Airport perimeter intrusion alarm system and method

By combining radar and video technology, embedded AI is used for target recognition, the problem of high false alarm rate and low real-time performance of the airport intrusion alarm system is solved, and all-weather accurate monitoring and high-accurate intrusion alarms are achieved.

CN115841727BActive Publication Date: 2024-11-12CHINA ELECTRONIC TECH GRP CORP NO 38 RES INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211509922.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-11-29
Publication Date
2024-11-12
Estimated Expiration
2042-11-29

AI Technical Summary

Technical Problem

The existing airport invasion alarm system has problems such as high false alarm rate, susceptibility to external interference, and poor real-time performance.

Method used

Adopt an intrusion alarm system combining radar and video, combined with embedded AI technology, quickly detect targets through phased array radar and capture pictures in conjunction with high-definition bolts, and use intelligent analysis box to identify targets in depth learning algorithms to achieve accurate all-weather and daytime Detection and has equipment blocking alarm function.

Benefits of technology

It significantly improves the accuracy and real-time nature of intrusion alarms, reduces the false alarm rate, ensures accurate alarms for targets approaching and climbing the boundary, and enhances the reliability and anti-interference ability of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115841727B_ABST
    Figure CN115841727B_ABST
Patent Text Reader

Abstract

The present invention provides an airport perimeter intrusion alarm system and method, including: a monitoring center device (1) and a field device (2); the monitoring center device (1) is connected to one or more sets of the field devices (2); the field devices (2) include a laser gun integrated machine (21), an audible and visual alarm (22), a router (23), an optical fiber transceiver (24), an intelligent analysis box (25), a waterproof equipment box and a rod; the monitoring center device (1) includes an application server (11), an optoelectronic switch (12) and a storage device (13). Through the radar linkage camera mode, a deep learning algorithm is used to detect the category and position of the intrusion target, detect the target approaching and climbing the perimeter, and have the target intrusion alarm and equipment obstruction alarm functions. The present invention solves the technical problems of high false alarm rate, susceptibility to external interference and poor real-time performance.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of airport security alarm, and in particular to an airport perimeter intrusion alarm system and method. Background Art

[0002] As an important public transportation hub, airport security is of vital importance. For airports, it is particularly important to adopt safer, more reliable, practical and effective perimeter security system technology.

[0003] At present, the front-end detection technologies used in airport perimeter intrusion alarm systems at home and abroad mainly include infrared counter-radiation, electronic fences, vibration optical fibers, and intelligent video surveillance. These traditional methods have defects such as high false alarm rate, susceptibility to weather and environmental influences, and low degree of automation, and can no longer meet the airport's needs for perimeter protection. Taking the widely used vibration optical fiber as an example, strong winds and vehicle vibrations can cause false alarms. At the same time, the vibration optical fiber can only alarm for climbing and crossing, and cannot provide early warning for targets close to the perimeter.

[0004] The Chinese invention patent application document with publication number CN113205654A discloses "An airport perimeter security monitoring alarm system and method based on AI information technology", in which the system realizes the functions of perimeter anti-cutting and anti-climbing through the combination of multiple sensors, but the application still has problems such as high false alarm rate and susceptibility to external interference. The Chinese invention patent application document with publication number CN114913654A discloses "An airport perimeter intrusion pre-alarm processing device and method based on edge computing", which adopts image intelligent early warning technology based on infrared thermal imaging and visible light video, and combines it with vibration sensors to filter out the common environmental factors of airport perimeters. This application requires the real-time application of video analysis algorithms, which has a large processing volume and low real-time performance, and is still affected by weather interference.

[0005] In summary, the existing technology has technical problems such as high false alarm rate, susceptibility to external interference and poor real-time performance. Summary of the invention

[0006] The technical problem to be solved by the present invention is how to solve the technical problems in the prior art of high false alarm rate, susceptibility to external interference and poor real-time performance.

