Front-end monitoring method and system based on vibration optical fiber and video fusion

Abstract

The invention discloses a front-end monitoring method and system based on vibration optical fiber and video fusion, and the method comprises the steps: employing an armored single-mode special optical cable sensitive to an acoustic signal as a sensor, and enabling the armored single-mode special optical cable to be disposed at the middle part and the bottom part of a rolling cage and a fence at two sides of a line or to be buried in the ground, so as to monitor an external vibration signal in real time. The cameras are installed on the vertical rod near the guardrail, the installation height is 3.5 m, and the cameras are arranged along the fence at intervals of 200 m so as to collect videos and images on site in real time. And identifying the vibration optical fiber signals, analyzing the probability that the vibration optical fiber signals are working conditions such as personnel, trains, motor vehicles, wind, rain, vegetation shaking and the like, and generating a vibration optical fiber alarm result. And performing deep learning related algorithm processing on the video and the image, identifying various objects in the scene, and generating a video alarm result. Therefore, by fusing the advantages of the vibration optical fiber and the video image, accurate identification and positioning of the intrusion behavior are realized, and the overall performance of the monitoring system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of security monitoring, and in particular to a front-end monitoring method and system based on vibration optical fiber and video fusion. Background Art

[0002] In the safety monitoring of long lines such as railway roadbed sections, perimeter protection is an important part of ensuring the safe operation of railways. Traditional monitoring methods such as infrared radiation and microwave detection are susceptible to environmental interference, high false alarm rates, and insufficient positioning accuracy. The vibration fiber optic technology, with its electrical insulation, corrosion resistance, long signal transmission distance, low loss, and strong anti-electromagnetic interference capabilities, has shown significant advantages in the long roadbed monitoring scenario. However, the environment along the railway is complex, and external factors such as train vibration, motor vehicle passing, and strong winds will interfere with the vibration fiber optic, resulting in false alarms or missed alarms, affecting the monitoring effect.

[0003] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0004] The purpose of the present invention is to provide a front-end monitoring method and system based on the fusion of vibrating optical fiber and video, which aims to solve the problems of high false alarm rate, low missed alarm rate, insufficient positioning accuracy, etc. in the perimeter monitoring of railway roadbed sections in the prior art. By integrating the advantages of vibrating optical fiber and video images, accurate identification and positioning of intrusion behaviors can be achieved, thereby improving the overall performance of the monitoring system.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a front-end monitoring method based on vibrating optical fiber and video fusion, comprising:

[0006] Use armored single-mode special optical cables that are sensitive to acoustic signals as sensors and deploy them in the middle and bottom of the rolling cages, fences or buried underground on both sides of the line to monitor external vibration signals in real time;

[0007] Cameras are installed on poles near the guardrail at a height of 3.5 meters and deployed every 200 meters along the fence to collect real-time video and images of the scene;

[0008] Identify the vibration optical fiber signal, analyze the probability of it being a person, a train, a motor vehicle, wind, rain, vegetation shaking, etc., and generate a vibration optical fiber alarm result;

[0009] Process the video and image using deep learning-related algorithms to identify various objects in the scene and generate video alarm results;

[0010] Different weight coefficients are set up in combination with the vibration optical fiber alarm result, the video alarm result and weather conditions, and a comprehensive analysis is performed to finally make a joint decision alarm.

[0011] In one embodiment of the present invention, the front-end monitoring method based on vibrating optical fiber and video fusion further includes:

[0012] The front-end monitoring equipment sends real-time monitoring data to the monitoring center;

[0013] When an alarm occurs, the front-end monitoring device sends alarm data to the monitoring center until receiving automatic confirmation information of alarm reception from the monitoring center;

[0014] After the alarm is lifted, the front-end monitoring device sends the alarm video to the monitoring center;

[0015] The front-end monitoring device periodically sends device status data to the monitoring center.

[0016] In one embodiment of the present invention, the front-end monitoring method based on vibration optical fiber and video fusion also includes: the front-end monitoring equipment uses the monitoring center to which it belongs as the time source, obtains the time synchronization signal based on the NTP protocol, performs time synchronization with the NTP server of the monitoring center once a day, and immediately makes a time synchronization request when detecting its own time abnormality.

[0017] In one embodiment of the present invention, the front-end monitoring method based on vibrating optical fiber and video fusion also includes: periodic self-inspection of equipment status, sending alarm information and triggering sound and light alarm in case of abnormality, the optical cable has the functions of both sensor and communication equipment, when the optical cable is damaged, the broken fiber position is located according to the characteristics of the optical fiber signal at the broken fiber location, and the broken fiber positioning time is less than 3 seconds.

[0018] In one embodiment of the present invention, the front-end monitoring method based on vibration optical fiber and video fusion also includes: utilizing the video resources of the existing comprehensive video surveillance system, accessing them to the intelligent analysis host through the B / H interface, combining the vibration optical fiber alarm information and the intelligent analysis algorithm, determining whether there is an intrusion target in the scene, and pushing the alarm results to the perimeter intrusion alarm monitoring platform.

[0019] In the second aspect, the present invention provides a front-end monitoring system based on vibration optical fiber and video fusion, including: a monitoring module, an acquisition module, a first generation module, a second generation module and an alarm module. The monitoring module is used to use an armored single-mode special optical cable that is sensitive to acoustic signals as a sensor, and deploy it in the middle and bottom of the rolling cage, fence or buried on both sides of the line to monitor the external vibration signal in real time. The acquisition module is used to install a camera on a pole near the guardrail, with an installation height of 3.5 meters, and deploy it at intervals of every 200 meters along the fence to collect real-time video and images of the scene. The first generation module is used to identify the vibration optical fiber signal, analyze the probability of it being a person, train, motor vehicle, wind, rain, vegetation shaking and other working conditions, and generate a vibration optical fiber alarm result. The second generation module is used to perform deep learning related algorithm processing on the video and image, identify various objects in the scene, and generate a video alarm result. And the alarm module is used to combine the vibration optical fiber alarm result, the video alarm result and the weather conditions, set different weight coefficients, conduct a comprehensive analysis, and finally make a joint decision alarm.

