At this time, the diagnostic server 4 sends the calculated poor routing state of the track section and the station interlocking code position information to the monitoring and early warning terminal 5 through the network. After obtaining the poor routing information, the monitoring and early warning terminal 5 displays the station yard map and code position information in real time, and at the same time issues an alarm for the section with poor routing, and issues a poor routing alarm to the train operation and signaling personnel. The monitoring and early warning terminal 5 is also connected to the microcomputer monitoring system 7 or CTC system 8 through the interface machine, and sends the poor routing track information and alarm information to the microcomputer monitoring system 7 or CTC system 8 (dispatch centralized system) and other systems.

[0045] To achieve effective acquisition and visualization of video information from all areas of the station, the station track area can be segmented and the layout of video acquisition equipment optimized. This solution divides the track areas according to the track section distribution on the station map and the actual station conditions, using a nine-square grid method. A schematic diagram of the specific cameras installed at the station and the accompanying images is shown below. Figure 2As shown, video capture zones are set up in the station according to tracks, non-intersection sections at station boundaries, and station throat areas. Cameras with different viewing distances and wide angles are deployed in the capture zones based on the camera's capture distance and range. Considering the long track length and the inability to capture video information from the other end of the track when vehicles obstruct the view, cameras are installed at both ends of the track to achieve full track coverage. A single camera can cover a single track or multiple tracks, depending on the actual camera capture range and image size. However, considering design and construction costs, a reasonable model is created based on the actual camera performance and the size of the station map to calculate the minimum camera installation scheme, thereby reducing construction costs.

[0046] The commonly used video area division method is based on the nine-grid area division method. This method is implemented using layout files, which fill the entire station's monitoring screen and station site map with M rows and N columns of cells. The division area can be determined by setting appropriate cells.

[0047] Meanwhile, based on the actual situation of railway stations and in conjunction with station routes, we have integrated and improved the method of dividing the area. When delineating cell areas, we have fully considered the interlocking route routing relationship, the section to which the turnout belongs, and other factors. We have integrated the division of areas and divided related areas into unified cells (or the same defined area), reducing the number of station area divisions and increasing the rationality of the division.

[0048] by Figure 2 Taking the track distribution as an example, the actual camera installation locations and layouts are divided below. The camera numbering and naming rule is: "Station Number_Track Number_Orientation". For example, if the station number is 13, the track number is 6G, and the camera is installed on the right, the orientation is R, therefore the camera number is 13_6G_R.

[0049] The video capture device 2 can easily identify the video area captured by the camera 1 and the corresponding station track area by the camera number.

[0050] By dividing the station map into a nine-square grid and combining it with the actual track layout, Figure 2 The entire station layout is divided into 5 rows and 4 columns. Cameras are deployed on both sides according to the 6 tracks and the entrance throat area. Considering that the middle tracks IG and IIG are mainline tracks, a single wide-range camera is used to cover these two tracks. Cameras are installed on both sides of the other tracks to provide full coverage of the station tracks. Cameras are installed on the tracks SSG, SMG, XJG, and SJG in the non-interchangeable section at the station boundary for area data collection. The cameras in the boundary area can also perform video tracking of entering vehicles and locomotives.

[0051] Based on the above detection system, this invention proposes a method for detecting track shunting defects based on video analysis. (See attached document.) Figure 3 The method includes the following:

[0052] S1. Collect the first image data from the railway station and number the first image data according to the track area;

[0053] S2. Perform visual analysis on the first image data with the number to determine the occupancy and clearance status of each track in the station, and determine the area to which it belongs based on the number;

[0054] S3. Based on the area to which the track belongs, compare it with the corresponding track occupancy and clearance status issued by the computer interlocking system to determine whether there is a problem with track routing.

[0055] S4. If the track occupancy clearance status obtained by visual analysis is inconsistent with the occupancy clearance status issued by the computer interlocking system, the track routing is faulty and an alarm will be displayed for this faulty status; otherwise, the track routing is normal and no alarm will be issued.

[0056] Further, S1 includes the following steps:

[0057] S11. The station camera sends video data to the video acquisition unit, which performs filtering, noise reduction, and grayscale conversion on the video data.

[0058] S12. The video capture machine converts the video data into first image data and numbers the first image data.

[0059] Furthermore, S2 also includes the following steps:

[0060] S21. Determine the track region to which the first image data belongs according to the numbering rules.

[0061] S22. Using the second image data of the static station, machine learning and deep learning algorithms are used to separate the vehicles and background of the track section from the collected first image data, reducing interference from background and lighting.

[0062] S23. Obtain the occupancy and clearing status of each track from the first image data.

[0063] Furthermore, the bad track information in S3 refers to the situation where the track section is actually occupied, but the computer interlocking system 6 obtains that the track section is in a cleared state.

[0064] Furthermore, S4 includes the following: sending information on faulty tracks and alarm information to the microcomputer monitoring system 7 and the CTC system 8.