[0007] The present invention adopts the following technical solutions to solve the above technical problems: an airport perimeter intrusion alarm system includes: no less than 2 field devices and monitoring center equipment, the field equipment includes: a lightning gun integrated machine, an audible and visual alarm, an intelligent analysis box, a router, an optical fiber transceiver and a rod, wherein:

[0008] The integrated laser gun is hoisted with the sound and light alarm on the cross arm of the pole, and the pole is installed on the inner side of the airport perimeter to detect whether there is a moving target in the preset warning area. When a moving target approaches or passes through the perimeter in the preset warning area, the integrated laser gun is used to shoot through radar linkage, capture the warning area picture, and send the warning area picture to the intelligent analysis box;

[0009] The intelligent analysis box is used to receive and detect the warning zone image by using preset intelligent processing logic, and judge whether there is a target invading the warning zone, so as to generate alarm trigger information; the intelligent analysis box is connected to the sound and light alarm through a router, and is used to send alarm trigger information to the sound and light alarm to trigger the alarm when there is a target invading the warning zone, wherein the intelligent analysis box also includes: a device blocking alarm unit, which is used to generate device blocking alarm trigger information when the device is blocked;

[0010] Monitoring center equipment, which is connected to the field equipment, including: application server, optoelectronic switch and storage device;

[0011] An application server is used to obtain and display alarm trigger information and device cover alarm trigger information, and send the alarm trigger information and device cover alarm trigger information to a storage device;

[0012] The storage device is used to receive and store alarm trigger information and equipment cover alarm trigger information, and the storage device is connected to the application server.

[0013] The present invention combines the advantages of radar and video, adopts embedded AI technology, and can realize accurate detection of targets around the clock. It can also monitor the behavior of targets approaching and climbing the fence at the same time, and realize alarm when the target approaches and climbs the fence. It has high accuracy, extremely low false alarm rate, and strong real-time performance. The system itself has anti-obstruction alarm function to ensure the actual deployment and application effect of airport fences.

[0014] In a more specific technical solution, the integrated laser gun includes: a phased array radar, no less than 2 high-definition gun cameras, and a laser sensor;

[0015] Phased array radar, used to detect moving targets within a preset warning zone;

[0016] High-definition gun camera, used to capture pictures of the preset warning area when there is a moving target in the warning area;

[0017] Laser sensor, used to detect the distance to the obstruction.

[0018] The integrated lightning gun of the present invention adopts an integrated structure, has a detection distance range of 0-200 meters, has no blind spots, and has target intrusion alarm and equipment obstruction alarm functions, which effectively improves the reliability and practicality of the system.

[0019] In a more specific technical solution, the monitoring center equipment also includes: an optoelectronic switch, the optoelectronic switch is connected to the application server and the storage device respectively;

[0020] In a more specific technical solution, the monitoring center equipment also includes: an optoelectronic switch, which is connected to the optical fiber transceiver of each set of field equipment. The optical fiber transceiver is installed in a waterproof equipment box of the field equipment to receive and send optoelectronic signals between the monitoring center equipment and the field equipment.

[0021] In a more specific technical solution, the intelligent analysis box includes: an intelligent data processing unit for detecting the warning zone image using the following logic to determine whether there is a target invading the warning zone, so as to generate alarm trigger information.

[0022] In a more specific technical solution, the intelligent data processing unit also includes:

[0023] A target recognition model parameter processing component is used to use a preset perimeter radar target recognition data set to offline train a perimeter radar target detection network to obtain target recognition model parameters;

[0024] A detection network initialization component is used to import target recognition model parameters to initialize the perimeter radar target detection network, and the detection network initialization component is connected to the target recognition model parameter processing component;

[0025] A forward reasoning component is used to input the warning area image into the perimeter radar target, and use the perimeter radar target detection network to perform forward reasoning calculations to obtain forward advancement results. The detection network initialization component is connected to the forward reasoning component;

[0026] The intrusion category and position output component is used to output the category and position information of the intrusion target in the warning area image based on the forward advancement result of the perimeter radar target detection network. The intrusion category and position output component is connected to the forward reasoning component and the detection network initialization component.

[0027] The intrusion alarm system of the present invention detects targets quickly through phased array radar, automatically links high-definition gun cameras to capture target images and sends them to the intelligent analysis box, which uses a self-designed and trained deep learning algorithm to identify the target category and automatically eliminate interference. The present invention uses a self-designed and trained deep learning algorithm for target identification, which can eliminate the influence of non-intrusive targets, with high identification accuracy and extremely low false alarm rate.

[0028] In a more specific technical solution, the on-site equipment also includes: a router, a laser gun all-in-one machine, an intelligent analysis box and an audible and visual alarm are all connected through the router.

[0029] In a more specific technical solution, the router includes: LAN port, WAN port and 485 interface.