[0020] In one embodiment of the present invention, the front-end monitoring system based on vibration optical fiber and video fusion also includes: a first sending module, a second sending module, a third sending module and a fourth sending module. The first sending module is used for the front-end monitoring device to send real-time monitoring data to the monitoring center. The second sending module is used for the front-end monitoring device to send alarm data to the monitoring center when an alarm occurs, until receiving the automatic confirmation information of the alarm reception from the monitoring center. The third sending module is used for the front-end monitoring device to send alarm video to the monitoring center. The fourth sending module is used for the front-end monitoring device to periodically send device status data to the monitoring center.

[0021] In one embodiment of the present invention, the front-end monitoring system based on vibration optical fiber and video fusion also includes: a synchronization module, which is used for the front-end monitoring equipment to use the monitoring center to which it belongs as the time source, obtain the time synchronization signal based on the NTP protocol, and perform time synchronization with the NTP server of the monitoring center once a day, and immediately make a time synchronization request when detecting its own time abnormality.

[0022] In one embodiment of the present invention, the front-end monitoring system based on vibrating optical fiber and video fusion also includes: a positioning module, which is used for periodic self-inspection of equipment status, sending alarm information and triggering sound and light alarms when abnormalities occur. The optical cable has the functions of both sensor and communication equipment. When the optical cable is damaged, the broken fiber position is located according to the characteristics of the optical fiber signal at the broken fiber location, and the broken fiber positioning time is less than 3 seconds.

[0023] In one embodiment of the present invention, the front-end monitoring system based on vibration optical fiber and video fusion also includes: a push module, which is used to utilize the video resources of the existing comprehensive video surveillance system, access to the intelligent analysis host through the B / H interface, combine the vibration optical fiber alarm information and the intelligent analysis algorithm, determine whether there is an intrusion target in the scene, and push the alarm result to the perimeter intrusion alarm monitoring platform.

[0024] Compared with the prior art, the front-end monitoring method and system based on vibrating optical fiber and video fusion according to the present invention has the following beneficial effects:

[0025] 1. By integrating the information of vibrating optical fiber and video images, the present invention can accurately identify and locate intrusion behaviors and improve monitoring accuracy;

[0026] 2. By adopting advanced signal processing and fusion algorithms, the present invention can effectively filter environmental interference signals, reduce false alarm rate and missed alarm rate, and improve the reliability of the monitoring system;

[0027] 3. The present invention is suitable for perimeter monitoring of railway roadbed sections under different terrain and climatic conditions, and has strong adaptability and flexibility;

[0028] 4. By utilizing the video resources of the existing integrated video surveillance system, the present invention can realize the reuse of old equipment, reduce construction costs, and improve the overall benefits of the monitoring system;

[0029] 5. The present invention can realize real-time monitoring and alarm, provide timely and accurate information support for railway safety management departments, and help to quickly respond to and handle intrusion incidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of a front-end monitoring method based on vibrating optical fiber and video fusion in Embodiment 1 of the present invention;

[0031] Figure 2 It is a structural schematic diagram of a front-end monitoring system based on vibration optical fiber and video fusion in embodiment 2 of the present invention. DETAILED DESCRIPTION

[0032] The embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the embodiments of the present invention, rather than to limit the embodiments of the present invention. It is also necessary to explain that, for ease of description, only parts related to the embodiments of the present invention are shown in the accompanying drawings, rather than all structures.

[0033] To facilitate understanding, the main implementation concepts of the embodiments of the present invention are first briefly described.

[0034] The roadbed section has a long line length. In this scenario, the main risk to be prevented by the perimeter is human intrusion, and the main intrusion methods are to destroy the barbed wire cage and the fence. The vibration fiber technology itself has inherent properties such as electrical insulation, corrosion resistance, long signal transmission distance, and low loss. At the same time, the vibration fiber has strong resistance to electromagnetic interference and is not affected by high voltage and high current. Since optical fiber does not require additional electricity and transmission network, is easy to install, and is relatively soft and easy to combine with physical protection facilities along the line, it has obvious cost advantages over other technologies in the long roadbed monitoring scenario. The advantages of vibration light itself are as follows:

[0035] 1. Using optical fiber sensors as the detection means, it has low power consumption per unit distance, strong anti-electromagnetic interference ability, high spatial resolution and strong environmental adaptability.

[0036] 2. The irradiated cross-linked optical cable has strong resistance to weathering and corrosion, long service life and easy maintenance.

[0037] 3. The optical cable is used as both a sensor and a communication line, saving cable costs.

[0038] 4. Design optical cable laying plans according to different terrains, with low laying costs and flexible and feasible laying methods.

[0039] However, the difficulty of vibration optical fiber + video front-end monitoring technology lies in the fact that the environment along the railway is complex, and train vibration, motor vehicle passing, strong winds, etc. will cause certain interference to the vibration optical fiber. At the same time, meeting the requirements of zero missed reports and low false alarms poses great challenges to the equipment hardware and processing algorithms.

[0040] By discovering the defects of the prior art as mentioned above, the inventors provide a front-end monitoring method and system based on the fusion of vibrating optical fiber and video, aiming to solve the problems of high false alarm rate, low missed alarm rate, insufficient positioning accuracy, etc. in the perimeter monitoring of railway roadbed sections in the prior art. By integrating the advantages of vibrating optical fiber and video images, accurate identification and positioning of intrusion behaviors can be achieved, thereby improving the overall performance of the monitoring system.

