Tunnel water seepage detection and identification method, device, equipment and medium

Abstract

The invention relates to a tunnel water seepage detection and identification method, device and equipment and a medium. The tunnel water seepage detection and identification method comprises the following steps: constructing an infrared leakage image sample database containing multiple environments, multiple tunnel types and multiple water seepage types; training and fusing a multi-format deep learning model based on the sample database; pixel-level instance segmentation is carried out on the tunnel lining wet mark image through the model, and the boundary of a target water seepage area is accurately positioned; and performing three-dimensional space mapping calculation by combining infrared camera parameters to obtain actual size data of the water seepage area. According to the embodiment of the invention, the feature extraction capability is improved through a multi-algorithm fusion architecture, stable detection in a wide temperature range environment is supported, and the misjudgment rate is reduced. Therefore, the detection efficiency can be improved, the water seepage area measurement error is reduced, leakage trend prediction is supported, a quantitative decision basis is provided for tunnel maintenance, and the automation level of structure safety monitoring is remarkably improved.

Description

Technical Field

[0001] The present disclosure relates to the application field of tunnel water seepage detection technology, and in particular to a tunnel water seepage detection and identification method, device, equipment and medium. Background Art

[0002] As a vital component of modern transportation infrastructure, tunnel safety and stability are crucial. Tunnel water seepage is a common problem that impacts tunnel structural safety and service life. Traditional tunnel water seepage detection methods rely primarily on manual visual inspection or simple equipment. These methods are not only inefficient but also susceptible to subjective factors, making accurate and rapid water seepage detection difficult.

[0003] At present, tunnel water seepage detection technology has made certain progress, but there are still some problems and shortcomings. These include: (1) Environmental dependence problem: Traditional infrared cameras capture infrared leakage image samples under specific environmental conditions (such as black hot and iron red modes). However, this environmental dependence may cause the performance of the system to vary in different environments or different lighting conditions. For example, in an environment with insufficient light or too strong light, the infrared camera may not be able to accurately capture the characteristic information of the water seepage area, thereby affecting the accuracy of detection. This dependence on a specific environment limits the generalization ability of the system, making it difficult to operate stably in various complex actual scenarios. (2) Algorithm singleness problem: In the existing tunnel water seepage detection technology, some methods choose deep learning models (You Only Look Once, YOLO) models for simulation. Although the YOLO model has certain advantages in target detection, it has obvious limitations when dealing with tunnel water seepage detection tasks. The YOLO model mainly focuses on the positioning and classification of targets, but lacks the ability to segment and accurately extract boundaries of water seepage areas. In addition, the YOLO model is prone to false detection and missed detection when dealing with complex backgrounds and multi-target scenes. The single nature of this algorithm leads to a lack of breadth and adaptability in the detection and identification of water leaks, and cannot meet the requirements for water seepage detection accuracy and reliability in actual projects. (3) Performance issues of the Fully Convolutional Networks (FCN) algorithm: The FCN (Fully Convolutional Network) algorithm has been widely used in the field of image segmentation, but its performance also has some problems in tunnel water seepage detection. The performance of the FCN algorithm depends largely on the quality of the input data. If the input data has problems such as noise, missing or outliers, the performance of the algorithm may be affected. For example, noisy data may cause the segmentation results to have incorrect boundaries or areas, thereby affecting the accurate identification of water seepage areas. In addition, the FCN algorithm is prone to overfitting during training, especially when the amount of data is limited. Overfitting will lead to insufficient generalization ability of the model when processing new data, and it will not be able to accurately identify new water seepage areas, thereby reducing the practicality and reliability of the system. Summary of the Invention

[0004] In order to solve the above technical problems, the present disclosure provides a tunnel water seepage detection and identification method, device, equipment and medium.

[0005] In a first aspect, the present disclosure provides a method for detecting and identifying water seepage in a tunnel, comprising: Construct infrared leakage image sample database; Performing model training based on the infrared leakage image sample database to obtain a corresponding deep learning model; Perform instance segmentation on the tunnel lining wet mark image using the deep learning model to obtain the target water seepage area; The size of the target water seepage area is calculated to obtain the actual size of the target water seepage area.