[0030] In a more specific technical solution, the equipment occlusion alarm unit includes:

[0031] The obstruction capture component is used to use the laser sensor to detect the distance of the obstruction in real time. The high-definition gun camera regularly captures the device obstruction picture and sends it to the intelligent analysis box;

[0032] The obstruction judgment component is used to judge whether the radar or the gun camera is blocked according to the device obstruction picture. The obstruction judgment component is connected to the obstruction capture component;

[0033] The device obstruction alarm component is used to control the sound and light alarm to alarm when the radar or gun is blocked by the intelligent analysis box, and push the alarm information to the application server for storage and display. The device obstruction alarm component is connected to the obstruction judgment component.

[0034] The present invention uses radar + video + embedded AI technology to greatly improve the accuracy and real-time performance of intrusion alarms and significantly reduce the false alarm rate. By setting up a perimeter warning zone, it can realize target approach warning and climbing alarm. At the same time, the radar gun integrated machine has an occlusion alarm function, which improves the system reliability and anti-interference ability. It can detect intrusion behaviors approaching and climbing the perimeter all day and all weather without being affected by the environment.

[0035] In a more specific technical solution, the cross arm is provided with no less than two anti-climbing structures.

[0036] In a more specific technical solution, an airport perimeter intrusion alarm method includes:

[0037] S1, detecting whether there is a moving target in the preset warning area. When a moving target approaches or passes through the boundary in the preset warning area, using the integrated laser gun (21) to shoot through radar linkage, capture a picture of the warning area, and send the picture of the warning area to the intelligent analysis box (25);

[0038] S2, using preset intelligent processing logic, receiving and detecting the warning area image, judging whether there is a target invading the warning area, and generating alarm trigger information;

[0039] S3, when a target intrudes into the warning zone, sending alarm trigger information to the sound and light alarm (22) to trigger an alarm, wherein step S2 further comprises: when the device is blocked, generating device blocking alarm trigger information;

[0040] S4, obtaining and displaying alarm triggering information and device shielding alarm triggering information, and sending the alarm triggering information and device shielding alarm triggering information to a storage device (13);

[0041] S5. Receive and store alarm trigger information and device occlusion alarm trigger information.

[0042] Compared with the prior art, the present invention has the following advantages: the present invention combines the advantages of radar and video, adopts embedded AI technology, and can realize accurate detection of targets around the clock. It can simultaneously monitor the behavior of targets approaching and climbing the fence, and realize alarm of target approaching and climbing the fence. It has high accuracy, extremely low false alarm rate, strong real-time performance, and the system itself has anti-obstruction alarm function to ensure the actual deployment and application effect of the airport fence.

[0043] The integrated lightning gun of the present invention adopts an integrated structure, has a detection distance range of 0-200 meters, has no blind spots, and has target intrusion alarm and equipment obstruction alarm functions, which effectively improves the reliability and practicality of the system.

[0044] The intrusion alarm system of the present invention detects targets quickly through phased array radar, automatically links high-definition gun cameras to capture target images and sends them to the intelligent analysis box, which uses independently designed and trained intelligent algorithms to identify target categories and automatically eliminate interference. The present invention uses independently designed and trained deep learning algorithms for target identification, which can eliminate the influence of non-intrusive targets, with high identification accuracy and extremely low false alarm rate.

[0045] The present invention uses radar + video + embedded AI technology to greatly improve the accuracy and real-time performance of intrusion alarms and significantly reduce the false alarm rate; by setting up a fence warning area, it can realize target approach warning and climbing alarm. At the same time, the integrated radar gun has an occlusion alarm function, which improves the system reliability and anti-interference ability, and can detect intrusion behaviors approaching and climbing the fence all day and all weather without being disturbed by the environment. The present invention solves the technical problems existing in the prior art, such as high false alarm rate, susceptibility to external interference, and poor real-time performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a schematic diagram of the architecture of an airport perimeter intrusion alarm system according to Embodiment 1 of the present invention;

[0047] Figure 2 This is a schematic diagram of the workflow of the airport perimeter intrusion alarm method according to Embodiment 1 of the present invention;

[0048] Figure 3 This is a schematic diagram of an artificial intelligence algorithm network according to Embodiment 1 of the present invention;

[0049] Figure 4 This is a schematic diagram of specific steps of detection and identification processing according to Embodiment 1 of the present invention;

[0050] Figure 5 This is a schematic diagram of the installation of the airport iron fence equipment according to Embodiment 1 of the present invention;

[0051] Figure 6This is a schematic diagram of the installation of the airport brick wall perimeter equipment according to Example 1 of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Example 1

[0054] like Figure 1 As shown, the present invention discloses an airport perimeter intrusion alarm system, comprising:

[0055] One or more sets of field equipment 2 and one set of monitoring center equipment 1.