[0041] Embodiment 1

[0042] Figure 1 is a flow chart of a front-end monitoring method based on vibrating optical fiber and video fusion in Embodiment 1 of the present invention, such as Figure 1 As shown, embodiment 1 provides a front-end monitoring method based on vibrating optical fiber and video fusion, including:

[0043] Step S100, using an armored single-mode optical cable that is sensitive to acoustic signals as a sensor, and deploying it in the middle and bottom of the rolling cage, fence or buried in the ground on both sides of the line to monitor external vibration signals in real time;

[0044] Specifically, first of all, the armored single-mode special optical cable that is sensitive to acoustic signals is selected as a sensor based on its excellent performance characteristics. This optical cable not only has electrical insulation and corrosion resistance, but also maintains low loss when transmitting signals over long distances. At the same time, it has strong anti-electromagnetic interference ability and can effectively resist the interference of complex environmental factors such as high voltage and high current. These characteristics make this optical cable an ideal monitoring sensor in complex environments such as railway roadbed sections. Secondly, the optical cable is deployed in the middle and bottom of the rolling cages and fences on both sides of the line or buried underground to achieve all-round monitoring of perimeter intrusion behavior. As important physical protection facilities in the railway roadbed section, the location of the rolling cages and fences is directly related to the coverage and monitoring effect of the monitoring system. By fitting the optical cable closely to these facilities, it can be ensured that any attempt to destroy or cross can be perceived in time. At the same time, according to different terrains and monitoring needs, the laying method of the optical cable can be flexibly adjusted, such as underground laying, to further improve the adaptability and flexibility of the monitoring system. Finally, real-time monitoring of external vibration signals is the core purpose of this step. The vibration signals collected by the optical cable can be used to analyze the behavioral characteristics of the intruder, such as climbing, crossing, digging, etc., so as to accurately identify and locate the intrusion behavior. This real-time monitoring capability is crucial for railway safety management departments. It can help staff quickly respond to and handle intrusion incidents and effectively ensure the safety and stability of railway operations.

[0045] Based on the above analysis, it can be seen that using armored single-mode special optical cables that are sensitive to acoustic signals as sensors and deploying them in the middle and bottom of the rolling cages and fences on both sides of the line or laying them underground to monitor external vibration signals in real time is a key technical means in the present invention. It not only improves the accuracy and reliability of the monitoring system, but also enhances the adaptability and flexibility of the system, providing an efficient and reliable solution for perimeter monitoring in complex environments such as railway roadbed sections.

[0046] Step S200, installing cameras on poles near the guardrail at a height of 3.5 meters and deploying them at intervals of 200 meters along the fence to collect real-time video and images of the scene;

[0047] Specifically, the camera is installed on a pole near the guardrail. This location selection ensures that the camera's field of view covers key areas of the railway embankment, such as physical protection facilities such as rolling cages and fences, thereby achieving direct monitoring of intrusions. The installation height of 3.5 meters is carefully designed to ensure that the camera can capture sufficiently clear and comprehensive image information while avoiding the problem of being easily damaged or blocked due to installation too low. The cameras are deployed every 200 meters along the fence to ensure the monitoring range while reasonably controlling costs. This interval deployment method can not only achieve full-line monitoring of the railway embankment section, but also ensure that it can quickly locate and respond when an intrusion occurs. By collecting real-time video and images of the scene, the camera can capture the behavioral characteristics of the intruder, such as climbing, climbing, digging, etc., providing key evidence for the subsequent identification and location of intrusions. In addition, this deployment method is also convenient for linkage with vibration fiber optic sensors. When the vibration fiber optic sensor detects a suspicious vibration signal, it can immediately trigger a nearby camera to take photos or videos, thereby achieving double verification of intrusions. This integrated monitoring method not only improves the accuracy and reliability of the monitoring system, but also reduces the false alarm rate and missed alarm rate, providing railway safety management departments with a more comprehensive and effective monitoring method.

[0048] Based on the above analysis, it can be seen that installing cameras on poles near the guardrail and deploying them along the fence at reasonable intervals is one of the key steps in the present invention to achieve real-time video monitoring. It can not only directly monitor the key areas of the railway roadbed section, but also complement the vibration fiber optic sensor to jointly improve the overall performance of the monitoring system.

[0049] Step S300, identifying the vibration optical fiber signal, analyzing the probability of it being a person, a train, a motor vehicle, wind, rain, vegetation shaking, etc., and generating a vibration optical fiber alarm result;

[0050] Specifically, this step uses advanced signal processing technology to conduct an in-depth analysis of the vibration signals collected by the vibration fiber sensor to identify different working conditions that cause vibration. The vibration fiber sensor can monitor various external vibration signals in real time. These signals come from natural environmental factors such as personnel intrusion, train operation, motor vehicle traffic, wind and rain, and non-invasive interference such as vegetation shaking. Since the vibration signals generated by these working conditions differ in characteristics such as frequency, amplitude, and duration, these signals can be accurately distinguished through signal processing algorithms. During the signal processing process, the system will perform preprocessing such as filtering and denoising on the vibration signal to improve the signal-to-noise ratio and recognition accuracy of the signal. Subsequently, the system uses advanced technologies such as pattern recognition and machine learning to extract and classify the processed vibration signals. These features may include, for example, the frequency component, amplitude change, duration, etc. of the signal. By comparing the signal feature library of known working conditions, the system can determine the working condition that the current vibration signal is most likely to correspond to. After determining the working condition, the system will generate the corresponding vibration fiber alarm result. For high-risk conditions such as personnel intrusion, the system will immediately trigger the alarm mechanism and upload the alarm information to the monitoring center so that the safety management personnel can respond and handle it in time. At the same time, for normal conditions such as train operation and motor vehicle traffic, the system will also record and analyze them to optimize the performance of the monitoring system and reduce the false alarm rate. In addition, this step also takes into account the impact of different weather conditions on vibration signal recognition. For example, in strong winds or heavy rain, the vibration signals generated by natural environmental factors may be enhanced, thereby increasing the risk of false alarms. Therefore, the system will adjust the threshold and weight coefficient of condition recognition according to the real-time weather conditions to improve the accuracy and reliability of recognition.

[0051] Based on the above analysis, it can be seen that identifying and analyzing the vibration fiber signal and generating the vibration fiber alarm result are the key steps to achieve accurate monitoring and alarm in the present invention. Through this step, the system can accurately distinguish the vibration signals under different working conditions, effectively filter the environmental interference signals, reduce the false alarm rate and missed alarm rate, and improve the overall performance and reliability of the monitoring system.

