Airport pipeline safety monitoring method based on airport underground three-dimensional pipeline model

Abstract

This disclosure presents an embodiment of an airport pipeline safety monitoring method based on an airport underground 3D pipeline model. One specific implementation of this method includes: acquiring a multi-source device sensing dataset for airport underground pipelines; synchronizing the multi-source device sensing dataset to the airport underground 3D pipeline model to obtain a current 3D pipeline model; performing signal change analysis on the multi-source device sensing data in the multi-source device sensing dataset to generate a signal analysis result set; based on the current 3D pipeline model, calling the corresponding cameras in the underground pipelines to capture pipeline scene videos at corresponding locations; performing pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment identification information; and updating the historical pipeline environment record table to generate an underground pipeline safety monitoring record table. This implementation improves the accuracy of airport underground pipeline safety monitoring.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the fields of computer technology, image recognition technology, and three-dimensional pipeline modeling technology, specifically to an airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model. Background Technology

[0002] The safety monitoring of underground pipelines at airports has a decisive impact on the safe operation of airports. Currently, safety monitoring of underground pipelines at airports is usually carried out manually. However, due to the complexity and large coverage of underground pipelines at airports, manual monitoring is prone to blind spots (for example, ignoring some hidden defects in the pipelines, or making it difficult to directly observe animal damage to underground pipelines due to their natural tendency to avoid them, thus making it difficult to directly determine the factors causing pipeline anomalies). This results in slow response times to pipeline anomalies, leading to inadequate safety monitoring of underground pipelines. Consequently, it is difficult to respond to and handle pipeline anomalies in a timely manner, resulting in serious safety hazards (such as the rapid reproduction of rodents, which greatly increases the probability of damage to cables, pipe joints, or rubber anti-corrosion coatings in the pipelines if not controlled in time). Summary of the Invention

[0003] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0004] Some embodiments of this disclosure propose an airport pipeline safety monitoring method based on an underground three-dimensional pipeline model of the airport, in order to solve the technical problems mentioned in the background section above.

[0005] In a first aspect, some embodiments of this disclosure provide an airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model. The method includes: acquiring a multi-source device sensing dataset for the airport underground pipeline; synchronizing the multi-source device sensing dataset to the airport underground three-dimensional pipeline model to obtain a current three-dimensional pipeline model; wherein the multi-source device sensing data is collected by sensors in a sensor assembly, and each sensor is bound to corresponding sensor coding information in the current airport underground three-dimensional pipeline model, the sensor coding information including corresponding sensor coordinates, sensor number, sensor type, underground passage number, and pipeline type number; performing signal change analysis on the multi-source device sensing data in the multi-source device sensing dataset to generate a signal analysis result set; and responding to determining the signal... The analysis results set contains signal analysis results representing signal changes. For each signal analysis result representing a signal change, based on the current 3D pipeline model, the corresponding camera in the underground pipeline is invoked to capture pipeline scene videos at the corresponding locations. Pipeline scene image recognition is performed on each captured pipeline scene video to generate pipeline environment identification information. The pipeline environment identification information includes pipeline environment anomaly type identifiers and pipeline anomaly identification information. The pipeline environment anomaly type identifiers represent pipeline static environment anomaly types or pipeline dynamic environment anomaly types. The pipeline anomaly identification information corresponding to the pipeline dynamic environment anomaly type includes pipeline hazard animal identifiers. Based on the generated pipeline environment identification information set, the historical pipeline environment record table is updated to generate an underground pipeline safety monitoring record table.

[0006] Secondly, some embodiments of this disclosure provide an airport pipeline safety monitoring device based on an airport underground three-dimensional pipeline model. The device includes: an acquisition and synchronization unit configured to acquire a multi-source device sensing dataset for airport underground pipelines and synchronize the multi-source device sensing dataset to the airport underground three-dimensional pipeline model to obtain a current three-dimensional pipeline model. The multi-source device sensing data is collected by sensors in a sensor assembly, and each sensor is bound to corresponding sensor coding information in the current airport underground three-dimensional pipeline model. The sensor coding information includes corresponding sensor coordinates, sensor number, sensor type, underground passage number, and pipeline type number; a signal change analysis unit configured to perform signal change analysis on the multi-source device sensing data in the multi-source device sensing dataset to generate a signal analysis result set; and a calling unit configured to respond to a confirmation... The above signal analysis results set contains signal analysis results that represent signal changes. For each signal analysis result representing a signal change, based on the current three-dimensional pipeline model, the corresponding camera in the underground pipeline is called to capture pipeline scene video at the corresponding location. The pipeline scene image recognition unit is configured to perform pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment recognition information. The pipeline environment recognition information includes a pipeline environment anomaly type identifier and pipeline anomaly recognition information. The pipeline environment anomaly type identifier represents a static or dynamic environment anomaly type. The pipeline anomaly recognition information corresponding to the dynamic environment anomaly type includes a pipeline hazard animal identifier. The update unit is configured to update the historical pipeline environment record table based on the generated pipeline environment recognition information set to generate an underground pipeline safety monitoring record table.

[0007] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0008] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0009] The above-described embodiments of this disclosure have the following beneficial effects: the airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model, as described in some embodiments of this disclosure, can improve the accuracy of airport underground pipeline safety monitoring, enabling timely and targeted handling of abnormal situations, thereby reducing airport safety hazards. Specifically, the reason for the existence of serious safety hazards is that airport underground pipelines are complex and cover a large area, while manual monitoring is prone to detection blind spots (for example, ignoring some hidden defects in the pipelines, or making it difficult to directly observe animal damage to underground pipelines due to animals' avoidance instincts, thus making it difficult to directly determine the factors causing pipeline anomalies), resulting in slow response speed to pipeline anomalies, and therefore, inadequate underground pipeline safety monitoring. Based on this, the airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model, as described in some embodiments of this disclosure, firstly acquires a multi-source equipment sensing dataset for airport underground pipelines, and then synchronizes the multi-source equipment sensing dataset to the airport underground three-dimensional pipeline model to obtain the current three-dimensional pipeline model. The multi-source sensing data is collected by sensors in the sensor assembly. Each sensor is bound to corresponding sensor coding information in the current airport underground 3D pipeline model. This sensor coding information includes the corresponding sensor coordinates, sensor number, sensor type, underground passage number, and pipeline type number. By introducing the airport underground 3D pipeline model, various sensors laid within the airport underground passages can be uniformly associated, replacing manual monitoring. Then, signal change analysis is performed on the multi-source sensing data in the aforementioned multi-source sensing dataset to generate a signal analysis result set. This signal change analysis determines whether the signal analysis results indicate a signal change, thereby identifying any anomalies in the underground pipeline. Next, in response to the determination that a signal analysis result indicating a signal change exists in the signal analysis result set, for each signal analysis result indicating a signal change, the corresponding camera in the underground pipeline is invoked to capture a video of the pipeline scene at the corresponding location, based on the current 3D pipeline model. The introduction of cameras allows for integration with sensors, enabling the capture of pipeline scene videos at abnormal locations. Finally, pipeline scene image recognition is performed on each captured pipeline scene video to generate pipeline environment identification information. The pipeline environment identification information includes pipeline environment anomaly type identifiers and pipeline anomaly identification information. The pipeline environment anomaly type identifiers represent either static or dynamic environmental anomalies, while the pipeline anomaly identification information corresponding to dynamic environmental anomalies includes identifiers of hazardous animals. Therefore, image recognition can be used to accurately identify the real factors causing pipeline anomalies. Finally, based on the generated pipeline environment identification information set, the historical pipeline environment record table is updated to generate an underground pipeline safety monitoring record table. This improves the accuracy of airport underground pipeline safety monitoring, thereby reducing airport safety hazards. Attached Figure Description