[0056] In this embodiment, each set of field equipment 2 includes: a lightning gun integrated machine 21, an audible and visual alarm 22, a router 23, a fiber optic transceiver 24, an intelligent analysis box 25, a waterproof equipment box 26, and a set of rods 27.

[0057] In this embodiment, the integrated laser gun machine 21 includes: a phased array radar, five high-definition gun cameras and a laser sensor, and the router 23 includes: a LAN port, a WAN port and a 485 interface.

[0058] In this embodiment, the effective coverage range of the integrated lightning gun device 21 includes: 0-200 meters, and the field devices 2 are deployed end to end according to the actual length of the airport perimeter to cover the entire airport perimeter.

[0059] In this embodiment, the phased array radar of the integrated laser gun 21 detects moving targets appearing in the warning area, automatically links the high-definition gun camera to capture the moving targets and sends the captured images to the intelligent analysis box 25 for confirmation.

[0060] In this embodiment, the laser sensor of the laser gun integrated machine 21 detects the distance of the obstruction in real time, the high-definition gun camera captures pictures at regular intervals, and the intelligent analysis box 25 determines the obstruction event. If obstruction occurs, the intelligent analysis box 25 controls the sound and light alarm to alarm and send the alarm information to the application server 11.

[0061] In this embodiment, the integrated lightning gun 21 is connected to the LAN port of the router 23 via a network cable, the sound and light alarm 22 is connected to the 485 interface of the router 23 via a 485 line, the intelligent analysis box 25 is connected to the LAN port of the router 23 via a network cable, the WAN port of the router 23 is connected to the network port of the fiber optic transceiver 24, and the fiber optic transceiver 24 sends data via optical fiber.

[0062] In this embodiment, the field equipment 2 is deployed on the airport fence, the integrated lightning gun 21 and the sound and light alarm 22 are hoisted on the cross arm of the pole; the waterproof equipment box is installed on the pole column; the router 23, the intelligent analysis box 25 and the optical fiber transceiver 24 are installed in the waterproof equipment box; the pole column is installed on the inner side of the airport fence, and is deployed at a maximum interval of 200 meters according to the actual situation of the airport fence, and is connected end to end.

[0063] In this embodiment, the monitoring center device 1 is deployed in a computer room or a monitoring center, and the monitoring center device 1 includes an application server 11 , an optoelectronic switch 12 , and a storage device 13 .

[0064] In this embodiment, the optoelectronic switch 12 receives data transmitted by each set of field equipment 2 through multiple optical fibers, the application server 11 is connected to the optoelectronic switch 12 through a network cable, and the storage device 13 is connected to the optoelectronic switch 12 through a network cable.

[0065] In this embodiment, the smart analysis box 25 uses a self-designed and trained deep learning algorithm to detect the target in the captured image sent by the laser gun integrated machine 21, and at the same time sends a command to control the sound and light alarm 22 to sound an alarm prompt, and sends the alarm information to the application server 11. The warning area can cover the airport fence area and the external area, and can simultaneously alarm the target approaching and climbing the fence; the smart analysis box 25 can determine the occlusion event and alarm.

[0066] In this embodiment, the application server 11 receives the alarm information sent by the intelligent analysis box 25 and stores it in the database, and queries, counts, and displays the alarm information stored in the database. The alarm information includes target intrusion alarms and equipment obstruction alarms. The target intrusion alarm information includes the number of intrusion targets, intrusion target types, intrusion target locations, intrusion time, and alarm pictures of intrusion targets.

[0067] In this embodiment, the application server 11 provides an interface for the third-party management platform to call so that it can obtain alarm information for analysis or display.

[0068] The system can divide the monitoring range of the airport perimeter field equipment 2 into a target warning area and an exclusion area.

[0069] In this embodiment, the warning area is within the coverage of the airport perimeter field device 2, and is used to detect whether the target appears. The exclusion area can be set within the monitoring range of the airport perimeter field device 2, and the exclusion area does not perform intrusion detection.

[0070] In this embodiment, airport security personnel can flexibly and conveniently set warning areas and exclusion areas by graphically displaying and dragging the mouse.