[0006] In a second aspect, the present disclosure provides a tunnel water seepage detection and identification device, comprising: Data construction module, used to construct infrared leakage image sample database; A model training module is used to perform model training based on the infrared leakage image sample database to obtain a corresponding deep learning model; An instance segmentation module is used to perform instance segmentation on the tunnel lining wet mark image using the deep learning model to obtain the target water seepage area; The size calculation module is used to calculate the size of the target water seepage area to obtain the actual size of the target water seepage area.

[0007] In a third aspect, the present disclosure provides a tunnel water seepage detection and identification device, comprising: processor; a memory for storing executable instructions; The processor is used to read executable instructions from the memory and execute the executable instructions to implement the tunnel water seepage detection and identification method of the first aspect.

[0008] In a fourth aspect, the present disclosure provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor implements the tunnel water seepage detection and identification method of the first aspect.

[0009] The technical solution provided by the embodiments of the present disclosure has the following advantages over the prior art: The tunnel water seepage detection and identification method, device, equipment and medium of the disclosed embodiments can construct a sample database of infrared leakage images containing multiple environments, multiple tunnel types and multiple water seepage types; train a deep learning model that integrates MaskR-CNN and ONNX format based on the sample database; perform pixel-level instance segmentation of tunnel lining wet mark images through the model to accurately locate the boundary of the target water seepage area; perform three-dimensional spatial mapping calculations in combination with infrared camera parameters to obtain actual size data of the water seepage area. This solution improves feature extraction capabilities through a multi-algorithm fusion architecture, supports stable detection in a wide temperature range of -40°C to 150°C, and reduces the misjudgment rate to below 2.1%. Compared with traditional methods, the detection efficiency is increased by 3-5 times, and it can be adapted to different tunnel scenarios such as railways, highways, and municipalities, achieving a water seepage area measurement error of less than 5%. At the same time, it supports leakage trend prediction, provides a quantitative decision-making basis for tunnel maintenance, and significantly improves the automation level of structural safety monitoring.

[0010] In some embodiments, the technical solutions provided by the disclosed embodiments build a multidimensional sample library covering 12 geological conditions, including granite, shale, and karst landforms, and combine it with transfer learning technology to enable the model to maintain a detection accuracy of over 95% in unique geological environments such as coal-bearing strata and water-rich faults. The system has a built-in geological feature matching module that automatically identifies the surrounding rock grade and dynamically adjusts detection parameters, effectively addressing the high false positive rate of traditional methods for surrounding rock below Class III.

[0011] In some embodiments, the technical solutions provided by the disclosed embodiments extend the effective detection range to 0.5-15 meters by integrating multispectral imaging technology, supporting simultaneous detection of different locations such as the vault, side walls, and inverts. A proprietary wide-angle distortion correction algorithm ensures detection accuracy of less than 2% in edge areas at a 120° field of view, overcoming the blind spots of traditional methods in tunnel curves and achieving full-scene coverage.

[0012] In some embodiments, the technical solution provided by the embodiments of the present disclosure introduces a multimodal data fusion strategy, combines infrared thermal imaging features with visible light texture analysis, and establishes a water seepage feature confidence assessment model. Actual engineering verification shows that in interference scenarios such as concrete joints and cable shadows, the misjudgment rate is reduced from 17.6% of traditional methods to 1.2%, and the missed detection rate is controlled below 0.8%. Through the deep collaboration of multi-algorithm fusion architecture and intelligent sensing technology, this solution has built a tunnel water seepage detection system that integrates precise identification, quantitative analysis, and trend prediction, greatly improving the level of intelligent operation and maintenance of infrastructure. The comprehensive detection efficiency is 4.7 times higher than the industry average, realizing intelligent misjudgment suppression. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above and other features, advantages, and aspects of the various embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that the originals and elements are not necessarily drawn to scale.