[0010] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0011] Figure 1 This is a flowchart of some embodiments of the airport pipeline safety monitoring method based on the airport underground three-dimensional pipeline model according to the present disclosure;

[0012] Figure 2 This is a schematic diagram of the pipe numbers in the initial three-dimensional piping model;

[0013] Figure 3 This is a schematic diagram of the model structure of a lightweight image recognition model;

[0014] Figure 4 This is a partial top view of the distribution of animal activity caused by pipeline hazards;

[0015] Figure 5 This is a structural schematic diagram of some embodiments of the airport pipeline safety monitoring device based on the airport underground three-dimensional pipeline model according to the present disclosure;

[0016] Figure 6 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0017] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0018] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0019] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0020] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0021] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0022] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] Figure 1 This is a flowchart illustrating some embodiments of the airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model according to this disclosure. The airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model includes the following steps:

[0024] Step 101: Obtain the multi-source equipment perception dataset for the airport's underground pipelines, and synchronize the multi-source equipment perception dataset to the airport's underground 3D pipeline model to obtain the current 3D pipeline model.

[0025] In some embodiments, the execution entity (e.g., a computing device) of the airport pipeline safety monitoring method based on an airport underground 3D pipeline model can acquire a multi-source device sensing dataset for the airport underground pipeline via a wired or wireless connection, and synchronize the multi-source device sensing dataset to the airport underground 3D pipeline model to obtain the current 3D pipeline model. The multi-source device sensing data is collected by sensors in a sensor assembly. The sensor assembly can include various types of sensors, such as infrared sensors, vibration sensors, distributed fiber optic sensors, and gravity sensors. Each sensor is bound to corresponding sensor coding information in the current airport underground 3D pipeline model. The sensor coding information includes the corresponding sensor coordinates, sensor number, sensor type, underground passage number, and pipeline type number. Here, the multi-source device sensing data can be transmitted to the corresponding sensor according to the correspondence between the sensor and the multi-source device sensing data to obtain the current 3D pipeline model.

[0026] It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (Ultra Wide Band) connections, and other currently known or future wireless connection methods.

[0027] It should be noted that the aforementioned computing devices can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed on the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0028] Optionally, the above-mentioned three-dimensional underground pipeline model of the airport is generated through the following steps:

[0029] Step S1: Based on the pipeline dimension data in the pre-set airport underground pipeline construction drawings, establish an underground three-dimensional tunnel model. This model includes a set of underground tunnel numbers, with each number representing a segment of the underground tunnel. Here, the airport underground pipeline construction drawings can be three-dimensional dimension drawings, and the pipeline dimension data includes construction data for each underground tunnel within the airport, such as tunnel length, width, and height. A tunnel segment can be defined as the distance from one connecting point (or tunnel endpoint) to another. This facilitates the creation of corresponding underground tunnel numbers for each segment, aiding in subsequent location tracking. Secondly, the underground three-dimensional tunnel model is established within the airport's three-dimensional coordinate system. This system can be established with any location within the airport area (e.g., the airport center) as its origin.

[0030] Step S2: Based on the preset underground pipeline data file, simulated pipelines are established within the underground passage area of ​​the aforementioned underground 3D passage model, and a corresponding pipeline type number is assigned to each simulated pipeline, resulting in an initial 3D pipeline model. The underground pipeline data file can include information about the pipelines laid in the underground passage. For example, pipeline information includes pipeline type information, pipeline transmission medium, pipeline laying location in the underground passage, pipeline dimensions, etc. Therefore, simulated pipelines can be added to the underground 3D passage model according to the actual airport pipeline design based on the underground pipeline data file. Furthermore, a corresponding pipeline type number can be assigned to each pipeline segment based on the underground passage number where the pipeline is located, resulting in an initial 3D pipeline model.

[0031] As an example, see Figure 2The diagram shows the pipe numbering in the initial 3D pipework model. Here, a specific underground passage T101 in the initial 3D pipework model, used for power supply and communication, can include communication fiber optic cable line D001, communication fiber optic cable line D002, monitoring cable G011, power cable conduit X001, and power cable conduit X003. Different types of pipes can be distinguished by different prefix symbols followed by the pipe number of the same type. This not only facilitates the differentiation of different types of pipes but also allows for the differentiation of pipes of the same type. Furthermore, it enables the mapping between underground passage numbers and pipe type numbers.

[0032] Step S3: Based on the preset underground sensor deployment data file, simulated sensors and corresponding simulated cameras are added to the initial 3D pipeline model, and corresponding sensor and camera coding information is established to obtain the airport underground 3D pipeline model. The underground sensor deployment data file includes camera identifiers and coordinates for multiple cameras, and sensor identifiers and coordinates for each sensor in the sensor assembly. There is a correspondence between camera identifiers and sensor identifiers; each camera identifier corresponds to at least one underground passage number or at least one pipeline type number. Here, the shapes of the simulated sensors and cameras can be schematic and used to establish the correspondence between sensors and underground passages and pipelines in the airport underground 3D pipeline model. Furthermore, the cameras actually deployed in the underground pipelines correspond one-to-one with the simulated cameras in the airport underground 3D pipeline model.

[0033] In practice, to better identify the actual factors causing pipeline abnormalities in underground passages, multiple cameras are installed within these passages. For example, cameras are placed at passageway interfaces, and for longer passages, cameras can be added in the middle. Considering that infrared thermal imaging cameras are easily affected by ambient temperature—for example, if the temperature inside underground pipelines is similar to an animal's body temperature, it can be difficult to distinguish the animal's location—the cameras can be configured as infrared night vision monitoring cameras. Thus, video recording using infrared thermal imaging can be done both day and night, facilitating real-time monitoring. Secondly, corresponding sensor deployment methods are used for different pipelines and different underground passage locations. Specifically, for steel pipelines such as domestic water pipelines, fire water pipelines, hot water pipelines, and compressed air pipelines, vibration sensors can be installed at the sealed interfaces. For plastic pipes such as communication fiber optic cable pipelines and monitoring cable pipelines, not only are distributed fiber optic sensors installed, but vibration sensors are also installed at the junctions to detect damage to the pipelines by animals (e.g., rodents). Additionally, gravity sensors and infrared sensors can be installed at passageway intersections to detect the presence of animals. Here, infrared sensors can be installed inside the channel at preset intervals to detect the presence of living organisms.

[0034] Step 102: Perform signal change analysis on the multi-source device sensing data in the multi-source device sensing dataset to generate a signal analysis result set.