[0071] like Figure 2 As shown, this embodiment discloses an airport perimeter intrusion alarm method, including a target intrusion alarm method and a device shielding alarm method, including the following steps:

[0072] S1. Phased array radar detects whether there are moving targets in the warning area;

[0073] S2. If there is a moving target in the warning area, the high-definition camera of the laser gun integrated machine will capture the picture and send it to the intelligent analysis box. If not, it will return to the target intrusion alarm module;

[0074] S3, the intelligent analysis box 25 confirms the intrusion target in the captured image. In this embodiment, the intelligent analysis box detects and identifies whether there are targets such as people, cars, animals, etc. in the captured image through a built-in intelligent algorithm. The algorithm network is as follows: Figure 3 As shown;

[0075] like Figure 4 As shown, in this embodiment, step S3 also includes the following specific detection and identification processing steps:

[0076] S31, using the perimeter radar target recognition data set to train the independently designed perimeter radar target detection network to obtain target recognition model parameters;

[0077] In this embodiment, the loss function used in the process of training the perimeter radar target detection network is as follows:

[0078]

[0079]

[0080] Among them, S represents the scale of the detection module used by the perimeter radar target detection network, and B represents the number of target boxes predicted by each cell at each detection scale of the detection module; represents the balance parameter at each scale, α box represents the weight of coordinate loss, α obj represents the weight of target confidence loss, α cls Represents the weight of the category loss; IOU represents the intersection-and-union ratio between the predicted target box and the labeled true value box, ω gt Indicates the width of the labeled truth box, hgt represents the height of the labeled truth box, b w Indicates the width of the predicted target box, h w Indicates the height of the predicted target box; Indicates whether the j-th predicted target box of the ith cell at a certain scale contains the target. If it does, If the target is not included C i Represents the confidence of the i-th cell at a certain scale; They represent the confidence of the marked target, class represents the target category to be detected, and p i (c) is the predicted probability for each category, is the true probability of each category;

[0081] S32, importing the target recognition model parameters trained offline to initialize the self-designed perimeter radar target detection network;

[0082] S33, the warning area image is input into the perimeter radar target detection network for forward reasoning calculation;

[0083] S34, the perimeter radar target detection network outputs the category and location information of the intrusion target in the warning area image.

[0084] S4. If a target appears in the captured image, the intelligent analysis box 25 controls the sound and light alarm 22 in the field device 2 to sound an alarm, and at the same time pushes the alarm information to the application server for storage and display.

[0085] In this embodiment, the device blocking alarm step includes:

[0086] S1', high-definition bolt camera to take pictures regularly;

[0087] S2', the intelligent analysis box 25 comprehensively determines whether the image is blocked according to the clarity, contrast and brightness of the captured image;

[0088] S1”, the laser sensor detects the distance of the obstruction in real time, and the acquired data is sent to the intelligent analysis box 25;

[0089] S2”, intelligent analysis box 25 automatically determines whether the radar is blocked;

[0090] S3”, if the gun or radar is blocked, push the alarm information to the application server 11, if not, return to execute steps S2' and S2”;

[0091] S4”, the application server displays the alarm information and saves the alarm information to the database. In this embodiment, the intelligent analysis box 25 controls the sound and light alarm 22 in the field device 2 to alarm, and pushes the alarm information to the application server 11 for storage and display.

[0092] like Figure 5 and Figure 6 As shown, this embodiment discloses an installation example of an intrusion alarm system in an airport perimeter.

[0093] Since the types of airport fences mainly include iron mesh and brick walls, the installation methods of iron mesh and brick walls are similar. For the installation of iron mesh, please refer to Figure 5 , for brick wall installation, refer to Figure 6 .

[0094] The pole column is installed in the inner area of ​​the airport fence. In order to ensure the detection effect, both the outer area of ​​the fence and the top area of ​​the fence can be detected. The horizontal arm is about 1 meter higher than the top of the fence, and the horizontal arm extends 0.3 meters from the edge of the fence. The on-site equipment 2, the integrated lightning gun 21, is hoisted at the end of the horizontal arm. In order to prevent the equipment from falling, it is equipped with an anti-fall safety rope during installation; the sound and light alarm 22 is hoisted in the middle of the horizontal arm near the top of the fence, facing down, and barbs are installed on the horizontal arm to prevent intruders from climbing through the horizontal arm. The waterproof equipment box is clamped on the column, and the integrated lightning gun 21 is arranged end to end according to the actual situation on site, with a maximum spacing of 200 meters, so as to achieve full coverage of the prevention area.