[0065] This method proposes to combine the station's computer interlocking system and video intelligent analysis technology to identify and resolve the problem of faulty track switching that has long plagued railway transportation safety. This method is highly feasible and reliable. With the development of railway intelligence, technologies such as video intelligent analysis and artificial intelligence will inevitably bring revolutionary solutions and approaches to difficult railway problems.

[0066] In summary, the monitoring system and method proposed in this invention combine a video intelligent analysis system and a computer interlocking system. By combining traditional and artificial intelligence technologies, the accuracy of fault detection is improved. A monitoring and early warning terminal for power supply and train operation personnel is proposed, enabling real-time detection of faulty circuits and sending alarms to relevant maintenance and train route handling personnel. Faulty circuit warning information can also be transmitted to the microcomputer monitoring system and CTC system via an interface device, avoiding the risk of transportation accidents caused by human error in route judgment. Furthermore, a video data sharing strategy based on video acquisition equipment is proposed, allowing a single station video acquisition system to meet various application needs and reduce the construction cost of a single system.

[0067] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A method for detecting track shunting defects based on video analysis, characterized in that, Includes the following steps: S1. Collect the first image data from the railway station and number the first image data according to the track area; S1 includes the following steps: S11, The station camera (1) sends the track video data to the video acquisition machine (2), and the video acquisition machine (2) performs processing on the video data, including filtering, noise reduction and grayscale conversion. S12, The video acquisition machine (2) converts the video data into first image data and numbers the first image data; The method also includes dividing the track area using a nine-square grid method and numbering the first image data according to the battlefield camera number; S2. Perform visual analysis on the first image data with the number to determine the occupancy and clearance status of each track in the station, and determine the area to which it belongs based on the number; S2 further includes the following steps: S21. Determine the track region to which the first image data belongs according to the numbering rules. S22. Combining the second image data of the static station, machine learning and deep learning algorithms are used to separate the vehicles and background of the track section from the collected first image data in order to reduce the interference of background and lights; S23. Obtain the occupancy and clearance status of each track from the first image data; S3. Based on the track area, compare the track occupancy and clearance status obtained from the first image data with the corresponding track occupancy and clearance status sent by the computer interlocking system to determine whether there is a track malfunction. S4. If the track occupancy clearance status obtained by visual analysis is inconsistent with the occupancy clearance status issued by the computer interlocking system, the track routing is faulty and an alarm will be displayed for this faulty status; if the occupancy clearance status of the two is consistent, the track routing is normal and no alarm will be issued.

2. The method for detecting track shunting defects based on video analysis as described in claim 1, characterized in that, The bad track information in S3 refers to the track section being actually occupied, but the computer interlocking system (6) obtains that the track section is in a cleared state.

3. The method for detecting track shunting defects based on video analysis as described in claim 1, characterized in that, S4 includes the following: S41. Send the information on the faulty track and the station interlocking code to the monitoring and early warning terminal (5). S42. After obtaining the fault information of the branch, the monitoring and early warning terminal (5) displays the station site map and code position information in real time and issues an alarm.

4. The method for detecting track shunting defects based on video analysis as described in claim 3, characterized in that, It also includes: S5, sending alarm information of faulty track to the microcomputer monitoring system (7) and CTC system (8).

5. A video analytics-based track shunting defect detection system, used to implement the detection method described in any one of claims 1 to 4, characterized in that, The detection system includes: Camera (1) is installed at both ends of the station track to monitor the track; The video acquisition unit (2) is connected to the camera (1) and is used to initially process the video data transmitted from the camera (1) and convert the video data into first image data; The analysis server (3) is connected to the video capture machine (2) and is used to analyze and process the first image data to obtain the occupancy and clearing status of each track; The diagnostic server (4) is connected to the analysis server (3) and the computer interlocking system (6) respectively. It is used to receive the track occupancy and clearance status sent by the analysis server (3) and compare it with the track occupancy and clearance status sent by the computer interlocking system (6) to analyze the bad status of the track. The monitoring and early warning terminal (5) is connected to the diagnostic server (4) to receive the adverse status issued by the diagnostic server (4) and display the station site map in real time. If an adverse status is received, an alarm is issued.

6. The track shunting defect detection system based on video analysis as described in claim 5, characterized in that, The cameras are numbered according to the area where the camera (1) is located.

7. The track shunting defect detection system based on video analysis as described in claim 6, characterized in that, The video acquisition machine (2) numbers each first image data converted from video data according to the number of the camera (1).

8. The track shunting defect detection system based on video analysis as described in claim 5, characterized in that, The analysis server (3) is connected to the computer interlocking system (6) through the interlocking interface machine.

9. The track shunting defect detection system based on video analysis as described in claim 5, characterized in that, The detection system is also connected to the microcomputer monitoring system (7) and the CTC system (8); the monitoring and early warning terminal (5) is connected to the microcomputer monitoring system (7) and the CTC system (8) to send alarm information of bad status to the microcomputer monitoring system (7) and the CTC system (8).

Images