[0035] In some embodiments, the aforementioned executing entity can perform signal change analysis on the multi-source device sensing data in the aforementioned multi-source device sensing dataset to generate a signal analysis result set. For the multi-source device sensing data from infrared sensors, considering the high speed of animal movement, processing to remove instantaneous interference cannot be performed to avoid filtering out the true detection signal. Therefore, the difference between the real-time temperature value and the reference temperature value in the multi-source device sensing data from the infrared sensor can be detected. If the difference exceeds a preset temperature difference, a signal analysis result indicating a signal change is generated. Here, the reference temperature value can vary with different seasons, weather conditions, and times. Similarly, for the multi-source device sensing data corresponding to vibration sensors, the difference between the amplitude of the vibration signal and the reference amplitude can be determined. If the difference exceeds a preset amplitude difference, a signal analysis result indicating a signal change is generated. Likewise, for the multi-source device sensing data corresponding to gravity sensors, a signal analysis result indicating whether a signal change has occurred can also be generated. Furthermore, for multi-source sensing data corresponding to distributed fiber optic sensors, due to their wide detection range, interference signals from aircraft takeoffs and landings at airports often interfere with the fiber optic sensors. Therefore, wavelet transform algorithms can be used for noise reduction to avoid numerous misjudgments. Then, the absolute deviation between the real-time light intensity of the processed fiber optic signal and the baseline light intensity is determined. If this absolute deviation is greater than a preset deviation threshold, a signal analysis result representing the signal change is generated.

[0036] In practice, considering the numerous influencing factors in the environment of airport underground pipelines (e.g., aircraft takeoff and landing affecting fiber optic and vibration signals, animals entering and damaging underground pipelines, foreign objects falling into underground pipelines, etc.), relying solely on fiber optic sensors for signal processing requires highly precise noise segmentation of the signal. This necessitates more complex signal processing algorithms, resulting in significant computational resource consumption for the large volume of signals in the airport environment and a potential slowdown in signal processing due to the sheer number of signals. Therefore, this application aims to avoid this situation by accelerating response times and more accurately identifying the true factors causing pipeline anomalies. Consequently, signal processing operations for each deployed sensor are minimized to maintain signal sensitivity. Furthermore, the introduction of cameras for auxiliary identification of anomaly locations improves the response efficiency and accuracy of security monitoring.

[0037] Step 103: In response to the determination that there are signal analysis results in the signal analysis result set that represent signal changes, for each signal analysis result that represents signal changes, according to the current three-dimensional pipeline model, call the corresponding camera in the underground pipeline to capture pipeline scene video at the corresponding location.

[0038] In some embodiments, the execution entity may, in response to determining that there is a signal analysis result in the set of signal analysis results indicating a signal change, call the corresponding camera in the underground pipeline to capture video of the pipeline scene at the corresponding location for each signal analysis result indicating a signal change, based on the current three-dimensional pipeline model. Specifically, for the sensor coordinates corresponding to each signal analysis result indicating a signal change, the camera corresponding to the coordinates of the nearest camera located in the same underground passage can be identified as the target camera to be called.

[0039] In some optional implementations of certain embodiments, the execution entity, based on the signal analysis results for each characteristic of a signal change and according to the current three-dimensional pipeline model, calls the corresponding camera in the underground pipeline to capture video of the pipeline scene at the corresponding location, including:

[0040] Step S1: Based on the current 3D pipeline model, determine the camera identifier corresponding to the signal analysis result, which serves as the target camera identifier. The camera corresponding to the target camera identifier shares the same underground passage number and pipeline type number as the signal analysis result, and is in standby mode. Here, sensor identifiers corresponding to the signal analysis result can be found in the current 3D pipeline model. This allows for the identification of the nearest camera identifier in the same underground passage as the sensor identifier, which is then used as the target camera identifier. The camera corresponding to this target camera identifier is the target camera. This avoids situations where the target is too small to be easily identified due to the camera being too far away.

[0041] In practice, protective covers can be installed on cameras installed in underground passages to greatly prevent rodents from damaging the cameras.

[0042] Step S2 involves performing camera detection on the cameras corresponding to the target camera identifiers to determine if any camera malfunctions. Considering potential issues such as signal abnormalities, abnormal shooting angles, and abnormal captured images, camera detection is necessary before capturing scene video. This involves checking the online status of the camera corresponding to the target camera identifier. If the camera is online, it is determined that the camera is not malfunctioning. Additionally, the presence of obstructions can be determined by examining the captured images. Finally, any obstructions can be reported to the target terminal for removal.

[0043] Step S3: In response to determining that the camera is not malfunctioning, the camera corresponding to the target camera identifier is controlled to adjust its shooting direction to capture the environment of the pipeline corresponding to the signal analysis result, thereby obtaining a pipeline scene video. The shooting direction is generated based on the sensor coordinates corresponding to the target camera identifier. Here, the shooting direction can be the direction from the camera to the location of the sensor coordinates. Therefore, the camera can be controlled to turn to the aforementioned shooting direction to capture the environment and obtain the pipeline scene video.

[0044] Step 104: Perform pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment recognition information.

[0045] In some embodiments, the aforementioned executing entity can perform pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment identification information. This pipeline environment identification information includes a pipeline environment anomaly type identifier and pipeline anomaly identification information. The pipeline environment anomaly type identifier represents either a static or dynamic pipeline environment anomaly type. Static pipeline environment anomaly types represent pipeline anomalies caused by non-biological factors (e.g., pipeline fire, channel collapse). Dynamic pipeline environment anomaly types represent pipeline anomalies caused by biological factors (e.g., rodents gnaw on underground communication pipelines). The pipeline anomaly identification information corresponding to dynamic pipeline environment anomaly types includes a pipeline hazard animal identifier, which can be the species identifier of the identified hazard animal. The pipeline environment identification information can be obtained by recognizing pipeline scene images in pipeline scene videos using a pre-trained convolutional network model. Here, the convolutional network model is a model that takes pipeline scene images as input and pipeline hazard animal identifiers as output. For example, if a convolutional network model identifies rodents as a pest species in a pipeline scene image, then the output pipeline pest identifier could be "shu". Thus, the pipeline pest identifier can be used to identify pipeline environmental information.

[0046] In some optional implementations of certain embodiments, the aforementioned execution entity performs pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment identification information, including:

[0047] Step S1: Sample images from the pipeline scene video to obtain a pipeline scene image sequence. Here, images from the pipeline scene video can be acquired at preset frame intervals to obtain the pipeline scene image sequence.

[0048] Step S2 involves performing rapid target detection on the aforementioned pipeline scene image sequence to generate a first target detection identifier. This first target detection identifier indicates whether biological features have been detected. Here, a pre-defined rapid target detection model can be used to perform rapid target detection on the aforementioned pipeline scene image sequence to generate the first target detection identifier.