[0095] The airport perimeter intrusion alarm system and method provided in this embodiment uses a deep learning algorithm to detect the type and location of intrusion targets, with high accuracy and extremely low false alarm rate. Through the radar-linked camera mode, the number of images that the algorithm needs to process is reduced while ensuring detection accuracy, greatly improving the real-time performance of the system. At the same time, the system can detect the behavior of targets approaching and climbing the perimeter, and has an occlusion alarm function, which greatly improves the reliability and practicality of the system.

[0096] In summary, the present invention combines the advantages of radar and video, adopts embedded AI technology, and can realize accurate detection of targets around the clock. It can also monitor the behavior of targets approaching and climbing the fence at the same time, and realize alarm of targets approaching and climbing the fence. It has high accuracy, extremely low false alarm rate, strong real-time performance, and the system itself has anti-obstruction alarm function to ensure the actual deployment and application effect of airport fences.

[0097] The integrated lightning gun of the present invention adopts an integrated structure, has a detection distance range of 0-200 meters, has no blind spots, and has target intrusion alarm and equipment obstruction alarm functions, which effectively improves the reliability and practicality of the system.

[0098] The intrusion alarm system of the present invention detects targets quickly through phased array radar, automatically links high-definition gun cameras to capture target images and sends them to the intelligent analysis box, which uses independently designed and trained intelligent algorithms to identify target categories and automatically eliminate interference. The present invention uses independently designed and trained deep learning algorithms for target identification, which can eliminate the influence of non-intrusive targets, with high identification accuracy and extremely low false alarm rate.

[0099] The present invention uses radar + video + embedded AI technology to greatly improve the accuracy and real-time performance of intrusion alarms and significantly reduce the false alarm rate; by setting up a fence warning area, it can realize target approach warning and climbing alarm. At the same time, the integrated radar gun has an occlusion alarm function, which improves the system reliability and anti-interference ability, and can detect intrusion behaviors approaching and climbing the fence all day and all weather without being disturbed by the environment. The present invention solves the technical problems existing in the prior art, such as high false alarm rate, susceptibility to external interference, and poor real-time performance.

[0100] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An airport perimeter intrusion alarm system, characterized in that: The system includes: no less than 2 on-site devices and monitoring center equipment, the on-site equipment includes: a laser gun integrated machine, an audible and visual alarm, an intelligent analysis box and a rod, wherein: The integrated laser gun and the sound and light alarm are hoisted on the cross arm of the rod, and the rod is installed on the inner side of the airport perimeter to detect whether there is a moving target in the preset warning area. When a moving target approaches or passes through the perimeter in the preset warning area, the integrated laser gun is used to shoot through radar linkage, capture a picture of the warning area, and send the picture of the warning area to the intelligent analysis box; The intelligent analysis box is used to receive and detect the warning area image using preset intelligent processing logic, thereby determining whether the target intruding the warning area appears, so as to generate alarm trigger information; the intelligent analysis box includes: an intelligent data processing unit, which is used to detect the warning area image using the following logic, thereby determining whether the target intruding the warning area appears, so as to generate alarm trigger information; Wherein, the intelligent data processing unit also includes: A target recognition model parameter processing component is used to use a preset perimeter radar target recognition data set to offline train a perimeter radar target detection network to obtain target recognition model parameters; In the process of training the perimeter radar target detection network, the loss function used is as follows: In the formula, S represents the scale of the detection module used by the perimeter radar target detection network, and B represents the number of predicted target boxes for each cell at each detection scale of the detection module; represents the balance parameter at each scale, α box represents the weight of coordinate loss, α obj represents the weight of target confidence loss, α cls Represents the weight of the category loss; IOU represents the intersection-over-union ratio between the predicted target box and the labeled true value box, ω gt represents the width of the labeled truth box, h gt represents the height of the labeled truth box, b w Indicates the width of the predicted target box, h w Represents the height of the predicted target box; Indicates whether the j-th predicted target box of the ith cell at a certain scale contains the target. If it does, If the target is not included C i Represents the confidence of the i-th cell at a certain scale; They represent the confidence of the marked target, class represents the target category to be detected, and p i (c) is the predicted probability for each category, is the true probability of each category, ν represents the measure of the difference in aspect ratio between the predicted target box and the labeled true value box; A detection network initialization component, used for importing the target recognition model parameters to initialize the perimeter radar target detection network, wherein the detection network initialization component is connected to the target recognition model parameter processing component; A forward reasoning component, used to input the warning area image into the perimeter radar target, and use the perimeter radar target detection network to perform forward reasoning calculation to obtain a forward advancement result, and the detection network initialization component is connected to the forward reasoning component; An intrusion category and position output component, used to output the category and position information of the intrusion target in the warning area picture according to the forward advancement result using the perimeter radar target detection network, and the intrusion category and position output component is connected to the forward reasoning component and the detection network initialization component; The smart analysis box is connected to the sound and light alarm, and is used to send the alarm trigger information to the sound and light alarm to trigger the alarm when the target intruding the warning zone appears, wherein the smart analysis box also includes: a device blocking alarm unit, which is used to generate device blocking alarm trigger information when the device is blocked; wherein the device blocking alarm unit includes: The obstruction capture component is used to detect the distance of the obstruction in real time using a laser sensor, and the high-definition gun camera regularly captures the device obstruction picture and sends it to the intelligent analysis box; An obstruction judgment component is used to judge whether the radar and the gun are blocked according to the device obstruction picture, and the obstruction judgment component is connected to the obstruction capture component; The monitoring center device is connected to the field device, and the monitoring center device includes: an application server and a storage device; An application server, used for acquiring and displaying the alarm triggering information and the device shielding alarm triggering information, and sending the alarm triggering information and the device shielding alarm triggering information to a storage device; The device blocking alarm component is used to control the sound and light alarm to alarm when the radar and the gun are blocked by the intelligent analysis box, and push the alarm information to the application server for storage and display. The device blocking alarm component is connected to the blocking object judgment component; The storage device is used to receive and store the alarm triggering information and the device shielding alarm triggering information, and the storage device is connected to the application server.

