Perimeter intrusion detection method, device, equipment, medium and program product

Abstract

Provided are a perimeter intrusion detection method, apparatus and device, a medium and a program product, used in the technical field of perimeter intrusion detection, the method comprising: acquiring sensing data of at least one sensor in a sensing area, the sensing data indicating a state change of the sensing area; the state change comprises at least one of the conditions that the shielding device in the sensing area is in contact with an object due to external force, the sensing area is shielded, and the shielding device vibrates or deforms; if the sensing data triggers the preset condition, acquiring at least one frame of image of the sensing area; performing target identification on the at least one frame of image to determine whether a target object and / or a target behavior of the target object exists in the at least one frame of image; and if the target object and / or the target behavior of the target object exists in the at least one frame of image, triggering a preset operation.

Description

Technical Field

[0001] This application relates to the field of intrusion detection technology, and in particular to a perimeter intrusion detection method, apparatus, equipment, medium, and program product. Background Technology

[0002] Perimeter intrusion detection and alarm technology has been widely applied in perimeter security management in industrial parks, airports, railway lines, petrochemical bases, substations, and other locations. Its core relies on hardware devices such as microwave detection, wire mesh vibration sensing, infrared beam sensors, fiber optic sensors, and laser scanning to detect intrusion. When an object moves within the monitoring range or triggers a sensing threshold, an alarm signal is directly triggered, sending an intrusion alert to management personnel. Management personnel then review the alarm through video surveillance to confirm whether it is an intrusion alarm or a false alarm. If the video review confirms intrusion, the alarm is considered an intrusion alarm, and subsequent intrusion handling procedures are initiated. If the video review confirms no intrusion, it is determined to be a false alarm. The cause of the alarm is analyzed, and management personnel either process the alarm information or initiate equipment inspection and maintenance procedures. However, current intrusion detection technology still suffers from a high false alarm rate. Summary of the Invention

[0003] In view of the above problems, this application provides perimeter intrusion detection methods, apparatus, devices, media, and program products.

[0004] According to a first aspect of this application, a perimeter intrusion detection method is provided, comprising: acquiring sensing data from at least one sensor in a sensing area, the sensing data indicating a state change in the sensing area; the state change including at least one of the following: a shielding device in the sensing area being subjected to contact with an object by an external force, the sensing area being blocked, the shielding device vibrating, or the shielding device deforming; if the sensing data triggers a preset condition, acquiring at least one frame of an image of the sensing area; performing target recognition on the at least one frame of the image to determine whether a target object and / or the target behavior of the target object exists therein; if the at least one frame of the image contains a target object and / or the target behavior of the target object, triggering a preset operation.

[0005] According to an embodiment of this application, the sensing data triggering preset conditions includes: mapping the sensing data of at least one sensor to target score ranges according to a preset intrusion risk correlation degree; calculating a comprehensive sensing value score based on the sensing data in the target score range and the weight value corresponding to each sensor; and determining that the sensing signal triggers the preset conditions if the sensing value score exceeds a preset threshold.

[0006] According to an embodiment of this application, the method for setting a preset threshold includes: acquiring weather forecast data for at least one sensing area within a target time period; inputting the weather forecast data into a pre-trained threshold adjustment model and outputting a preset threshold for the target time period; the threshold adjustment model is obtained through training based on the fusion of weather forecast data and threshold samples.

[0007] According to an embodiment of this application, the method further includes: obtaining the absolute value of the difference between the current time sensing data and the previous time interval sensing data of the same sensor; obtaining the dynamic trend index of the current time data based on the ratio of the absolute value of the difference to the standard deviation of the normal difference, wherein the standard deviation of the normal difference is determined based on the sensor data when there are no personnel or target behaviors in the same target time period; if the dynamic trend index is greater than the dynamic trend judgment threshold, determining whether a preset condition has been met based on the current time data.

[0008] According to an embodiment of this application, the preset threshold is set by multiplying a basic threshold, a regional correction coefficient, and a meteorological correction coefficient to obtain the preset threshold. The basic threshold is a preset value applicable to multiple sensing areas, the regional correction coefficient is determined based on the geographical location of the current sensing area, and the meteorological correction coefficient is determined based on weather forecast data of at least one sensing area during the target time period.

[0009] According to an embodiment of this application, the method further includes: obtaining image recognition confirmation results after the sensing data of at least two adjacent sensors reach a preset condition within a preset time interval; if the confirmation results all indicate that the sensing data reaching the preset condition is an invalid alarm, then after the sensing data of the current sensor reaches the preset condition, at least one frame of image of the sensing area is obtained and target recognition is performed without triggering the preset operation.

[0010] According to an embodiment of this application, the method for setting the weight value includes: acquiring weather forecast data for at least one sensing area during a target time period; and determining the weight value of the corresponding type of sensor based on the mapping relationship between the forecast items in the weather forecast data and the threshold values ​​of the corresponding type of sensor.