[0049] As an example, a fast target detection model can include an input layer, a feature extraction layer, a feature fusion layer, and a binary classification regression head. Specifically, the input layer size can be 224 × 224 × 3. The feature extraction layer can employ a lightweight MobileNetV3-Small backbone network (removing two high-dimensional features and reducing the lightweight attention module (Squeeze-and-Excitation Module, lightweight channel attention mechanism) from 16x to 8x, reducing the compression ratio and decreasing the number of parameters by 40%). This includes pointwise convolutional layers (i.e., 1 × 1 convolutional layers for channel expansion), activation functions (Hard-Swish activation function, used to increase the dimensionality of features and perform nonlinear transformations to achieve fast extraction of pipeline edge features), deep convolutional layers, batch normalization layers, activation functions, lightweight attention modules, pointwise convolutional layers, and residual connection layers. Here, the lightweight structure of the feature extraction layer facilitates the extraction of core features, achieving a 90% reduction in computation in practice. The feature fusion layer can include a global average pooling layer and a convolutional layer. The binary classification regression head is a fully connected layer with an output dimension of 2. Therefore, recognition results can be output within tens of milliseconds or even a few milliseconds, achieving lightweight object detection. This yields the probability of the presence and absence of biometric labels, and finally, the label with the higher probability is determined as the first object detection identifier.

[0050] Step S3: In response to determining that the first target detection identifier indicates the detection of biological features, target feature point detection is performed on the pipeline scene image sequence to obtain a target feature point sequence. First, the pipeline scene images in the pipeline scene image sequence can be segmented using a target segmentation algorithm to separate the regions containing biological features as target sub-images. Then, the coordinates of the center points of each target sub-image can be determined as target feature points, thereby obtaining the target feature point sequence.

[0051] In practice, target segmentation algorithms can employ either the lightweight SAM (Segment Anything Model) image segmentation algorithm or the lightweight SSD (Single Shot MultiBox Detector) target detection algorithm.

[0052] Step S4: Determine the target distance value sequence based on the aforementioned target feature point sequence. The target distance value represents the distance between the detected target and the corresponding camera location. Here, the coordinates of the aforementioned target feature points can be transformed from the image coordinate system to the camera coordinate system of the corresponding camera through coordinate transformation to obtain three-dimensional feature points. Then, the distance between the aforementioned three-dimensional feature points and the origin is determined as the target distance value, resulting in the target distance value sequence. In practice, the target distance value can represent the distance between the animal corresponding to the biometric feature and the camera.

[0053] Step S5: In response to determining that the above target distance value sequence satisfies a first preset distance condition, biometric target recognition is performed on the above pipeline scene image sequence to obtain pipeline anomaly identification information. Here, the first preset distance condition may be the existence of N consecutive target distance values ​​less than or equal to a preset distance threshold. Here, the preset distance threshold may be the distance from the midpoint between two cameras to the camera. Next, a preset target biometric recognition algorithm can be used to perform biometric target recognition on the target sub-images corresponding to the target distance values ​​that satisfy the first preset distance condition in the above pipeline scene image sequence to obtain a pipeline hazard animal identification sequence. Finally, the pipeline hazard animal identification sequence with the most identical and numerous identifications can be used as pipeline anomaly identification information.

[0054] As an example, the target biometric algorithm can be an SVM (Support Vector Machine) or a pre-trained lightweight image recognition model. See also... Figure 3As shown, the lightweight image recognition model can be structured similarly to the feature extraction layer in the aforementioned lightweight MobileNetV3-Small backbone network. The lightweight image recognition model can include: an input layer, a preprocessing module, a feature extraction layer, a fine-grained bi-branch feature enhancement module, a feature fusion layer, a fine-grained classification head, and an output layer. Here, since the target sub-image is a segmented sub-image with a variable image size, a preprocessing module is introduced to adaptively scale the image to a scale of 64 × 64. Then, the number of channels in the convolutional layers of the feature extraction layer is reduced from 8 to 5. The depthwise convolutional layer with a stride of 2 is removed to avoid compressing small feature maps to a single pixel size. The fine-grained bi-branch feature enhancement module can include a local feature branch and a proportional feature branch. Specifically, taking rodent recognition as an example, the feature extraction layer can include: a pointwise convolutional layer, an activation function, a depthwise convolutional layer (3 × 3), a batch normalization layer, an activation function, a lightweight attention module, a residual connection layer, and a pointwise convolutional layer. Here, residual connections only take effect when the input and output channels / sizes match, used for gradient backpropagation to avoid gradient vanishing. The local feature branch can consist of pixel-level attention pooling layers, pointwise convolutional layers, rectified linear activation functions, and L2 normalization layers. Pixel-level attention pooling can include a spatial attention mask pre-created based on key points of rodents (e.g., ears, tails). Here, the spatial attention mask can be 64 × 64. The weights of keypoint regions in the spatial attention mask can be 1, while the weights of other regions are set to zero. This facilitates the extraction of fine-grained features such as ear shape and tail texture. The proportional feature branch can include adaptive pooling layers, pointwise convolutional layers, and rectified linear activation functions. This facilitates the extraction of features related to the body length and tail length of the target. The feature fusion layer can be used to concatenate the features from the local feature branch and the proportional feature branch to obtain 128-dimensional features, which are then compressed using a 1 × 1 convolution to obtain 64-dimensional comprehensive features. The fine-grained classification head can include a first fully connected layer and a second fully connected layer. Here, the output layer can output the distribution probabilities of multiple rodent species. Here, the number of rodent species can be pre-set based on the airport's rodent infestation situation. For example, if the airport's underground pipelines contain three types of rodents: brown rats, house mice, and yellow-breasted rats, then the distribution probability of these three species can be output. Finally, the rodent icon corresponding to the species with the highest distribution probability can be designated as the pipeline hazard icon.

[0055] Step S6: The pipeline environment anomaly type identifier, which characterizes the type of pipeline dynamic environment anomaly, and the aforementioned pipeline anomaly identification information are determined as pipeline environment identification information. For example, the pipeline environment anomaly type identifier, which characterizes the type of pipeline dynamic environment anomaly, is "1".

[0056] Step S7: In response to determining that the target distance value sequence does not meet the first preset distance condition but meets the second preset distance condition, the second camera is invoked to capture a second scene video, and the pipeline scene image recognition is performed again using the second scene video to generate pipeline environment recognition information. The target distance value sequence not meeting the first preset distance condition can occur if there are N consecutive target distance values ​​greater than a preset distance threshold, indicating that the distance between the target and the current camera is increasing, thus requiring the invocation of other cameras for continued tracking and recognition. Secondly, the camera closest to the target camera and corresponding to the same underground passage number or pipeline type number can be selected as the second camera. Here, the second camera is positioned in the shooting direction of the target camera. Therefore, the pipeline scene video can be captured again using the second camera, and pipeline scene image recognition can be performed again using the above step 104 to generate pipeline environment recognition information. If the second camera is already occupied, the current camera can be used to continue performing the biometric target recognition step.

[0057] In practice, in airport underground pipeline scenarios, to detect whether the factors causing pipeline anomalies are caused by animals (e.g., rodents), it is necessary to first determine whether there are biological features in the pipeline scene images. Here, if we were to identify whether there is any damage to the underground pipeline by recognizing videos of the underground pipeline captured by cameras in all time zones, it would require significant computational resources. Furthermore, animals in underground pipelines are not present all day; only a small portion of the videos captured in all time zones would show animals in the pipeline. Therefore, recognizing animals in all time zone videos would waste considerable computational resources. Thus, this application can capture video only when an anomaly is detected in the pipeline (e.g., abnormal vibration, detection of animals passing by), and then perform corresponding image recognition, thereby greatly reducing the number of video segments where animals are not captured. This reduces the consumption of computational resources. Furthermore, considering that rodents in underground pipeline scenarios move back and forth at relatively high speeds, this approach is more efficient. Therefore, directly identifying pipeline scene images is not only difficult to capture the small target features of rodents, but also challenging due to their fast movement or occlusion, leading to unsatisfactory identification results. To address this, target feature point detection can quickly pinpoint the target's distance. This determines whether the target is within the optimal observation range. If not (i.e., the target is far away and has few features in the image), a second camera can be used to capture images in the direction of the target's movement, thus capturing more images of the pipeline scene with biological features. Finally, by introducing a target biometrics algorithm, biological targets can be classified to determine their species, thereby achieving accurate safety monitoring. Furthermore, the above-described implementation of this application does not require numbering each identified rodent individual or identifying whether they belong to the same monitoring target; identification is based solely on species characteristics, achieving rodent control.