[0052] Step S400, performing deep learning-related algorithm processing on the video and image, identifying various objects in the scene, and generating a video alarm result;

[0053] Specifically, in the processing of the deep learning algorithm, the system first pre-processes the video and image, including image enhancement, denoising, scaling and other operations, to improve the accuracy and efficiency of the subsequent recognition algorithm. Subsequently, the system uses pre-trained deep learning models, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to extract and classify the processed images. These deep learning models can automatically learn complex features in images, such as edges, textures, shapes, etc., to achieve accurate recognition of different objects. During the recognition process, the system will further analyze and judge the identified objects according to preset thresholds and rules. For example, when a person is identified to enter the preset warning area, the system will immediately trigger the alarm mechanism, generate video alarm results, and upload relevant information to the monitoring center. At the same time, the system will also record and count other identified objects for subsequent data analysis and processing. In addition, this step also fully considers the impact of different weather and lighting conditions on video image recognition. By introducing adaptive adjustment algorithms and reinforcement learning technology, the system can automatically adjust the recognition parameters and thresholds according to real-time weather and lighting conditions to ensure that high recognition accuracy and stability can be maintained in different environments.

[0054] Based on the above analysis, it can be seen that deep learning-related algorithm processing of videos and images, identifying various objects in the scene, and generating video alarm results are key steps in realizing intelligent monitoring and accurate alarm in the present invention. Through this step, the system can achieve rapid response and effective processing of intrusion behaviors, providing strong technical support and guarantee for railway safety management departments.

[0055] Step S500, combining the vibration optical fiber alarm result, the video alarm result and the weather conditions, setting different weight coefficients, conducting a comprehensive analysis, and finally making a joint decision alarm;

[0056] Specifically, the system sets different weight coefficients for each information source. These weight coefficients are determined based on historical data, experimental verification, and experience accumulated in actual application scenarios, and are intended to reflect the reliability and importance of different information sources in specific environments. For example, in severe weather conditions with wind and rain, the vibration fiber optic sensor may be greatly disturbed, resulting in an increase in the false alarm rate. At this time, the system may reduce the weight of the vibration fiber optic alarm result and rely more on the video image analysis results. On the contrary, in clear and windless weather, the accuracy of the vibration fiber optic sensor may be higher, so the weight of its alarm result will also increase accordingly. By comprehensively analyzing these information sources and combining them with real-time weather conditions, the system can more accurately determine whether there is a real intrusion behavior and reduce the occurrence of false alarms and missed alarms. When multiple information sources point to the same intrusion event, the system will trigger the alarm mechanism, generate a joint decision-making alarm result, and report the relevant information to the monitoring center in a timely manner. This comprehensive analysis and joint decision-making method not only improves the accuracy and reliability of the monitoring system, but also provides more comprehensive and effective information support for railway safety management departments, helping them to quickly respond to and handle potential intrusion events.

[0057] In this embodiment, the front-end monitoring method based on vibrating optical fiber and video fusion further includes:

[0058] The front-end monitoring equipment sends real-time monitoring data to the monitoring center;

[0059] When an alarm occurs, the front-end monitoring device sends alarm data to the monitoring center until receiving automatic confirmation information of alarm reception from the monitoring center;

[0060] After the alarm is lifted, the front-end monitoring device sends the alarm video to the monitoring center;

[0061] The front-end monitoring device periodically sends device status data to the monitoring center.

[0062] In this embodiment, the front-end monitoring method based on vibration optical fiber and video fusion also includes: the front-end monitoring equipment uses the monitoring center to which it belongs as the time source, obtains the time synchronization signal based on the NTP protocol, performs time synchronization with the NTP server of the monitoring center once a day, and immediately makes a time synchronization request when detecting its own time abnormality.

[0063] In this embodiment, the front-end monitoring method based on vibrating optical fiber and video fusion also includes: periodic self-inspection of equipment status, sending alarm information and triggering sound and light alarm in case of abnormality, the optical cable has the functions of both sensor and communication equipment, and when the optical cable is damaged, the broken fiber position is located according to the characteristics of the optical fiber signal at the broken fiber location, and the broken fiber positioning time is less than 3 seconds.

[0064] In this embodiment, the front-end monitoring method based on vibration optical fiber and video fusion also includes: utilizing the video resources of the existing comprehensive video surveillance system, accessing them to the intelligent analysis host through the B / H interface, combining the vibration optical fiber alarm information and the intelligent analysis algorithm, determining whether there is an intrusion target in the scene, and pushing the alarm results to the perimeter intrusion alarm monitoring platform.

[0065] Embodiment 2

[0066] Figure 2 is a schematic diagram of the structure of a front-end monitoring system based on vibration optical fiber and video fusion in the second embodiment of the present invention, such as Figure 2 As shown, embodiment 2 provides a front-end monitoring system based on vibration optical fiber and video fusion, including: a monitoring module, an acquisition module, a first generation module, a second generation module and an alarm module. The monitoring module is used to use an armored single-mode special optical cable that is sensitive to acoustic signals as a sensor, and deploy it in the middle and bottom of the rolling cage, fence or buried on both sides of the line to monitor external vibration signals in real time. The acquisition module is used to install a camera on a pole near the guardrail, with an installation height of 3.5 meters, and deploy it at intervals of every 200 meters along the fence to collect real-time video and images of the scene. The first generation module is used to identify the vibration optical fiber signal, analyze the probability of it being a person, train, motor vehicle, wind, rain, vegetation shaking and other working conditions, and generate a vibration optical fiber alarm result. The second generation module is used to perform deep learning related algorithm processing on the video and image, identify various objects in the scene, and generate a video alarm result. And the alarm module is used to combine the vibration optical fiber alarm result, the video alarm result and the weather conditions, set different weight coefficients, conduct a comprehensive analysis, and finally make a joint decision alarm.