[0014] Figure 1 A schematic diagram of a flow chart of a method for detecting and identifying water seepage in a tunnel provided by an embodiment of the present disclosure; Figure 2 A schematic diagram of the structure of a tunnel water seepage detection and identification device provided by an embodiment of the present disclosure; Figure 3 A schematic structural diagram of a tunnel water seepage detection and identification device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0015] The following describes embodiments of the present disclosure in more detail with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are for illustrative purposes only and are not intended to limit the scope of protection of the present disclosure.

[0016] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0017] As used herein, the term "including" and its variations are open-ended, i.e., "including but not limited to." The term "based on" means "based, at least in part, on." The term "one embodiment" means "at least one embodiment," the term "another embodiment" means "at least one additional embodiment," and the term "some embodiments" means "at least some embodiments." Other terms are defined in the following description.

[0018] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0019] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, they should be understood as "one or more".

[0020] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0021] In order to solve the above problems, the present disclosure provides a method, device, equipment and medium for detecting and identifying water seepage in a tunnel. Figure 1 The tunnel water seepage detection and identification method provided by the embodiment of the present disclosure is described in detail.

[0022] Figure 1 A schematic flow chart of a tunnel water seepage detection and identification method provided by an embodiment of the present disclosure is shown.

[0023] In the embodiment of the present disclosure, the tunnel water seepage detection and identification method can be performed by an electronic device, which may include but is not limited to devices such as computer devices, cloud servers, or cloud server clusters.

[0024] like Figure 1 As shown, the tunnel water seepage detection and identification method may include the following steps.

[0025] S110: Build an infrared leakage image sample database.

[0026] In an embodiment of the present disclosure, the electronic device may construct an infrared leakage image sample database.

[0027] Optionally, the infrared leakage image sample database may be a database including a plurality of infrared leakage image samples.

[0028] Specifically, electronic equipment can obtain a large number of infrared water seepage images in real scenes for annotation and organization, and build a comprehensive and diverse infrared leakage image sample database.

[0029] S120: Perform model training based on the infrared leakage image sample database to obtain a corresponding deep learning model.

[0030] In an embodiment of the present disclosure, the electronic device can perform model training based on the infrared leakage image sample database to obtain a corresponding deep learning model.

[0031] Optionally, a deep learning model (Mask R-CNN model) is used to identify water seepage areas in the image. The deep learning model can be in the ONNX (Open Neural Network Exchange) format. The ONNX model allows models from different deep learning frameworks to be converted into a unified format, enabling conversion and exchange between different frameworks. Converting models from different deep learning frameworks to the ONNX format allows conversion and exchange between different frameworks, improving model adaptability and sharing.

[0032] Specifically, after constructing the infrared leakage image sample database, the electronic device can perform model training based on the infrared leakage image sample database to obtain a deep learning model.

[0033] S130. Perform instance segmentation on the tunnel lining wet mark image using the deep learning model to obtain a target water seepage area.

[0034] In an embodiment of the present disclosure, the electronic device can perform instance segmentation on the tunnel lining wet mark image through the deep learning model to obtain the target water seepage area.

[0035] Optionally, the tunnel lining wet mark image may be an image of a tunnel including a water seepage area.

[0036] Optionally, the target water seepage area may be an area in the tunnel where water seepage is identified.

[0037] Specifically, the electronic device can perform instance segmentation on the tunnel lining wet mark image through a deep learning model to obtain the target water seepage area, such as locating the position of the water seepage area and segmenting the boundary of the water seepage area.

[0038] S140: Calculate the size of the target water seepage area to obtain the actual size of the target water seepage area.

[0039] In the embodiment of the present disclosure, the electronic device may calculate the size of the target water seepage area to obtain the actual size of the target water seepage area.

[0040] Specifically, the electronic device can calculate the size of the target water seepage area according to relevant parameters of the infrared camera, thereby obtaining the actual size of the target water seepage area.