[0011] A second aspect of this application provides a perimeter intrusion detection device, comprising: a sensing data acquisition module for acquiring sensing data from at least one sensor in a sensing area, the sensing data indicating state changes in the sensing area; the state changes including at least one of the following: a shielding device in the sensing area being subjected to contact with an object by an external force, the sensing area being blocked, the shielding device vibrating, or the shielding device deforming; an image acquisition module for acquiring at least one frame of an image of the sensing area if the sensing data triggers a preset condition; a target recognition module for performing target recognition on the at least one frame of the image to determine whether a target object and / or the target behavior of the target object exists therein; and an operation triggering module for triggering a preset operation if the at least one frame of the image contains a target object and / or the target behavior of the target object.

[0012] A third aspect of this application provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors perform the methods described above.

[0013] A fourth aspect of this application also provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to perform the methods described above.

[0014] The fifth aspect of this application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method. Attached Figure Description

[0015] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0016] Figure 1 The illustrations depict application scenarios of perimeter intrusion detection methods, apparatuses, devices, media, and program products according to embodiments of this application.

[0017] Figure 2 A flowchart illustrating a perimeter intrusion detection method according to an embodiment of this application is shown schematically.

[0018] Figure 3 This schematic diagram illustrates the principle of triggering preset conditions based on sensing data according to an embodiment of this application.

[0019] Figure 4 A flowchart illustrating another perimeter intrusion detection method according to an embodiment of this application is shown schematically;

[0020] Figure 5 A schematic diagram illustrating the structure of a perimeter intrusion detection device according to an embodiment of this application is shown.

[0021] Figure 6 A schematic diagram illustrating a structural block diagram of another perimeter intrusion detection device according to an embodiment of this application; and

[0022] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a perimeter intrusion detection method according to an embodiment of this application. Detailed Implementation

[0023] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0024] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0025] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0026] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0027] It should be noted that in the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.

[0028] In the technical solution of this application, the collection, storage, use, processing, transmission, provision, application and application of user personal information all comply with the provisions of relevant laws and regulations, take necessary security measures, and do not violate public order and good morals.

[0029] In the technical solution of this application, the acquisition, collection, storage, use, processing, transmission, provision, application and application of data all comply with the provisions of relevant laws and regulations, take necessary security measures, and do not violate public order and good morals.

[0030] Perimeter refers to the outer boundary of the overall protection area defined in a security or monitoring system. It is the physical or virtual boundary that distinguishes the area inside and outside the protection area, and its purpose is to prevent external intrusion and define the security area.

[0031] Related perimeter detection relies on hardware-sensored signals (such as vibration, beam obstruction, and changes in microwave reflection) to determine whether to trigger an alarm. It cannot distinguish target types or determine whether an alarm is an intrusion or a false alarm. False alarms are triggered when animals (such as stray cats and birds), natural objects (such as branches swaying in the wind or fallen leaves), weather conditions (rain or snow), or non-intrusive human activities activate the sensors. Due to the variability and uncertainty of environmental factors, false alarms are numerous and the false alarm rate is high. This high false alarm rate means that genuine intrusion alarms may be buried under a sea of ​​false alarms, and genuine intrusion alarms may not be immediately acknowledged, identified, and responded to, reducing security management efficiency and posing security risks.

[0032] Detection methods based solely on hardware sensing signals lack target identification and judgment, making it impossible to filter alarm signals according to actual intrusion risk. For example, infrared beam systems are prone to false alarms due to environmental interference in heavy rain or dense fog, and cannot distinguish the risk differences between scenarios such as "person climbing over the fence" and "tree branch touching the fence." Regardless of whether it's a real alarm or a false alarm, management personnel need to review each alarm via video to confirm whether it's a real intrusion alarm or a false alarm, resulting in a low level of intelligence.

[0033] Frequent false alarms consume the communication bandwidth and storage resources of security systems. Administrators need to review a large number of alarms, requiring significant manpower for verification, resulting in wasted human resources and high labor costs. Improving detection efficiency and reducing false alarms are key concerns in the field of perimeter intrusion detection.

[0034] This application provides a perimeter intrusion detection method. It performs initial screening using sensor data, acquires image data under preset conditions, and uses algorithms for target recognition, forming a collaborative architecture. Before alarm triggering, it adds a step to determine personnel targets and behavioral characteristics, achieving accurate filtering of false alarms from perimeter devices. In some embodiments, relevant parameters are adjusted for different scenario characteristics to further reduce the false alarm rate.

[0035] Figure 1The illustration shows an application scenario of the perimeter intrusion detection method, apparatus, device, medium, and program product according to embodiments of this application.

[0036] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a perimeter 110, a first monitoring device 111, a second monitoring device 112, a monitored object 131, a monitored object 132, a network 140, and a server 120. The network 140 serves as a medium for providing a communication link between the first monitoring device 111, the second monitoring device 112, and the server 120. The network 140 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0037] The first monitoring device 111 and the second monitoring device 112 can interact with the server 120 via the network 140 to transmit collected video, images, and other monitoring data. The first monitoring device 111 and the second monitoring device 112 can have functions such as signal acquisition and motion capture, for example, various sensors, dome cameras, etc.

[0038] The first monitoring device 111 and the second monitoring device 112 can be various electronic devices that support video acquisition and network communication, including but not limited to bullet cameras, PTZ cameras, dome cameras, etc.