[0058] Optionally, the aforementioned executing entity performs pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment identification information, and further includes:

[0059] Step S1: In response to determining that the first target detection identifier indicates no biological features were detected, static target recognition is performed on the pipeline scene image sequence to generate pipeline anomaly identification information. The pipeline anomaly identification information includes pipeline static hazard identifiers. Here, a preset convolutional neural network can be used to perform static target recognition on the pipeline scene image sequence to obtain a pipeline static hazard identifier sequence. Furthermore, the pipeline static hazard identifier that is identical and has the highest number in the pipeline static hazard identifier sequence can be used as the pipeline anomaly identification information.

[0060] Step S2: The above-mentioned pipeline anomaly identification information and the pipeline environment anomaly type identifier that characterizes the type of pipeline static environment anomaly are determined as pipeline environment identification information. For example, the pipeline environment anomaly type identifier that characterizes the type of pipeline static environment anomaly is "2".

[0061] In practice, by generating static hazard indicators and abnormal environmental type indicators for pipelines, these can be used to reflect static hazards detected in airport underground pipelines, such as detected tunnel collapses, falling foreign objects, or even pipeline ruptures. This improves the comprehensiveness and accuracy of safety monitoring of airport underground pipelines.

[0062] Step 105: Update the historical pipeline environment record table based on the generated pipeline environment identification information set to generate an underground pipeline safety monitoring record table.

[0063] In some embodiments, the aforementioned executing entity can update the historical pipeline environment record table based on the generated pipeline environment identification information set to generate an underground pipeline safety monitoring record table. Specifically, the pipeline environment identification information can be added as a record to the historical pipeline environment record table in timestamp order to obtain the underground pipeline safety monitoring record table. Here, each record in the underground pipeline safety monitoring record table represents a single instance of animal presence detected in the underground passage.

[0064] Optionally, the aforementioned implementing entity can update the historical pipeline environment record table through the following steps: First, determine the timestamp corresponding to each pipeline environment identification information as the identification timestamp. Then, add the pipeline environment identification information to the historical pipeline environment record table in the following order: timestamp, pipeline dynamic environment anomaly type, pipeline hazardous animal identifier, pipeline static hazard identifier, sensor number, underground passage number, pipeline type number, camera number, and the corresponding clearest target sub-image, to obtain the underground pipeline safety monitoring record table. This enables accurate safety monitoring of airport underground pipelines.

[0065] In practice, when controlling rodents in underground pipelines at airports, rodent traps (such as rat traps and cages) and rodenticides are often placed in areas where rodents are suspected to be present. However, this approach is unlikely to be effective in controlling the problem comprehensively when the number of rodents in the area is uncertain.

[0066] Optionally, the aforementioned implementing entity may also include:

[0067] Step S1: For each corresponding pipeline dynamic environment anomaly type in the pipeline environment identification information set, a pipeline hazard animal activity route is generated based on at least one target feature point sequence corresponding to the pipeline environment identification information. Specifically, firstly, each target feature point in the at least one target feature point sequence corresponding to the pipeline environment identification information is transformed from the image coordinate system to the camera coordinate system to obtain a 3D target feature point sequence. Then, each 3D target feature point in the 3D target feature point sequence is transformed from the camera coordinate system to the airport 3D coordinate system of the current airport underground 3D pipeline model to obtain a transformed feature point sequence. Finally, each transformed feature point in the transformed feature point sequence is fitted to obtain the pipeline hazard animal activity route.

[0068] Step S2 involves generating a pipeline hazard animal activity distribution map using the generated set of pipeline hazard animal activity routes and historical animal activity routes, and then sending this map to the target display terminal. Specifically, each pipeline hazard animal activity route and each historical activity route can be used to create a three-dimensional pipeline hazard animal activity distribution map. Furthermore, each pipeline hazard animal activity route can be associated with corresponding records in the underground pipeline safety monitoring record table.

[0069] As an example, see Figure 4 .exist Figure 4 A camera 404 is installed within the passageway 401, allowing video recording when an anomaly is detected in the underground pipe 402 within the passageway 401. The recorded video of the pipe scene can then be used to generate a distribution map of the activity of pests affecting the pipes. This map shows multiple activity routes 403 of these pests. This not only provides a clear visual representation of the activity trajectories of rodents but also avoids the waste of computing and storage resources associated with continuous camera identification across all time zones.

[0070] In practice, during the safety monitoring of airport underground pipelines, the wide distribution of these pipelines results in a large area where invasive animals can roam. Relying solely on generated maps showing the distribution of these animals would be too labor-intensive and difficult to respond to promptly. Therefore, more granular monitoring and analysis results are needed to achieve precise monitoring in airport underground pipeline safety monitoring.

[0071] Optionally, the aforementioned implementing entity may also include:

[0072] Step S1 involves analyzing the activity routes of hazardous animals using the pipeline activity distribution map to generate an activity route analysis result set. This result set includes the coordinates of high-frequency activity intersections. First, the channel number groups corresponding to each activity route in the pipeline hazardous animal activity distribution map are determined, resulting in a channel number group set. Then, the number of each type of channel number in the channel number group set is determined, resulting in a channel number quantity set. Next, channel numbers whose quantity exceeds a preset threshold are identified as high-frequency channel numbers, resulting in a high-frequency channel number group. Finally, the coordinates of the intersections corresponding to each high-frequency channel in the high-frequency channel number group are determined as the high-frequency activity intersection coordinates. Thus, the activity route result analysis set is obtained.

[0073] In practice, because rodents are social animals, a higher frequency of animal activity in certain underground passages indicates a high probability of rodent nests within those passages. Therefore, by screening high-frequency underground passages along routes, the range of rodent nests can be located to some extent. This facilitates targeted rodent control. Specifically, the environment of airport underground pipelines is suitable for rodent survival, leading to rapid rodent reproduction. Furthermore, the complex and widely distributed nature of airport underground pipelines necessitates identifying the core areas (i.e., nest ranges) of rodents based on their social characteristics. This not only effectively controls rodent infestations but also significantly inhibits rodent reproduction, thereby greatly reducing the probability of rodent damage to underground pipelines.

[0074] Step S2: Add the above activity route analysis result set to the current three-dimensional pipeline model to obtain the target three-dimensional pipeline model. Specifically, the coordinates of high-frequency activity intersection points included in each of the above activity route analysis results can be marked in the current three-dimensional pipeline model, thereby obtaining the target three-dimensional pipeline model.