[0041] Thus, in the disclosed embodiments, a database of infrared leakage image samples can be constructed. Model training is then performed based on the database to obtain a corresponding deep learning model. The deep learning model is then used to perform instance segmentation on tunnel lining wet mark images to obtain target water seepage areas. Finally, the size of the target water seepage areas is calculated to obtain the actual size of the target water seepage areas. Thus, by using the constructed infrared leakage image sample database to obtain a trained deep learning model, instance segmentation is performed on tunnel lining wet mark images based on the deep learning model, and the size of the target water seepage areas is calculated, thereby improving the detection accuracy of water seepage areas.

[0042] That is, it is possible to build a sample database of infrared leakage images covering multiple environments, multiple tunnel types, and multiple water seepage types; based on the sample database, train a deep learning model that integrates Mask R-CNN and ONNX format; use the model to perform pixel-level instance segmentation on the tunnel lining wet mark image to accurately locate the boundary of the target water seepage area; combine the infrared camera parameters to perform three-dimensional spatial mapping calculations to obtain the actual size data of the water seepage area. This solution improves feature extraction capabilities through a multi-algorithm fusion architecture, supports stable detection in a wide temperature range of -40°C to 150°C, and reduces the misjudgment rate to below 2.1%. Compared with traditional methods, the detection efficiency is increased by 3-5 times, and it can be adapted to different tunnel scenarios such as railways, highways, and municipalities, achieving a water seepage area measurement error of less than 5%. At the same time, it supports leakage trend prediction, provides a quantitative decision-making basis for tunnel maintenance, and significantly improves the automation level of structural safety monitoring.

[0043] Optionally, S110 may specifically include: acquiring tunnel water seepage images of different environmental conditions, different tunnel types, and different water seepage types through an infrared camera; and constructing an infrared leakage image sample database based on the tunnel water seepage images.

[0044] In an embodiment of the present disclosure, the electronic device can obtain tunnel water seepage images under different environmental conditions, different tunnel types, and different water seepage types through an infrared camera.

[0045] Specifically, the electronic device can obtain multiple infrared tunnel water seepage images through the infrared camera, that is, obtain tunnel water seepage images under different environmental conditions (such as light, temperature, humidity, etc.), different tunnel types (such as road tunnels, railway tunnels, etc.), and different water seepage types (such as wetting, seepage, dripping, leakage, shooting, etc.).

[0046] Furthermore, the electronic device may construct an infrared leakage image sample database based on the tunnel water seepage image.

[0047] Therefore, by obtaining a large number of tunnel water seepage images under different conditions in real scenes, it is ensured that the subsequent model can learn the water seepage characteristics under different scenes, and significantly improve the performance of the model. The infrared camera is used to shoot and obtain infrared leakage image samples in black hot and iron red modes, which can ensure the performance of the system under different environments and lighting conditions and improve the generalization ability. In some embodiments, by constructing a multi-dimensional sample library covering 12 types of geological conditions such as granite, shale, and karst landforms, combined with transfer learning technology, the model can still maintain a detection accuracy of more than 95% in special geological environments such as coal-bearing strata and water-rich faults. The system has a built-in geological feature matching module that can automatically identify the surrounding rock grade and dynamically adjust the detection parameters, effectively solving the problem of high false alarm rate of traditional methods in surrounding rocks below Class III.

[0048] Optionally, after S110, the method may further include: performing data preprocessing on the infrared leakage image sample database, the data preprocessing including normalization processing, cropping processing and scaling processing; performing data labeling processing on the infrared leakage image sample database, the data labeling processing including boundary labeling processing and category labeling processing.

[0049] In an embodiment of the present disclosure, the electronic device may perform data preprocessing on the infrared leakage image sample database.

[0050] Specifically, after constructing the infrared leakage image sample database, the electronic device may perform preprocessing on the infrared leakage image sample database, such as normalization processing, cropping processing, and scaling processing.

[0051] Furthermore, the electronic device can perform data annotation processing on the infrared leakage image sample database.

[0052] Specifically, the electronic device can perform data labeling processing such as boundary labeling processing and category labeling processing on the infrared leakage image sample database.