[0067] In this embodiment, the front-end monitoring system based on vibration optical fiber and video fusion also includes: a first sending module, a second sending module, a third sending module and a fourth sending module. The first sending module is used for the front-end monitoring device to send real-time monitoring data to the monitoring center. The second sending module is used for the front-end monitoring device to send alarm data to the monitoring center when an alarm occurs, until the alarm reception automatic confirmation information of the monitoring center is received. The third sending module is used for the front-end monitoring device to send alarm video to the monitoring center. The fourth sending module is used for the front-end monitoring device to periodically send device status data to the monitoring center.

[0068] In this embodiment, the front-end monitoring system based on vibration optical fiber and video fusion also includes: a synchronization module, which is used for the front-end monitoring equipment to use the monitoring center to which it belongs as the time source, obtain the time synchronization signal based on the NTP protocol, and perform time synchronization with the NTP server of the monitoring center once a day, and immediately make a time synchronization request when detecting its own time abnormality.

[0069] In this embodiment, the front-end monitoring system based on vibrating optical fiber and video fusion also includes: a positioning module, which is used for periodic self-inspection of equipment status, sending alarm information and triggering sound and light alarms when abnormalities occur. The optical cable has the functions of both sensor and communication equipment. When the optical cable is damaged, the broken fiber position is located according to the characteristics of the optical fiber signal at the broken fiber location, and the broken fiber positioning time is less than 3 seconds.

[0070] In this embodiment, the front-end monitoring system based on vibration optical fiber and video fusion also includes: a push module, which is used to utilize the video resources of the existing comprehensive video surveillance system, access to the intelligent analysis host through the B / H interface, combine the vibration optical fiber alarm information and the intelligent analysis algorithm, determine whether there is an intrusion target in the scene, and push the alarm result to the perimeter intrusion alarm monitoring platform.

[0071] The various variations and specific examples of the front-end monitoring method based on vibration optical fiber and video fusion provided in Example 1 are also applicable to the front-end monitoring system based on vibration optical fiber and video fusion provided in this example. Through the above detailed description of a front-end monitoring method based on vibration optical fiber and video fusion, those skilled in the art can clearly know the implementation method of a front-end monitoring system based on vibration optical fiber and video fusion in this example. Therefore, for the sake of brevity of the specification, it will not be described in detail here.

[0072] In practical applications, the overall architecture of the present invention is specifically composed of vibration optical fiber + video front-end monitoring equipment, monitoring units and other equipment. The front-end monitoring / alarm equipment includes sensor optical fibers and cameras. The sensor optical fibers are passive, low-cost, and long-life, and can be deployed over long distances and over a large area; the cameras support all-weather operation under natural light and at night. The monitoring unit includes an optical fiber demodulator, an optical fiber processing host, an edge analysis device, a power conversion module, and network equipment. Based on the information obtained by the front-end monitoring equipment, the optical fiber processing host, the edge analysis device, etc. perform target identification and positioning, and generate alarm information, real-time video, associated pictures and videos, which are transmitted to the monitoring center through a data interface.

[0073] The vibration fiber + video front-end monitoring equipment transmits the location information monitored by the vibration fiber to the camera according to the characteristics of different technologies. The camera takes pictures according to the target information, and then monitors the intrusion behavior in real time. The vibration fiber and video identify the target and output the target feature information according to their respective technologies. The vibration fiber monitors the position, amplitude, sound wave and other signals of the target, and links the camera to perform target identification and model matching. The camera performs secondary compounding based on the target information, and outputs the alarm result through the fusion algorithm based on confidence evaluation, so as to monitor the intrusion behavior in real time.

[0074] The vibrating optical fiber collects on-site signals through the monitoring host, and the camera collects on-site image information and transmits it to the edge computing device. When a person climbs over, the fence optical cable is triggered, and the climbing and climbing behaviors can be identified according to the sequence of actions. When the intruder uses underground crossing, the outer and inner optical cables of the fence will be triggered in sequence, and the underground digging and crossing intrusion behaviors can be realized. The pulse laser converts the spatial domain feature changes of spatial propagation into the time domain analysis of the light return light. The spatial domain information obtained by reverse calculation of the time domain information can effectively locate the events such as personnel intrusion, climbing, underground crossing, and equipment digging and crossing along the optical fiber. When the front-end detection device finds the intrusion target, the on-site monitoring device reports the alarm information, equipment status information, etc. to the monitoring center through the corresponding interface. When the on-site monitoring device enters the alarm state, it can trigger the sound and light alarm device in the area, and the front-end alarm device sends an alarm signal to warn and drive away the intruder. Set the logical defense zone range according to the on-site needs. Support the monitoring center to deploy and disarm the defense zone. The hardware, software and data of the front-end monitoring device are protected, and have IoT security measures. The equipment status is self-checked periodically, and an alarm message is sent and an audible and visual alarm is triggered when an abnormality occurs. The optical cable has the functions of both a sensor and a communication device. When the optical cable is damaged and a cable break occurs, the fiber break position can be located based on the characteristics of the optical fiber signal at the broken fiber location, and the fiber break location time is less than 3 seconds.

[0075] The main functions of the front-end monitoring equipment based on vibration fiber + video technology include modules such as vibration signal detection, video analysis, and data output. The sensor fiber performs real-time monitoring around the clock, and triggers the camera to take pictures after detecting the target; the sensor fiber data is fused with the video analysis algorithm to finally determine the intrusion event. The data includes monitoring and alarm information such as the location, amplitude, sound wave, video, and image of the intruding target. According to the on-site data collection, the sensor fiber data and video image data are input and resolved, the initial interference and noise data are filtered out, and the location information of the intruding target is passed to the video analysis. The target in the video is tracked and identified through relevant algorithms, the target is classified, and the alarm information is verified by image. Finally, the alarm results, equipment status information, etc. are sent to the monitoring center.