[0075] Step S3: In the aforementioned target 3D pipeline model, determine the airport pipeline security coordinate group corresponding to the coordinates of each high-frequency activity intersection point, thus obtaining the airport pipeline security coordinate group set. First, according to a preset time period (e.g., one week), group the records in the underground pipeline safety monitoring record table to obtain a sequence of grouped record groups. Then, determine the number of each record in each grouped record group as the animal activity frequency, obtaining an animal activity frequency sequence. Afterwards, a trend test algorithm can be used to determine whether the number of animals (rodents) in the underground pipeline shows an increasing trend over a continuous time period based on the above animal activity frequency sequence.

[0076] As an example, trend verification algorithms can include the Mann-Kendall (MK) trend test algorithm and linear regression algorithms. For instance, a linear regression slope test can be performed on the above animal activity frequency sequence to obtain the linear regression slope value. Here, the time period T (T is an integer, ranging from (-n, 1)) can be used as the independent variable for the linear regression slope test, and the animal activity frequency can be used as the dependent variable to obtain the linear regression slope value. Finally, if the linear regression slope value is greater than 0.8, it is determined that the number of animals in the airport underground pipelines has an increasing trend over the continuous time period. Otherwise, it is determined that the number of animals in the airport underground pipelines has not been increasing over the continuous time period.

[0077] Finally, in response to the determination that the animal population is in an increasing trend, an escalation prevention coefficient is generated. Here, a preset coefficient value corresponding to the increasing trend can be determined as the escalation prevention coefficient. Then, using the escalation prevention coefficient, the airport pipeline security coordinate set corresponding to the coordinates of each high-frequency activity intersection point is determined. This results in a set of airport pipeline security coordinate sets. Additionally, in response to the determination that the animal population is not in an increasing trend, a basic prevention coefficient is generated. Here, a preset coefficient value corresponding to the absence of an increasing trend can be determined as the basic prevention coefficient. Finally, using the basic prevention coefficient, the airport pipeline security coordinate set corresponding to the coordinates of each high-frequency activity intersection point is determined. This results in a set of airport pipeline security coordinate sets. Specifically, the escalation prevention coefficient and the basic prevention coefficient can represent the number of association layers of the high-frequency activity intersection point coordinates. Furthermore, after obtaining each airport pipeline security coordinate, duplicate airport pipeline security coordinates can be removed through deduplication.

[0078] As an example, an upgraded security coefficient of 2 indicates that for each high-frequency activity intersection point coordinate, the intersection points of the passages within two adjacent associated layers need to be determined as airport pipeline security coordinates. For instance, the intersection point of passage A and passage B is a1b1. The intersection point of passage B and passage C is b2c1. The intersection point of passage C and passage D is c2d2. Then, for intersection point a1b1, the corresponding first associated layer is intersection point b2c1, and the corresponding second associated layer is intersection point c2d2. Therefore, intersection points a1b1, b2c1, and c2d2 can all be determined as airport pipeline security coordinates. Alternatively, a basic security coefficient of 1 indicates that for each high-frequency activity intersection point coordinate, the intersection points of the passages within one adjacent associated layer need to be determined as airport pipeline security coordinates. Therefore, intersection points a1b1 and b2c1 can be determined as airport pipeline security coordinates.

[0079] In practice, by generating airport pipeline security coordinates, these coordinates can be used to record high-frequency locations of (rodent) activity, thereby providing more accurate and comprehensive security data for airport underground pipelines.

[0080] Step S4 involves updating the aforementioned target 3D pipeline model and the aforementioned airport pipeline security coordinate set to the target display terminal, so as to display the location corresponding to each airport pipeline security coordinate. This facilitates the precise placement of rodent-proof boxes according to the airport pipeline security coordinates. The rodent-proof boxes can be rodent traps or boxes containing rodenticide. Additionally, for other underground pipeline areas with traces of rodent activity (e.g., the number of channel numbers does not exceed a preset threshold), a small number of rodent-proof boxes can be installed within the channels to facilitate rodent control before rodents breed.

[0081] In practice, due to the large distribution area and complex structure of airport underground passages (pipelines), if rodent-proof boxes are only placed in certain passage points during rodent control (extermination) operations, it will be impossible to control rodent infestations in other areas in a timely manner. Even when placed in areas with rodent infestations, the effect may be poor due to an excessive or insufficient number of rodents. For example, if there are many rodents in a certain area of ​​the underground pipeline, insufficient rodent-proof boxes will result in inadequate control. If the number of rodents is small, placing the same number of rodent-proof boxes will not be effective. Therefore, this application uses image recognition to reconstruct the routes of rodent activity within underground pipelines, thereby determining the distribution of rodent activity. Secondly, the analysis can be used to locate rodent nesting areas with coarse granularity. Thus, rodent-proof boxes can be placed around rodent nesting areas based on the frequency of rodent activity routes without needing to accurately determine the rodent population size. Furthermore, the number of boxes can be adaptively adjusted according to rodent growth. This greatly improves the probability of rodent control and extermination, and inhibits rodent reproduction. This significantly reduces the probability of rodents damaging underground pipelines, thereby reducing safety hazards.

[0082] The above-described embodiments of this disclosure have the following beneficial effects: the airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model, as described in some embodiments of this disclosure, can improve the accuracy of airport underground pipeline safety monitoring, enabling timely and targeted handling of abnormal situations, thereby reducing airport safety hazards. Specifically, the reason for the existence of serious safety hazards is that airport underground pipelines are complex and cover a large area, while manual monitoring is prone to detection blind spots (for example, ignoring some hidden defects in the pipelines, or making it difficult to directly observe animal damage to underground pipelines due to animals' avoidance instincts, thus making it difficult to directly determine the factors causing pipeline anomalies), resulting in slow response speed to pipeline anomalies, and therefore, inadequate underground pipeline safety monitoring. Based on this, the airport pipeline safety monitoring method based on an airport underground three-dimensional pipeline model, as described in some embodiments of this disclosure, firstly acquires a multi-source equipment sensing dataset for airport underground pipelines, and then synchronizes the multi-source equipment sensing dataset to the airport underground three-dimensional pipeline model to obtain the current three-dimensional pipeline model. The multi-source sensing data is collected by sensors in the sensor assembly. Each sensor is bound to corresponding sensor coding information in the current airport underground 3D pipeline model. This sensor coding information includes the corresponding sensor coordinates, sensor number, sensor type, underground passage number, and pipeline type number. By introducing the airport underground 3D pipeline model, various sensors laid within the airport underground passages can be uniformly associated, replacing manual monitoring. Then, signal change analysis is performed on the multi-source sensing data in the aforementioned multi-source sensing dataset to generate a signal analysis result set. This signal change analysis determines whether the signal analysis results indicate a signal change, thereby identifying any anomalies in the underground pipeline. Next, in response to the determination that a signal analysis result indicating a signal change exists in the signal analysis result set, for each signal analysis result indicating a signal change, the corresponding camera in the underground pipeline is invoked to capture a video of the pipeline scene at the corresponding location, based on the current 3D pipeline model. The introduction of cameras allows for integration with sensors, enabling the capture of pipeline scene videos at abnormal locations. Finally, pipeline scene image recognition is performed on each captured pipeline scene video to generate pipeline environment identification information. The pipeline environment identification information includes pipeline environment anomaly type identifiers and pipeline anomaly identification information. The pipeline environment anomaly type identifiers represent either static or dynamic environmental anomalies, while the pipeline anomaly identification information corresponding to dynamic environmental anomalies includes identifiers of hazardous animals. Therefore, image recognition can be used to accurately identify the real factors causing pipeline anomalies. Finally, based on the generated pipeline environment identification information set, the historical pipeline environment record table is updated to generate an underground pipeline safety monitoring record table. This improves the accuracy of airport underground pipeline safety monitoring, thereby reducing airport safety hazards.