[0039] Server 120 can be a server providing perimeter intrusion detection services, such as a backend management server that provides storage, identification, and analysis support for monitoring data of monitored objects (such as 131 and 132) collected by the first monitoring device 111 and the second monitoring device 112 (this is just an example). The backend management server can analyze and process the received monitoring data (such as personnel activity images, dynamic trajectory information, etc.) and feed back the processing results (such as abnormal behavior alarms, personnel flow statistics, etc.) to relevant terminals or store them.

[0040] It should be noted that the perimeter intrusion detection method provided in this application embodiment can generally be executed by server 120. Correspondingly, the perimeter intrusion detection device provided in this application embodiment can generally be installed in server 120. The perimeter intrusion detection method provided in this application embodiment can also be executed by a server or server cluster that is different from server 120 and capable of communicating with the first monitoring device 111, the second monitoring device 112, and / or server 120. Correspondingly, the perimeter intrusion detection device provided in this application embodiment can also be installed in a server or server cluster that is different from server 120 and capable of communicating with the first monitoring device 111, the second monitoring device 112, and / or server 120.

[0041] It should be understood that Figure 1The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0042] The following will be based on Figure 1 The described scene, through Figures 2-5 The perimeter intrusion detection method of the application embodiments is described in detail.

[0043] Figure 2 A flowchart illustrating a perimeter intrusion detection method according to an embodiment of this application is shown schematically.

[0044] like Figure 2 As shown, the perimeter intrusion detection in this embodiment includes operations S210 to S240.

[0045] In operation S210, sensing data from at least one sensor in the sensing area is acquired, and the sensing data indicates a change in the state of the sensing area; the change in state includes at least one of the following: the shielding device in the sensing area is subjected to contact with an object by an external force, the sensing area is blocked, the shielding device vibrates, or the shielding device is deformed.

[0046] A sensing zone is a specific monitoring point deployed on the perimeter to detect whether the perimeter has been breached or touched. The sensing zone can be a pre-defined physical space to be monitored, for example, a sensing zone can be defined every 50 meters starting from a point on the perimeter. Typically, devices, i.e., sensors, can be installed within the sensing zone to detect the status of the area. Sensors can be installed at key locations on the perimeter fence (such as fence posts and top crossbars) to detect whether the fence has been touched or obstructed by external forces.

[0047] Sensors can include pressure sensors, infrared sensors, vibration sensors, deformation sensors, proximity sensors, etc. The corresponding sensing data can include pressure values, infrared signal blocking, vibration frequency, deformation data, etc.

[0048] For example, a vibration sensor, an infrared beam sensor, and a deformation sensor are installed in a certain sensing area. The sensing data obtained from these sensors at a certain moment include: vibration frequency of 35 Hz, infrared blocking for 1.5 seconds, and deformation amplitude of 1 cm.

[0049] During operation S220, if the sensing data triggers a preset condition, at least one frame of image of the sensing area is acquired.

[0050] The preset conditions can be pre-defined thresholds or rules to trigger image acquisition. For example, preset thresholds may include: vibration sensor value ≥ 50 Hz, infrared occlusion duration ≥ 2 seconds, and deformation amplitude ≥ 2 cm. When any of the sensed data exceeds the threshold, the next operation is triggered. Alternatively, the next operation can be triggered when the combined score of multiple sensed data exceeds a certain value.

[0051] In the example above, assume the preset conditions are set as follows: if the vibration frequency is ≥30Hz, or the infrared blockage is ≥1 second, or the deformation amplitude is ≥0.5cm, then image acquisition is triggered. Since the sensing data of a vibration frequency of 35Hz, an infrared blockage of 1.5 seconds, and a deformation amplitude of 1cm all meet the preset conditions, the trigger command acquires continuous images or short videos of the sensing area captured by the camera on top of the wall.

[0052] If, during operation S210, the infrared sensor only outputs sensing data for 0.5 seconds of continuous occlusion, which is below the threshold in the preset conditions, and other sensors do not output sensing data, then image acquisition in operation S220 will not be triggered to avoid false triggering.

[0053] In operation S230, target recognition is performed on at least one frame of image to determine whether a target object and / or the target object's target behavior exist therein.

[0054] Target recognition refers to the analysis of image content using image algorithms to identify specific people or animals. For example, people recognition can distinguish people from other objects in an image, or it can distinguish specific people (such as staff and non-staff). Behavior recognition can identify whether there is a pre-defined behavior, such as people climbing fences or staying in a sensing area.

[0055] The embodiments of this application can receive image data captured by a camera via Ethernet, and extract information such as target contours, limb features, movement trajectories, and behavioral characteristics from the image through algorithms to determine whether the target is a person and whether there is any intrusion behavior.

[0056] In operation S240, if at least one frame of the image contains a target object and / or the target behavior of the target object, a preset operation is triggered.

[0057] Preset actions can be responses to confirmed intrusions, such as triggering alarms, sending alarm information to security personnel, activating sound and light deterrence, and saving video recordings.

[0058] In the example above, assuming that the presence of a person in the image is determined through target identification and that the target behavior of "climbing the fence" is observed, the sound and light alarm near the fence is immediately activated and the alarm information is pushed to the terminal device of the security personnel.