[0076] The data transmission content of the vibration optical fiber + video front-end monitoring device mainly includes monitoring data, alarm data, alarm video, equipment data, etc. Among them, the data type of monitoring data may include camera monitoring data and vibration optical fiber monitoring data. Among them, the data type of alarm data may include alarm information and alarm confirmation information. Among them, the data type of alarm video may include alarm video. Among them, the data type of equipment data may include camera status information and vibration optical fiber status information.

[0077] The front-end monitoring device sends real-time monitoring data to the monitoring center; when an alarm occurs, the front-end monitoring device sends alarm data to the monitoring center until it receives the automatic confirmation information of the alarm reception from the monitoring center; when the alarm state persists, the front-end monitoring device periodically sends alarm information, and the interval time is adjustable. After the alarm is released, the front-end monitoring device sends the alarm video (with the intrusion target mark box) from 15 seconds before the alarm starts to 15 seconds after the alarm ends to the monitoring center. The front-end monitoring device sends the device status data to the monitoring center, including the device communication status data and the device status data. Under normal circumstances, the front-end monitoring device sends 1 device status data to the monitoring center every minute. When the front-end monitoring device detects a change in its own state, it should immediately send the device status data to the monitoring center. The front-end monitoring device uses the monitoring center to which it belongs as the time source and obtains the time synchronization signal based on the NTP protocol. The monitoring device should perform time synchronization with the NTP server of the monitoring center once a day. When the front-end monitoring device detects an abnormality in its own time, it should immediately make a time synchronization request to the NTP server of the monitoring center, and the time synchronization uses the NTP protocol.

[0078] Positioning accuracy: less than 5m. Under normal working conditions, the monitoring equipment can effectively monitor human intrusion within a range of 40km on both sides. When the wind speed is greater than 20m / s (8-level gale or above) or the hourly rainfall is greater than 50mm (heavy rain or above), the recognition effect of the front-end monitoring equipment will be affected, which may lead to false alarms.

[0079] After the video reinforcement of the road base section, fixed-focus gun cameras, zoom gun cameras, ball cameras, etc. were deployed according to the on-site conditions. Through summary, it was concluded that on lines with a speed of 250 kilometers per hour and below, the camera monitoring range is 1.5 kilometers, which does not meet the monitoring plan requirements and requires additional cameras; on lines with a speed of 300 kilometers per hour, cameras are set every 400 meters. If the field of view and clarity of the existing cameras meet the requirements and basically meet the 200-meter monitoring needs, they can be reused, and an additional camera needs to be added to the uncovered area; on lines with a speed of 350 kilometers per hour, cameras are set every 200 meters. If the field of view and clarity of the existing cameras meet the requirements, they can be reused.

[0080] The perimeter intrusion alarm front-end monitoring equipment utilizes the video resources of integrated video surveillance, based on the Q / CR 575-2022 standard, through the B / H interface, combined with the architecture of the existing integrated video system, to obtain video resources from the video access node, access to the intelligent analysis host, combined with the vibration fiber optic alarm information and intelligent analysis algorithm, to determine whether there is an intrusion target in the scene, and push the alarm result to the perimeter intrusion alarm monitoring platform through the gateway.

[0081] Fiber Bragg grating array sensing technology is a combination of distributed fiber Rayleigh effect and weak grating array. The distributed acoustic wave sensing system combines coherent optical time domain reflection receiving technology with phase demodulation technology to complete the linear recovery of dynamic vibration signals, realize the long-distance and high temporal and spatial precision perception of vibration and sound field information, and improve the perception performance of vibration signals such as climbing, destruction, and drilling. In the echo detection method, the optical path coupling is realized by combining wavelength division multiplexing with time division multiple access, that is, continuous light and pulse light are combined on the same spatial signal light, and synchronous demodulation of multiple scatterings is realized on a sensing optical fiber, avoiding noise interference between multi-parameter signals during the detection process, and realizing multi-parameter synchronous detection at the same time, while simplifying the complexity of engineering applications, increasing the detection distance and significantly improving the spatial resolution. Considering the error caused by the construction of engineering optical cables, the engineering implementation accuracy is less than 5 meters. The equipment integrates key components such as power supply, clock, FPGA processor, ADC, coherent receiver, vibration detection module, laser, etc., as well as proprietary software and hardware such as tiny signal processing and high-speed sampling, FPGA signal conversion and processing algorithm, large-capacity parallel event analysis, and high-speed data transmission, to meet telecom-grade reliability.

[0082] Distributed fiber acoustic sensing (DAS) technology utilizes the highly sensitive property of coherent Rayleigh scattering stimulated by narrow-linewidth single-frequency laser in optical fiber to strain changes, and combines it with the principle of reflectometer to achieve long-distance and high-temporal and spatial precision perception of environmental vibration and sound field information.

[0083] Due to the high coherence of the DAS light source, there is multi-beam interference in the backscattered light. The changes in the polarization state and phase difference of the light wave will cause random fluctuations in the intensity of the interference light, resulting in signal fading when the light intensity is extremely weak. The project team conducts diversity reception detection based on the changes in signal fading with factors such as polarization state, frequency, phase, and mode, including polarization diversity, frequency diversity, phase diversity, mode diversity, and space diversity, and makes comprehensive judgments or vector synthesis on multi-carrier signals, thereby suppressing the impact of signal fading.

[0084] Strong wind interference is one of the main factors that cause false alarms in vibration fiber optic detection. The project team mainly solves the problem of strong wind and sand disturbance from three aspects:

[0085] A dedicated sensor fixing substrate is used, and the optical fiber sensors are arranged in parallel on the fixed substrate, so that wind and sand disturbance can be eliminated. Field experiments in a desert in Xinjiang have proved that this construction process is reliable and feasible. For areas with frequent strong winds, a local Z-shaped fence can also be used to greatly reduce the stress on the windward surface, forming a triangular layout to overcome the impact of wind resistance on the vibrating optical fiber. The sensor background demodulation system learns the environmental background noise in real time, and can perform adaptive filtering according to the changes in environmental noise to eliminate the impact of wind and sand.