2. The airport perimeter intrusion alarm system according to claim 1, characterized in that: The integrated laser gun machine includes: a phased array radar, no less than 2 high-definition gun cameras and a laser sensor; The phased array radar is used to detect the moving target in a preset warning area; The high-definition gun camera is used to capture a picture of the preset warning area when there is the moving target in the warning area; The laser sensor is used to detect the distance to the obstruction.

3. The airport perimeter intrusion alarm system according to claim 1, characterized in that: The monitoring center equipment also includes: an optoelectronic switch, which is connected to the application server and the storage device respectively, and is used for converting and transmitting optoelectronic signals between the application server and the field equipment through optical fibers.

4. The airport perimeter intrusion alarm system according to claim 3, characterized in that: The field equipment also includes: a waterproof equipment box, the optoelectronic switch is connected to the optical fiber transceiver of each set of the field equipment, and the optical fiber transceiver is installed in the waterproof equipment box to receive and send the optoelectronic signal between the monitoring center equipment and the field equipment.

5. The airport perimeter intrusion alarm system according to claim 3, characterized in that: The on-site equipment also includes: a router, which includes a LAN port, a WAN port and a 485 interface; the integrated lightning gun machine and the intelligent analysis box are both connected to the LAN port of the router, the sound and light alarm is connected to the 485 port of the router, and the optical fiber transceiver is connected to the WAN port of the router.

6. The airport perimeter intrusion alarm system according to claim 1, characterized in that: The cross arm is provided with no less than two anti-climbing structures.

7. An airport perimeter intrusion alarm method, used to control an airport perimeter intrusion alarm system according to any one of claims 1 to 6, characterized in that: The method comprises: S1, detecting whether there is a moving target in the preset warning area. When a moving target approaches or passes through the boundary of the preset warning area, the integrated laser gun device (21) is used to shoot with the radar, capture a picture of the warning area, and send the picture of the warning area to the intelligent analysis box (25); S2, using preset intelligent processing logic to receive and detect the warning area image, and determine whether there is a target invading the warning area, so as to generate alarm trigger information; S3, when the target intruding the warning zone appears, sending the alarm trigger information to the sound and light alarm (22) to trigger an alarm, wherein the step S2 further comprises: when the device is blocked, generating device blocking alarm trigger information; S4, acquiring and displaying the alarm triggering information and the device shielding alarm triggering information, and sending the alarm triggering information and the device shielding alarm triggering information to a storage device (13); S5. Receive and store the alarm triggering information and the device shielding alarm triggering information.