[0059] In the embodiments of this application, the sensing data includes various state changes such as the fence being contacted by external forces, area occlusion, vibration, and deformation, which can comprehensively capture potential anomalies in the perimeter and avoid missed intrusion detection due to a single detection dimension. Secondly, image acquisition and recognition are only triggered when the sensor data meets preset conditions, and preset operations are only executed after the image confirms the presence of personnel and target behavior. By first using sensors for initial screening and then using image target recognition for verification, invalid alarms caused by simple sensor false triggering are effectively filtered out, and the reliability of perimeter intrusion detection is greatly improved.

[0060] In some embodiments of this application, the sensing data triggering preset conditions includes: mapping the sensing data of at least one sensor to target score ranges according to a preset intrusion risk correlation degree; calculating a comprehensive sensing value score based on the sensing data of the target score range and the weight value corresponding to each sensor; and determining that the sensing signal triggers the preset conditions if the sensing value score exceeds a preset threshold.

[0061] Determining whether an intrusion has occurred based on a single sensor may lead to false positives due to the limitations of that sensor. Integrating detection data from different types of sensors can reduce false positives. By setting a comprehensive sensor value score and preset thresholds, the thresholds for multiple sensors can be adjusted.

[0062] Figure 3 The schematic diagram illustrates the principle of triggering preset conditions based on sensing data according to an embodiment of this application.

[0063] like Figure 3 As shown, vibration sensors, infrared beam sensors, and fiber optic strain sensors are arranged in the sensing area of ​​a certain perimeter.

[0064] At a certain moment, the vibration sensor detected a vibration acceleration of 4.5 m / s² due to climbing; the infrared beam sensor detected a continuous occlusion time of 1700 ms due to obstruction; and the fiber optic strain sensor detected a wall strain value of 350 microstrain due to climbing.

[0065] Assuming the vibration sensor data ranges from 0 to 5 m / s², a higher value indicates more severe vibration and a higher correlation with intrusion risk. The formula for mapping vibration sensor data to the target score range is: Vibration Score = min((Original Acceleration Value / 5) × 100, 100). The infrared beam sensor data ranges from 0 to 2000 milliseconds, a higher value indicates a longer blocking time and a higher correlation with intrusion risk. The formula for mapping infrared beam sensor data to the target score range is: Infrared Score = min((Original Blocking Time / 2000) × 100, 100). The fiber optic strain sensor data ranges from 0 to 500 microstrain, a higher value indicates more severe vibration and a higher correlation with intrusion risk. The formula for mapping fiber optic strain sensor data to the target score range is: Vibration Score = min((Original Strain Value / 500) × 100, 100).

[0066] The collected data is normalized to a score of 0-100: Vibration score = (4.5 / 5) × 100 = 90 points; Infrared score = (1700 / 2000) × 100 = 85 points; Strain score = (350 / 500) × 100 = 70 points. Assume the pre-set weights for the vibration sensor (0.4), infrared beam sensor (0.35), and fiber optic strain sensor (0.25). The preset threshold is 75 points. The weighted comprehensive sensing value score is calculated as 90 × 0.4 + 85 × 0.35 + 70 × 0.25 = 83.25 points. It is then determined whether the preset threshold has been reached. If not, no action is taken; if so, the score exceeds the threshold, and the sensing signal triggers the preset conditions.

[0067] In the embodiments of this application, different types of sensors have different reliability in responding to intrusion events. By mapping the sensor's sensing data to the target range and using corresponding weights to obtain a comprehensive score, the true state of the perimeter can be reflected more reasonably, avoiding misjudgment caused by abnormal data from individual sensors, improving the accuracy of intrusion determination, and reducing misjudgment.

[0068] In some embodiments of this application, the preset threshold is set by: acquiring weather forecast data for at least one sensing area during a target time period; inputting the weather forecast data into a pre-trained threshold adjustment model and outputting a preset threshold for the target time period; the threshold adjustment model is obtained by fusing and training the weather forecast data and threshold samples.

[0069] Under favorable weather conditions, sensors can accurately detect minor intrusions. However, in adverse weather conditions (such as rain or snow), environmental interference can significantly increase the sensor's baseline values. In such cases, low threshold values ​​can trigger numerous false alarms, leading to frequent acquisition of invalid images and repeated verification by security personnel. This not only wastes resources but also reduces the importance placed on genuine alarms and may even cause alarm devices to shut down. Conversely, setting a high, fixed threshold can cause minor but genuine intrusions to be missed on sunny days due to their values ​​falling below the threshold, resulting in security vulnerabilities. The embodiments of this application utilize dynamic threshold adjustment to adapt to different weather conditions.

[0070] For example, the weather forecast data obtained for a certain period includes: blizzard; snowfall of 30 cm; wind speed: 10 m / s; temperature: -5 degrees Celsius. The model references historical blizzard samples to output a threshold of 92 points. Snow accumulation during a blizzard causes a rise in the sensor's baseline values, but this only triggers an alarm when there is a genuine intrusion, thus avoiding invalid alarms caused by snow accumulation.

[0071] In the embodiments of this application, a preset threshold is dynamically set based on weather forecast data and a model, thereby achieving environmental adaptation of the threshold: weather factors directly affect the sensing state of the sensor. Traditional fixed thresholds cannot adapt to environmental changes and are prone to weather interference leading to false alarms or excessively high thresholds leading to missed alarms. However, after the threshold adjustment model is trained by fusing weather forecast data and threshold samples, it can output a matching threshold based on the weather characteristics of the target time period, making the threshold conform to the actual environmental conditions, reducing the detection error caused by environmental interference, and improving the stability of detection under different weather conditions.