[0086] The intrusion feature classification algorithm of vibrating optical fiber mainly consists of the following steps:

[0087] First, a distributed fiber optic sonar system is used to extract signals;

[0088] FPGA high-speed processor is used to calculate parameters such as main wave energy (MVE), center frequency (CF), signal spectrum (SFS), main detection peak value (MVMA), etc.

[0089] These parameters are classified and clustered using chaotic neural network algorithm;

[0090] The target features are extracted and identified using pattern recognition methods.

[0091] Signal Main Wave Energy (MVE): The integral of the square of the fiber optic detection signal amplitude. For discretely sampled fiber optic detection signals, the energy is the sum of the squares of the signal amplitudes at each point, as shown in the following formula:

[0092]

[0093] The main wave energy of the fiber optic detection signal reflects the strength information of the target signal, and the calculation result in the algorithm is used to measure the distance of the target. In the distributed fiber optic detection echo, the main wave energy (MVE) of the fiber optic detection sensor closer to the target is large, and gradually decays with the increase of distance. Generally, a larger attenuation rate indicates a heavier target weight.

[0094] Signal Main Lobe Time Delay (MVBW): The Main Lobe Time Delay (MVBW) of the fiber detection echo signal describes the speed of the intruder. The signal main lobe time delay reflected by different speed processes will change.

[0095] Signal Main Wave Q Value (MVQV): The Main Wave Q Value (MVQV) of the fiber optic detection echo signal describes the sharpness of the main energy of the incoming wave signal. The Main Wave Q Value (MVQV) of a flexible intrusion target with small energy is sharper, that is, the energy is concentrated, relatively narrow, sharp, and the Main Wave Q Value (MVQV) is large. On the contrary, the Main Wave Q Value (MVQV) of a bulky target with large energy is flatter, that is, the energy distribution is wide, the sharpness is small, and the Main Wave Q Value (MVQV) is small.

[0096] For excavators or large load-carrying vehicles, the main wave Q value (MVQV) is significantly greater than that of small animals or wind-blown sand and rockfall signals. Human intrusion signals are regular, and the main wave Q value (MVQV) is relatively fixed.

[0097] Signal Main Wave Peak-to-Peak Value (MDV): The peak-to-peak value of the echo signal in a single cycle of the fiber optic detection signal. For discretely sampled fiber optic detection signals, the main lobe peak-to-peak value measures the instantaneous strength information of the signal.

[0098] Signal center frequency (CF): The center frequency (CF) of the fiber optic detection echo signal describes the speed of the intrusion target. The center frequency (CF) of wind and sand intrusion is small and the range is large, while the center frequency (CF) of small animal intrusion is small.

[0099] Signal spectrum (SFS): The frequency distribution (SFS) of the fiber optic detection echo signal describes the action habits of the intrusion signal and is the main component of the voiceprint signal.

[0100] Signal Secondary Return Loss (2VEL): The secondary return loss (2VEL) of the optical fiber detection echo signal refers to the secondary callback peak value after the main wave oscillation of the impulse signal. This parameter describes the attenuation speed of the optical fiber detection wave signal.

[0101] Signal 3dB delay (3dBBW): The 3dB delay (3dBBW) of the fiber detection signal refers to the time it takes for the signal energy to decay to half of the original power. It can also be expressed by decomposing the square wave signal using the total power diagram of the entire signal.

[0102] Signal wave number (SSVC): The wave number (SSVC) of the fiber optic detection echo signal describes the rate of change of the target action and is the main component of the voiceprint signal.

[0103] Signal half-wave length (HTDT): The fiber optic detection echo signal is a periodic stationary signal. The signal half-wave length (HTDT) is the time from the rise of a peak to the end within the entire cycle. Some signals will have the next peak before the end. The half-wave length (HTDT) is based on the signal delay within the entire cycle.

[0104] For long and large roadbed monitoring sections, the main targets of perimeter intrusion are people. The vibration optical fiber and camera fusion alarm technology fully utilizes the respective advantages of distributed optical fiber and image intelligent recognition, reduces the system's false alarm rate and missed alarm rate, and improves the system's reliability.

[0105] Intrusion event identification - vibration optical fiber: The vibration optical fiber senses external vibration signals in real time, identifies the vibration characteristics through the recognition model, and analyzes the probability of it being a person, train, motor vehicle, wind, rain, vegetation shaking and other working conditions. When the vibration signal is suspected to be a human intrusion, the vibration alarm information is uploaded.

[0106] Alarm verification - video: The camera collects on-site pictures and videos in real time, and uses deep learning-related algorithms to identify various objects in the scene. When a person intrusion is identified, the video alarm information is uploaded.

[0107] Fusion decision: The fusion decision module combines the fiber optic alarm results, video alarm results and weather conditions, sets different weight coefficients, conducts comprehensive analysis, and finally makes a joint decision alarm.

[0108] Armored single-mode special optical cables that are extremely sensitive to acoustic signals are used as sensors. The optical cables are deployed on the rolling cages and the middle and bottom of the fences on both sides of the line (or buried underground). For key sections, the fiber optic layout is encrypted. The camera is installed on a pole near the guardrail. The installation height is 3.5 meters and it is deployed every 200 meters along the fence. The demodulator and processing host of the distributed optical fiber are deployed in the adjacent base station. A set of demodulators and processing hosts are deployed every 40 kilometers to achieve co-location access, fusion analysis and alarm of vibration signals and video information. A set of demodulators and processing hosts are deployed every 40 kilometers. On average, 800W of electricity is required per kilometer, of which 750W is for the camera (estimated at 200-meter intervals, 75W for a single camera).