[0083] Further reference Figure 5 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of an airport pipeline safety monitoring device based on a three-dimensional underground pipeline model of an airport. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, the airport pipeline safety monitoring device based on the airport underground three-dimensional pipeline model can be specifically applied to various electronic devices.

[0084] like Figure 5 As shown, an airport pipeline safety monitoring device 500 based on an airport underground 3D pipeline model in some embodiments includes: an acquisition and synchronization unit 501, a signal change analysis unit 502, a calling unit 503, a pipeline scene image recognition unit 504, and an update unit 505. The acquisition and synchronization unit 501 is configured to acquire a multi-source device sensing dataset for the airport underground pipeline and synchronize the multi-source device sensing dataset to the airport underground 3D pipeline model to obtain the current 3D pipeline model. The multi-source device sensing data is collected by sensors in the sensor assembly. Each sensor is bound to corresponding sensor coding information in the current airport underground 3D pipeline model. The sensor coding information includes the corresponding sensor coordinates, sensor number, sensor type, underground passage number, and pipeline type number. The signal change analysis unit 502 is configured to perform signal change analysis on the multi-source device sensing data in the multi-source device sensing dataset to generate a signal analysis result set. Calling unit 503 is configured to, in response to determining that there is a signal analysis result in the above signal analysis result set that represents a signal change, call the corresponding camera in the underground pipeline to capture a pipeline scene video at the corresponding location for each signal analysis result that represents a signal change, based on the above current three-dimensional pipeline model. Pipeline scene image recognition unit 504 is configured to perform pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment identification information. The pipeline environment identification information includes a pipeline environment anomaly type identifier and pipeline anomaly identification information. The pipeline environment anomaly type identifier represents a static or dynamic environment anomaly type, and the pipeline anomaly identification information corresponding to the dynamic environment anomaly type includes a pipeline hazard animal identifier. Update unit 505 is configured to update the historical pipeline environment record table based on the generated pipeline environment identification information set to generate an underground pipeline safety monitoring record table.

[0085] It is understandable that the units recorded in the airport pipeline safety monitoring device 500 based on the airport underground three-dimensional pipeline model are similar to the reference units. Figure 1The steps described in the method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the airport pipeline safety monitoring device 500 based on the airport underground three-dimensional pipeline model and the units contained therein, and will not be repeated here.

[0086] The following is for reference. Figure 6 It shows a schematic diagram of the structure of an electronic device 600 (e.g., a computing device) suitable for implementing some embodiments of the present disclosure. Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0087] like Figure 6 As shown, electronic device 600 may include processing device 601 (e.g., central processing unit, graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from storage device 608 into random access memory (RAM) 603. The random access memory 603 also stores various programs and data required for the operation of electronic device 600. Processing device 601, read-only memory 602, and random access memory 603 are interconnected via bus 604. Input / output (I / O) interface 605 is also connected to bus 604.

[0088] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic device 600 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 An electronic device 600 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 6 Each box shown can represent a device or multiple devices as needed.

[0089] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a read-only memory 602. When the computer program is executed by the processing device 601, it performs the functions defined above in the methods of some embodiments of this disclosure.

[0090] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may 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. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0091] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0092] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: acquire a multi-source device sensing dataset for airport underground pipelines, and synchronize the aforementioned multi-source device sensing dataset to the airport underground three-dimensional pipeline model to obtain a current three-dimensional pipeline model, wherein the multi-source device sensing data is acquired by sensors in a sensor assembly, and each sensor is bound to corresponding sensor coding information in the aforementioned current airport underground three-dimensional pipeline model, the sensor coding information including corresponding sensor coordinates, sensor number, sensor type, underground passage number, and pipeline type number; perform signal change analysis on the multi-source device sensing data in the aforementioned multi-source device sensing dataset to generate a signal analysis result set; and respond to determining the aforementioned signal... The analysis results set contains signal analysis results representing signal changes. For each signal analysis result representing a signal change, based on the current 3D pipeline model, the corresponding camera in the underground pipeline is invoked to capture pipeline scene videos at the corresponding locations. Pipeline scene image recognition is performed on each captured pipeline scene video to generate pipeline environment identification information. The pipeline environment identification information includes pipeline environment anomaly type identifiers and pipeline anomaly identification information. The pipeline environment anomaly type identifiers represent pipeline static environment anomaly types or pipeline dynamic environment anomaly types. The pipeline anomaly identification information corresponding to the pipeline dynamic environment anomaly type includes pipeline hazard animal identifiers. Based on the generated pipeline environment identification information set, the historical pipeline environment record table is updated to generate an underground pipeline safety monitoring record table.

[0093] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0094] 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 disclosure. 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 the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can 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.

[0095] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0096] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for airport pipeline safety monitoring based on a three-dimensional underground pipeline model of an airport, characterized in that, include: A multi-source equipment sensing dataset for airport underground pipelines is obtained, and the multi-source equipment sensing dataset is synchronized to the airport underground three-dimensional pipeline model to obtain the current three-dimensional pipeline model. The multi-source equipment sensing data is collected by sensors in the sensor component. Each sensor is bound to corresponding sensor coding information in the current airport underground three-dimensional pipeline model. The sensor coding information includes the corresponding sensor coordinates, sensor number, sensor type, underground passage number and pipeline type number. Signal change analysis is performed on the multi-source device sensing data in the multi-source device sensing dataset to generate a signal analysis result set; In response to determining that there is a signal analysis result in the set of signal analysis results that represents a signal change, for each signal analysis result that represents a signal change, according to the current three-dimensional pipeline model, the corresponding camera in the underground pipeline is called to capture a video of the pipeline scene at the corresponding location; Each captured video of a pipeline scene undergoes pipeline scene image recognition to generate pipeline environment identification information. This information includes a pipeline environment anomaly type identifier and pipeline anomaly identification information. The pipeline environment anomaly type identifier represents either a static or dynamic environmental anomaly type. The pipeline anomaly identification information corresponding to a dynamic environmental anomaly type includes a pipeline hazard animal identifier, including: Image sampling is performed on the pipeline scene video to obtain a pipeline scene image sequence; The pipeline scene image sequence is subjected to fast target detection to generate a first target detection identifier, wherein the first target detection identifier represents whether biological features are detected; In response to determining that the first target detection identifier represents the detection of biological features, target feature point detection is performed on the pipeline scene image sequence to obtain a target feature point sequence; Based on the target feature point sequence, a target distance value sequence is determined, wherein the target distance value represents the distance between the detected target and the corresponding camera location; In response to determining that the target distance value sequence satisfies a first preset distance condition, biological target recognition is performed on the pipeline scene image sequence to obtain pipeline anomaly recognition information; The pipeline environment anomaly type identifier, which characterizes the type of dynamic environmental anomaly in the pipeline, and the pipeline anomaly identification information are determined as pipeline environment identification information. In response to determining that the target distance value sequence does not meet the first preset distance condition but meets the second preset distance condition, the second camera is invoked to capture a second scene video, and the second scene video is used to perform pipeline scene image recognition again to generate pipeline environment recognition information, wherein the second camera is a camera that is selected again; In response to determining that the first target detection identifier indicates that no biological features were detected, static target recognition is performed on the pipeline scene image sequence to generate pipeline anomaly identification information, wherein the pipeline anomaly identification information includes pipeline static hazard identifiers; The pipeline anomaly identification information and the pipeline environment anomaly type identifier that characterizes the type of static environment anomaly of the pipeline are determined as pipeline environment identification information; Based on the generated pipeline environment identification information set, the historical pipeline environment record table is updated to generate an underground pipeline safety monitoring record table.