[0072] In some embodiments of this application, the method further includes: obtaining the absolute value of the difference between the current time sensing data of the same sensor and the sensing data of the previous time interval; obtaining the dynamic trend index of the current time data based on the ratio of the absolute value of the difference to the standard deviation of the normal difference, wherein the standard deviation of the normal difference is determined based on the sensor data when there are no personnel or target behaviors in the same target time period; if the dynamic trend index is greater than the dynamic trend judgment threshold, determining whether a preset condition has been met based on the current time data.

[0073] Based on weather forecast data, a preset threshold is raised during periods of unfavorable weather conditions, potentially leading to actual fence intrusion incidents. Analysis reveals that sensor data changes caused by normal weather variations are relatively gradual and continuous. For example, the data change of a vibration sensor during a rainy day with no actual intrusion incidents might be [11, 12, 10, 13, 18, 22, 25, 30, 35, 37, 36, 36]. In contrast, sensor data changes caused by actual intrusion incidents are often relatively sudden and brief. For example, the data change of a vibration sensor during a rainy day with intrusion incidents might be [11, 12, 10, 13, 14, 35, 14, 14, 15, 16, 14, 12]. Although the maximum values ​​are similar, the patterns of change differ.

[0074] Figure 4 A flowchart illustrating another perimeter intrusion detection method according to an embodiment of this application is shown schematically.

[0075] like Figure 4 As shown, vibration, infrared beam, and fiber optic strain sensors are deployed in a certain sensing area. Sensor data, after normalization, ranges from 0 to 100. On a rainy day with no intrusion, data is collected every minute, and the absolute difference between the current and previous data is calculated. The standard deviation is calculated from 60 sets of data to obtain the normal deviation standard deviation for this time period. For example, the normal deviation standard deviation for the vibration sensor is 2, for the infrared beam sensor it is 3, and for the fiber optic strain sensor it is 1.5. Assume the dynamic trend judgment threshold is 4. Only when the dynamic trend index is greater than 4 is the comprehensive score further judged to see if it reaches the preset threshold of 80 points; if the dynamic trend index is less than or equal to 4, it is directly excluded, and no comprehensive score judgment is needed.

[0076] Suppose that the acquired sensing data at time T includes: vibration sensor data (normalized to 85 points), infrared beam sensor data (normalized to 82 points), and fiber optic strain sensor data (normalized to 80 points). The corresponding sensing data at time T-1 are 63, 59, and 72 points respectively. The absolute values ​​of the current differences are calculated as follows: vibration sensor: |85-63|=22; infrared beam sensor: |82-59|=23; fiber optic strain sensor: |80-72|=8. The dynamic trend index is calculated as follows: vibration sensor: 22 / 2=11; infrared beam sensor: 23 / 3≈7.67; fiber optic strain sensor: 8 / 1.5≈5.33. The dynamic trend index is then checked to determine if it exceeds the dynamic trend judgment threshold. If so, whether the sensing data at time T meets the preset conditions. In this example, all three exceed the sensor dynamic trend judgment threshold.

[0077] The weighted summation of the current data is used to calculate the comprehensive score = (85×0.4) + (82×0.35) + (80×0.25) = 82.7 points, which is higher than the preset threshold of 80 points, thus triggering the preset conditions.

[0078] In the embodiments of this application, based on the different changing patterns of sensing data under weather changes and intrusion events, interference is filtered to reduce the false alarm rate; through dual verification of dynamic trend determination and threshold comparison, intrusion omissions caused by raising the threshold under severe weather conditions are avoided; complementing the dynamic threshold adjustment scheme, the anti-interference capability and reliability of the detection method are further improved.

[0079] In some embodiments of this application, the preset threshold is set by multiplying a basic threshold, a regional correction coefficient, and a meteorological correction coefficient to obtain the preset threshold. The basic threshold is a preset value applicable to multiple sensing areas, the regional correction coefficient is determined based on the geographical location of the current sensing area, and the meteorological correction coefficient is determined based on the weather forecast data of at least one sensing area during the target time period.

[0080] Different perimeter fencing areas have varying geographical locations and environmental interferences (e.g., adjacent to freight corridors, open areas, or windy areas). A uniform basic threshold will result in a high false alarm rate in areas with more interference and a high false negative rate in areas with less interference. The regional correction coefficient can be customized to adjust the threshold for specific characteristics of small areas, adapting to the detection needs of different regions.

[0081] For example, the baseline value for all fenced areas in a certain industrial park is 75 points. Point A at the corner of the western fence of the industrial park is located in a windy area, making it easier for sensors to detect vibrations in windy weather. Based on the geographical characteristics of point A, its coefficient relative to the ordinary fenced area is 1.0, and the area correction coefficient for this area is 1.1.

[0082] In a sunny day scenario, the weather forecast data includes: sunny, wind speed 2 m / s, no rainfall; the meteorological correction factor is set to 1.0. The preset threshold is calculated as 75 (basic threshold) × 1.1 (regional correction) × 1.0 (meteorological correction) = 82.5 points.