[0109] In summary, the front-end monitoring method and system based on vibrating optical fiber and video fusion of the present invention have the following beneficial effects:

[0110] 1. By integrating the information of vibrating optical fiber and video images, the present invention can accurately identify and locate intrusion behaviors and improve monitoring accuracy;

[0111] 2. By adopting advanced signal processing and fusion algorithms, the present invention can effectively filter environmental interference signals, reduce false alarm rate and missed alarm rate, and improve the reliability of the monitoring system;

[0112] 3. The present invention is suitable for perimeter monitoring of railway roadbed sections under different terrain and climatic conditions, and has strong adaptability and flexibility;

[0113] 4. By utilizing the video resources of the existing integrated video surveillance system, the present invention can realize the reuse of old equipment, reduce construction costs, and improve the overall benefits of the monitoring system;

[0114] 5. The present invention can realize real-time monitoring and alarm, provide timely and accurate information support for railway safety management departments, and help to quickly respond to and handle intrusion incidents.

[0115] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A front-end monitoring method based on vibrating optical fiber and video fusion, characterized in that: include: Use armored single-mode special optical cables that are sensitive to acoustic signals as sensors and deploy them in the middle and bottom of the rolling cages, fences or buried underground on both sides of the line to monitor external vibration signals in real time; Cameras are installed on poles near the guardrail at a height of 3.5 meters and deployed every 200 meters along the fence to collect real-time video and images of the scene; Identify the vibration optical fiber signal, analyze the probability of it being a person, a train, a motor vehicle, wind, rain, vegetation shaking, etc., and generate a vibration optical fiber alarm result; Process the video and image using deep learning-related algorithms to identify various objects in the scene and generate video alarm results; Different weight coefficients are set up in combination with the vibration optical fiber alarm result, the video alarm result and weather conditions, and a comprehensive analysis is performed to finally make a joint decision alarm.

2. The front-end monitoring method based on vibrating optical fiber and video fusion as claimed in claim 1, characterized in that: Also includes: The front-end monitoring equipment sends real-time monitoring data to the monitoring center; When an alarm occurs, the front-end monitoring device sends alarm data to the monitoring center until receiving automatic confirmation information of alarm reception from the monitoring center; After the alarm is lifted, the front-end monitoring device sends the alarm video to the monitoring center; The front-end monitoring device periodically sends device status data to the monitoring center.

3. The front-end monitoring method based on vibrating optical fiber and video fusion as claimed in claim 2, characterized in that: Also includes: The front-end monitoring device uses the monitoring center to which it belongs as the time source and obtains the time synchronization signal based on the NTP protocol. It should perform time synchronization with the NTP server of the monitoring center once a day and immediately make a time synchronization request when it detects an abnormality in its own time.

4. The front-end monitoring method based on vibrating optical fiber and video fusion as claimed in claim 2, characterized in that: Also includes: The equipment status is self-checked periodically, and an alarm message is sent and an audible and visual alarm is triggered when an abnormality occurs. The optical cable has the functions of both a sensor and a communication device. When the optical cable is damaged, the broken fiber position is located based on the characteristics of the optical fiber signal at the broken fiber location, and the broken fiber location positioning time is less than 3 seconds.

5. The front-end monitoring method based on vibrating optical fiber and video fusion as claimed in claim 2, characterized in that: Also includes: Utilize the video resources of the existing integrated video surveillance system, access it to the intelligent analysis host through the B / H interface, combine the vibration fiber optic alarm information and the intelligent analysis algorithm to determine whether there is an intrusion target in the scene, and push the alarm result to the perimeter intrusion alarm monitoring platform.

6. A front-end monitoring system based on vibrating optical fiber and video fusion, characterized in that: include: The monitoring module is used to use armored single-mode special optical cables that are sensitive to acoustic signals as sensors, and deploy them in the middle and bottom of the rolling cages, fences or buried in the ground on both sides of the line to monitor external vibration signals in real time; The acquisition module is used to install cameras on poles near the guardrail at a height of 3.5 meters and deploy them every 200 meters along the fence to collect real-time video and images of the scene; The first generating module is used to identify the vibration optical fiber signal, analyze the probability of it being a person, a train, a motor vehicle, wind, rain, vegetation shaking and other working conditions, and generate a vibration optical fiber alarm result; The second generation module is used to perform deep learning related algorithm processing on the video and image, identify various objects in the scene, and generate a video alarm result; as well as The alarm module is used to combine the vibration optical fiber alarm result, the video alarm result and weather conditions, set up different weight coefficients, conduct comprehensive analysis, and finally make a joint decision alarm.

7. The front-end monitoring system based on vibration optical fiber and video fusion as claimed in claim 6, characterized in that: Also includes: The first sending module is used for the front-end monitoring device to send real-time monitoring data to the monitoring center; A second sending module, used for, when an alarm occurs, the front-end monitoring device to send alarm data to the monitoring center until receiving the alarm reception automatic confirmation information from the monitoring center; A third sending module, used for the front-end monitoring device to send the alarm video to the monitoring center; The fourth sending module is used for the front-end monitoring device to periodically send device status data to the monitoring center.

8. The front-end monitoring system based on vibration optical fiber and video fusion as claimed in claim 7, characterized in that: Also includes: The synchronization module is used for the front-end monitoring device to use the monitoring center to which it belongs as the time source and obtain the time synchronization signal based on the NTP protocol. It should perform time synchronization with the NTP server of the monitoring center once a day and immediately make a time synchronization request when it detects an abnormality in its own time.

9. The front-end monitoring system based on vibration optical fiber and video fusion as claimed in claim 7, characterized in that: Also includes: The positioning module is used for periodic self-inspection of the equipment status. When an abnormality occurs, an alarm message is sent and an audible and visual alarm is triggered. The optical cable has the functions of both a sensor and a communication device. When the optical cable is damaged, the broken fiber position is located according to the characteristics of the optical fiber signal at the broken fiber position. The broken fiber positioning time is less than 3 seconds.

10. The front-end monitoring system based on vibration optical fiber and video fusion according to claim 7, characterized in that: Also includes: The push module is used to utilize the video resources of the existing comprehensive video surveillance system, access the intelligent analysis host through the B / H interface, combine the vibration fiber optic alarm information and the intelligent analysis algorithm to determine whether there is an intrusion target in the scene, and push the alarm result to the perimeter intrusion alarm monitoring platform.

Images