2. The method according to claim 1, characterized in that, The airport's underground three-dimensional pipeline model is generated through the following steps: Based on the pipeline size data in the pre-set airport underground pipeline construction drawings, an underground three-dimensional tunnel model is established. The underground three-dimensional tunnel model includes a set of underground tunnel numbers, and each underground tunnel number represents a segment of underground tunnel. Based on the preset underground pipeline data file, simulated pipelines are established within the underground passage range in the underground three-dimensional passage model, and a corresponding pipeline type number is established for each simulated pipeline to obtain the initial three-dimensional pipeline model. Based on the preset underground sensor deployment data file, simulated sensors and simulated cameras corresponding to cameras are added to the initial three-dimensional pipeline model, and corresponding sensor coding information and camera coding information are established to obtain the airport underground three-dimensional pipeline model. The underground sensor deployment data file includes camera identifiers and camera coordinates corresponding to multiple cameras, and sensor identifiers and sensor coordinates corresponding to each sensor in the sensor assembly. There is a correspondence between camera identifiers and sensor identifiers, and each camera identifier corresponds to at least one underground passage number or at least one pipeline type number.

3. The method according to claim 2, characterized in that, For each signal analysis result representing a signal change, based on the current 3D pipeline model, the corresponding camera in the underground pipeline is invoked to capture pipeline scene video at the corresponding location, including: Based on the current three-dimensional pipeline model, the camera identifier corresponding to the signal analysis result is determined as the target camera identifier. The camera corresponding to the target camera identifier has the same underground passage number and pipeline type number as the signal analysis result. The camera corresponding to the target camera identifier is in standby mode. Perform camera detection on the camera corresponding to the target camera identifier to determine whether the camera is malfunctioning; In response to determining that the camera has not malfunctioned, the camera corresponding to the target camera identifier is controlled to adjust its shooting direction to capture the environment of the pipeline corresponding to the signal analysis result, thereby obtaining a pipeline scene video. The shooting direction is generated based on the sensor coordinates corresponding to the target camera identifier.

4. The method according to claim 3, characterized in that, The method further includes: For each pipeline environment identification information corresponding to a pipeline dynamic environment anomaly type in the pipeline environment identification information set, a pipeline hazard animal activity route is generated based on at least one target feature point sequence corresponding to the pipeline environment identification information. Using the generated set of pipeline hazard animal activity routes and historical animal activity routes, a pipeline hazard animal activity distribution map is generated, and the pipeline hazard animal activity distribution map is sent to the target display terminal.

5. The method according to claim 4, characterized in that, The method further includes: Activity routes of animals that cause harm through pipelines are analyzed to generate a set of activity route analysis results, which includes the coordinates of high-frequency activity intersection points. The activity route analysis result set is added to the current three-dimensional pipeline model to obtain the target three-dimensional pipeline model; In the target three-dimensional pipeline model, the airport pipeline security coordinate set corresponding to the coordinates of each high-frequency activity intersection point is determined, thus obtaining the airport pipeline security coordinate set; The target 3D pipeline model and the airport pipeline security coordinate set are updated to the target display terminal to display the location corresponding to each airport pipeline security coordinate.

6. An airport pipeline safety monitoring device based on an airport underground three-dimensional pipeline model, characterized in that, include: The acquisition and synchronization unit is configured to acquire a multi-source device sensing dataset for airport underground pipelines and synchronize the multi-source device sensing dataset to the airport underground three-dimensional pipeline model to obtain the current three-dimensional pipeline model. The multi-source device sensing data is acquired by sensors in the sensor component. Each sensor is bound to corresponding sensor coding information in the current airport underground three-dimensional pipeline model. The sensor coding information includes the corresponding sensor coordinates, sensor number, sensor type, underground passage number and pipeline type number. The signal change analysis unit is configured to perform signal change analysis on the multi-source device sensing data in the multi-source device sensing dataset to generate a signal analysis result set. The calling unit is configured to respond to determining that there is a signal analysis result in the set of signal analysis results that represents a signal change, and for each signal analysis result that represents a signal change, to call the corresponding camera in the underground pipeline to capture a video of the pipeline scene at the corresponding location based on the current three-dimensional pipeline model; The pipeline scene image recognition unit is configured to perform pipeline scene image recognition on each captured pipeline scene video to generate pipeline environment recognition information. This information includes a pipeline environment anomaly type identifier and pipeline anomaly identification information. The pipeline environment anomaly type identifier represents either a static or dynamic environment anomaly type. The pipeline anomaly identification information corresponding to a dynamic environment anomaly type includes a pipeline hazard animal identifier, including: Image sampling is performed on the pipeline scene video to obtain a pipeline scene image sequence; The pipeline scene image sequence is subjected to fast target detection to generate a first target detection identifier, wherein the first target detection identifier represents whether biological features are detected; In response to determining that the first target detection identifier represents the detection of biological features, target feature point detection is performed on the pipeline scene image sequence to obtain a target feature point sequence; Based on the target feature point sequence, a target distance value sequence is determined, wherein the target distance value represents the distance between the detected target and the corresponding camera location; In response to determining that the target distance value sequence satisfies a first preset distance condition, biological target recognition is performed on the pipeline scene image sequence to obtain pipeline anomaly recognition information; The pipeline environment anomaly type identifier, which characterizes the type of dynamic environmental anomaly in the pipeline, and the pipeline anomaly identification information are determined as pipeline environment identification information. In response to determining that the target distance value sequence does not meet the first preset distance condition but meets the second preset distance condition, the second camera is invoked to capture a second scene video, and the second scene video is used to perform pipeline scene image recognition again to generate pipeline environment recognition information, wherein the second camera is a camera that is selected again; In response to determining that the first target detection identifier indicates that no biological features were detected, static target recognition is performed on the pipeline scene image sequence to generate pipeline anomaly identification information, wherein the pipeline anomaly identification information includes pipeline static hazard identifiers; The pipeline anomaly identification information and the pipeline environment anomaly type identifier that characterizes the type of static environment anomaly of the pipeline are determined as pipeline environment identification information; The update unit is configured to update the historical pipeline environment record table based on the generated pipeline environment identification information set to generate an underground pipeline safety monitoring record table.

7. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 5.

8. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 5.

Images