[0083] In a rainstorm scenario, the weather forecast data includes: wind speed 8 m / s, rainfall 55 mm; the meteorological correction factor is set to 1.2. The preset threshold calculation is: 75 × 1.1 × 1.2 = 99 points.

[0084] In the embodiments of this application, a composite threshold obtained through a basic threshold, a regional correction coefficient, and a meteorological correction coefficient is used to achieve refined scenario adaptation of the threshold. Introducing a regional correction coefficient to adapt to geographical location characteristics and a meteorological correction coefficient to adapt to short-term weather changes allows the final threshold to simultaneously match both long-term regional characteristics and short-term meteorological changes, further refining the scenario adaptability of the threshold and reducing the risk of false alarms and missed alarms caused by regional environmental and meteorological interference.

[0085] In some embodiments of this application, the method further includes: obtaining image recognition confirmation results after the sensing data of at least two adjacent sensors reach a preset condition within a preset time interval; if the confirmation results all indicate that the sensing data reaching the preset condition is an invalid alarm, then after the sensing data of the current sensor reaches the preset condition, at least one frame of image of the sensing area is obtained and target recognition is performed without triggering the preset operation.

[0086] Multiple sensors are prone to triggering invalid alarms due to local environmental interference. To avoid resource waste and interference caused by false triggers, an invalid alarm is triggered when the historical verification results of adjacent sensors all meet a preset condition. The triggering strategy for locally isolated sensors can be adjusted. By only recognizing images and not performing preset operations, invalid operations caused by environmental interference are filtered out while retaining image traceability capabilities.

[0087] For example, a perimeter fence has three adjacent vibration sensors, A, B, and C, spaced 5 meters apart. During a heavy rainstorm, the fence was affected by rain and gusts of wind, causing some areas to trigger preset vibration sensor conditions. Within the past hour, sensors A and C successively triggered their preset conditions. After triggering, the sensors acquired and identified images of the areas where A and C were located, automatically triggering alarms and notifying security personnel to check the video footage. It was confirmed that all triggers were false alarms caused by the violent impact of rainwater on the fence.

[0088] Currently, sensor B's overall sensor data score is 99 points (meeting the preset conditions) due to impact from localized gusts of wind and rain. Since the image recognition confirmation results of adjacent sensors A and C triggered within the past hour are all invalid alarms, after sensor B triggers the preset conditions, it will only acquire 3 frames of images of its sensing area and perform target identification, without triggering an alarm or notifying security personnel.

[0089] In the embodiments of this application, when adjacent sensors repeatedly issue invalid alarms recently, it indicates that there is common and similar environmental interference in the area. At this time, the current sensor is likely triggered by the same type of interference. By shielding invalid alarms caused by similar interference, while retaining the image recognition process to ensure that no real intrusion is missed, false alarms caused by common interference can be reduced in a targeted manner.

[0090] In some embodiments of this application, the weight values ​​are set by: acquiring weather forecast data for at least one sensing area during a target time period; and determining the weight values ​​of the corresponding type of sensor based on the mapping relationship between forecast items in the weather forecast data and threshold values ​​of the corresponding type of sensor.

[0091] Weather changes affect different types of sensors to varying degrees. For example, infrared beam sensors are easily blocked by rain during heavy rain, and vibration sensors are easily triggered by rain impacts, while fiber optic strain sensors are relatively less affected by heavy rain. Fixed weights can lead to data from highly interfered sensors dominating the overall score, causing misjudgments. Adjusting the weights by associating weather forecasts with sensor types can reduce the weight of highly interfered sensors and increase the weight of low-interference sensors, making the overall score more accurate and further reducing the misjudgment rate.

[0092] For example, a perimeter fence is equipped with infrared beam sensors, vibration sensors, and fiber optic strain sensors. The mapping relationship between weather forecast items and sensor thresholds includes: on sunny days, the weight of infrared beam sensors is 0.35, vibration sensors 0.4, and fiber optic strain sensors 0.25; on rainy days, the weight of infrared beam sensors, which are easily blocked by rain, is reduced to 0.2; the weight of vibration sensors, which are easily impacted by rain, is reduced to 0.3; and the weight of fiber optic strain sensors, which are less affected by rain, is increased to 0.5.

[0093] At a certain time, according to weather forecast data: heavy rain, rainfall of 55 mm, wind speed of 8 m / s; the weights are adjusted as follows: infrared beam sensor 0.2, vibration sensor 0.3, fiber optic strain sensor 0.5.

[0094] In the embodiments of this application, sensor weights are dynamically set based on the mapping relationship between weather forecast data and thresholds, so that the weighted score is more in line with the sensor reliability in the actual environment: different weather forecast items have different degrees of interference to different types of sensors. By adjusting the weights through the mapping relationship between forecast items and sensor thresholds, the weights of sensors that are more affected by weather interference can be reduced, and the weights of sensors that are less affected by interference can be increased. This makes the comprehensive sensing value score more accurately reflect the real perimeter state, avoids misjudgment caused by sensor data distortion due to weather, and improves the accuracy of detection under complex weather conditions.

[0095] Based on the above-described perimeter intrusion detection method, this application also provides a perimeter intrusion detection device. The following will be combined with... Figure 5 The device is described in detail.

[0096] Figure 5 A schematic block diagram of a perimeter intrusion detection device according to an embodiment of this application is shown.

[0097] like Figure 5 As shown, the perimeter intrusion detection device 500 of this embodiment includes a sensor data acquisition module 510, an image acquisition module 520, a target recognition module 530, and an operation triggering module 540.

[0098] The sensing data acquisition module 510 is used to acquire sensing data from at least one sensor in the sensing area. The sensing data indicates changes in the state of the sensing area. The changes in state include at least one of the following: the shielding device in the sensing area is subjected to contact with an object by an external force, the sensing area is blocked, the shielding device vibrates, or the shielding device deforms.

[0099] The image acquisition module 520 is used to acquire at least one frame of image of the sensing area if the sensing data triggers a preset condition.

[0100] The target recognition module 530 is used to perform target recognition on at least one frame of image to determine whether there is a target object and / or the target behavior of the target object.

[0101] The operation triggering module 540 is used to trigger a preset operation if at least one frame of the image contains a target object and / or the target behavior of the target object.

[0102] According to an embodiment of this application, the sensing data acquisition module 510 is further configured to map the sensing data of at least one sensor to a target score range according to a pre-set intrusion risk correlation degree; calculate a comprehensive sensing value score based on the sensing data of the target score range and the weight value corresponding to each sensor; and determine that the sensing signal triggers a preset condition if the sensing value score exceeds a preset threshold.

[0103] According to an embodiment of this application, the sensing data acquisition module 510 is further configured to acquire weather forecast data for at least one sensing area during a target time period; input the weather forecast data into a pre-trained threshold adjustment model and output a preset threshold for the target time period; the threshold adjustment model is obtained by fusing and training the weather forecast data and threshold samples.

[0104] According to an embodiment of this application, the sensing data acquisition module 510 is further configured to acquire the absolute value of the difference between the sensing data of the same sensor at the current time and the sensing data at the previous time interval; obtain the dynamic trend index of the current time data based on the ratio of the absolute value of the difference to the standard deviation of the normal difference, wherein the standard deviation of the normal difference is determined based on the sensor data when there are no personnel or target behaviors in the same target time period; if the dynamic trend index is greater than the dynamic trend judgment threshold, determine whether the preset conditions are met based on the current time data.

[0105] According to an embodiment of this application, the sensing data acquisition module 510 is further configured to obtain a preset threshold based on the product of a basic threshold, a regional correction coefficient, and a meteorological correction coefficient. The basic threshold is a preset value applicable to multiple sensing areas, the regional correction coefficient is determined based on the geographical location of the current sensing area, and the meteorological correction coefficient is determined based on the weather forecast data of at least one sensing area during the target time period.

[0106] According to an embodiment of this application, the image acquisition module 520 is further configured to acquire the image recognition confirmation result after the sensing data of at least two adjacent sensors reach the preset condition within a preset time interval; if the confirmation result indicates that the sensing data reaching the preset condition is an invalid alarm, then after the sensing data of the current sensor reaches the preset condition, at least one frame of image of the sensing area is acquired and target recognition is performed without triggering the preset operation.

[0107] According to an embodiment of this application, the sensing data acquisition module 510 is further configured to acquire weather forecast data for at least one sensing area during a target time period; and determine the weight value of the corresponding type of sensor based on the threshold mapping relationship between the forecast items in the weather forecast data and the corresponding type of sensor.

[0108] According to embodiments of this application, any multiple modules among the sensing data acquisition module 510, image acquisition module 520, target recognition module 530, and operation triggering module 540 can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least some of the functions of one or more of these modules can be combined with at least some of the functions of other modules and implemented in one module. According to embodiments of this application, at least one of the sensing data acquisition module 510, image acquisition module 520, target recognition module 530, and operation triggering module 540 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the sensing data acquisition module 510, image acquisition module 520, target recognition module 530, and operation triggering module 540 may be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0109] Figure 6 A schematic block diagram of another perimeter intrusion detection device according to an embodiment of this application is shown.

[0110] like Figure 6 As shown, the perimeter intrusion detection device includes a perimeter fence sensing module, a signal triggering module, a camera acquisition module, an algorithm recognition module, and an alarm decision module.

[0111] The perimeter fence sensing module is installed at key locations on the perimeter fence (such as fence posts and top crossbars) to detect whether the fence is touched or obstructed by external forces. The camera acquisition module is installed in the monitoring area corresponding to the perimeter fence sensing module and can use a high-definition network camera. The camera acquisition module is linked to the signal triggering module. When the sensing data detected by the perimeter fence sensing module triggers preset conditions, the signal triggering module immediately sends an "image acquisition command" to the camera acquisition module, which immediately starts and captures continuous images or short videos. The algorithm recognition module receives the image data captured by the camera via Ethernet and extracts information such as target contours, limb features, movement trajectories, and behavioral characteristics from the images using algorithms to determine whether the target is a "person" and whether there is intrusion behavior. When the algorithm identifies a person and whether the person is engaging in intrusion behavior, the alarm decision module immediately triggers an audible and visual alarm and remotely pushes alarm information to the management terminal. If the target is not identified as a person, no alarm is triggered in certain scenarios; if the target is identified as a person but the person is not engaging in intrusion behavior, no alarm is triggered; the anomaly is recorded in the log.

[0112] The embodiments of this application, by reducing invalid alarms, eliminate the need for administrators to handle unnecessary alarms, while preventing genuine intrusion alarms from being masked by false alarms, thus improving perimeter security. Abnormal information that does not trigger alarms is only stored in the logs, reducing the amount of alarm information transmitted and stored, lowering system hardware resource consumption, and extending equipment lifespan.

[0113] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a perimeter intrusion detection method according to an embodiment of this application.

[0114] like Figure 7 As shown, an electronic device 700 according to an embodiment of this application includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage portion 708 into a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0115] RAM 703 stores various programs and data required for the operation of electronic device 700. Processor 701, ROM 702, and RAM 703 are interconnected via bus 704. Processor 701 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 702 and / or RAM 703. It should be noted that programs may also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in one or more memories.

[0116] According to embodiments of this application, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to a bus 704. The electronic device 700 may also include one or more of the following components connected to the I / O interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 707 including a network interface card such as a LAN card, modem, etc. The communication section 707 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0117] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0118] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 702 and / or RAM 703 and / or one or more memories other than ROM 702 and RAM 703 described above.

[0119] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code enables the computer system to implement the perimeter intrusion detection method provided in the embodiments of this application.

[0120] When the computer program is executed by the processor 701, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0121] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 707, and / or installed from a removable medium 711. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0122] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 707, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0123] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, p9thon, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0124] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0125] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this application can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments and / or claims of this application can be combined or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

[0126] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this application is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this application, and all such substitutions and modifications should fall within the scope of this application.

Claims

1. A perimeter intrusion detection method, characterized in that, include: Acquire sensing data from at least one sensor in the sensing area, the sensing data indicating state changes in the sensing area; the state changes include at least one of the following: the shielding device in the sensing area is subjected to external force and comes into contact with an object, the sensing area is blocked, the shielding device vibrates, or the shielding device deforms. If the sensing data triggers a preset condition, at least one frame of image of the sensing area is acquired; Target recognition is performed on the at least one frame of image to determine whether a target object and / or the target object's target behavior exist therein; If the target object and / or the target behavior of the target object are present in at least one frame of the image, a preset operation is triggered.

2. The method according to claim 1, characterized in that, The sensing data triggers preset conditions, including: The sensing data from the at least one sensor are mapped to target score ranges according to a pre-set intrusion risk correlation degree. Based on the sensing data within the target score range and the weight value corresponding to each sensor, a comprehensive sensing value score is calculated. If the sensor value score exceeds a preset threshold, the sensor signal is determined to trigger a preset condition.

3. The method according to claim 2, characterized in that, The preset threshold can be set in the following ways: Acquire weather forecast data for at least one sensing area within the target time period; The weather forecast data is input into a pre-trained threshold adjustment model, which outputs a preset threshold for the target time period. The threshold adjustment model is obtained by fusing weather forecast data and threshold samples.

4. The method according to claim 3, characterized in that, Also includes: Obtain the absolute value of the difference between the current sensing data and the sensing data at the previous time interval of the same sensor; The dynamic trend index of the current time data is obtained by the ratio of the absolute value of the difference to the standard deviation of the normal difference, wherein the standard deviation of the normal difference is determined based on sensor data when there are no personnel or target behaviors in the same target time period. If the dynamic trend index is greater than the dynamic trend determination threshold, it is determined whether the preset conditions have been met based on the current time data.

5. The method according to claim 2, characterized in that, The preset threshold can be set in the following ways: The preset threshold is obtained by multiplying a basic threshold, a regional correction coefficient, and a meteorological correction coefficient. The basic threshold is a preset value applicable to multiple sensing areas. The regional correction coefficient is determined based on the current geographical location of the sensing area. The meteorological correction coefficient is determined based on weather forecast data of at least one sensing area during the target time period.

6. The method according to claim 1, characterized in that, Also includes: Obtain the image recognition confirmation result after the sensing data of at least two adjacent sensors reach the preset conditions within the preset time interval; If all the confirmation results indicate that the sensing data has reached the preset condition and is an invalid alarm, then after the current sensor's sensing data reaches the preset condition, at least one frame of image of the sensing area is acquired and target recognition is performed without triggering the preset operation.

7. The method according to claim 2, characterized in that, The methods for setting the weight values ​​include: Acquire weather forecast data for at least one sensing area within the target time period; The weight value of the corresponding sensor type is determined based on the mapping relationship between the forecast items in the weather forecast data and the threshold values ​​of the corresponding sensor types.

8. A perimeter intrusion detection device, characterized in that, The device includes: A sensing data acquisition module is used to acquire sensing data from at least one sensor in the sensing area, wherein the sensing data indicates a state change in the sensing area; the state change includes at least one of the following: the shielding device in the sensing area is subjected to external force and comes into contact with an object, the sensing area is blocked, the shielding device vibrates, or the shielding device deforms. The image acquisition module is used to acquire at least one frame of image of the sensing area if the sensing data triggers a preset condition; The target recognition module is configured to perform target recognition on the at least one frame of image to determine whether a target object exists therein and / or the target behavior of the target object; and An operation triggering module is used to trigger a preset operation if the target object and / or the target object's target behavior are present in at least one frame of the image.

9. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 7.

11. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 7.

Images