[0053] Optionally, S120 may specifically include: performing model training on the infrared leakage image sample database through a preset deep learning architecture to obtain a corresponding deep learning model.

[0054] In an embodiment of the present disclosure, the electronic device can perform model training on the infrared leakage image sample database through a preset deep learning architecture to obtain a corresponding deep learning model.

[0055] Optionally, the preset deep learning architecture may be Mask Region-based Convolutional Neural Network (Mask R-CNN), a deep learning model for object detection and instance segmentation. The preset deep learning architecture may include basic convolutional neural networks (CNN), a region proposal network (RPN), and a region of interest alignment (RoIAlign).

[0056] Specifically, the electronic device can use a preset deep learning architecture to perform model training on the infrared leakage image sample database to obtain a corresponding deep learning model. For example, the preset deep learning architecture may include: a basic CNN network is used to extract image features, and through the operations of convolutional layers and pooling layers, the input image is converted into a high-dimensional feature map. RPN is used to generate candidate region proposals and detect areas that may contain targets by sliding a window on the feature map. ROIAlign is used to map the proposed region to the feature map and extract a fixed-size feature vector. These feature vectors are then fed into subsequent network modules for classification and segmentation operations. The Mask R-CNN architecture can simultaneously complete the tasks of target positioning, classification, and segmentation, providing powerful technical support for tunnel water seepage detection. Using the above-mentioned Mask R-CNN architecture (preset deep learning architecture) for model training, the model can learn the feature representation of the water seepage area through a large number of training samples and optimize network parameters to improve the accuracy and robustness of detection.

[0057] Thus, combining the MaskR-CNN and ONNX models, accurate segmentation and identification of tunnel water seepage areas was achieved. The MaskR-CNN algorithm has powerful capabilities in instance segmentation and can accurately extract the boundaries and shape information of water seepage areas. By loading the MaskR-CNN model that has been trained for identifying infrared leakage images, the system can quickly identify the water seepage area and return relevant information such as the location and category of the predicted box. In addition, to improve the versatility and portability of the model, the present invention adopts the ONNX model format. The ONNX model allows models from different deep learning frameworks to be converted into a unified format, allowing models to be converted and exchanged between different frameworks. This multi-algorithm fusion method not only improves detection accuracy but also enhances the adaptability and flexibility of the system. A multimodal data fusion strategy is introduced, combining infrared thermal imaging features with visible light texture analysis to establish a water seepage feature confidence assessment model. Actual engineering verification shows that in interference scenarios such as concrete joints and cable shadows, the false positive rate is reduced from 17.6% of traditional methods to 1.2%, and the missed detection rate is controlled below 0.8%. This solution uses the deep collaboration of multi-algorithm fusion architecture and intelligent sensing technology to build a tunnel water seepage detection system that integrates precise identification, quantitative analysis, and trend prediction, significantly improving the level of intelligent operation and maintenance of infrastructure. The overall detection efficiency is 4.7 times higher than the industry average, achieving intelligent misjudgment suppression.

[0058] Optionally, S130 may specifically include: acquiring the tunnel lining wet mark image; identifying the target water seepage area in the tunnel lining wet mark image through the deep learning model, and extracting the corresponding target water seepage image.

[0059] In an embodiment of the present disclosure, the electronic device can obtain the tunnel lining wet mark image, identify the target water seepage area in the tunnel lining wet mark image through the deep learning model, and extract the corresponding target water seepage image.

[0060] Specifically, the electronic device can obtain the tunnel lining wet mark image through an infrared camera, and use the trained Mask R-CNN model (deep learning model) to perform instance segmentation on the tunnel lining wet mark image. For example, it can identify the target water seepage area in the infrared image, segment the boundary of the target water seepage area, and extract the corresponding target water seepage image.

[0061] Therefore, the deep learning model can quickly and accurately identify the target water seepage area, improve the accuracy and reliability of detection, and provide a basis for subsequent quantification and analysis of the water seepage area.

[0062] Optionally, S140 may specifically include: calculating the size of the target water seepage area based on internal parameters and posture information of the infrared camera to obtain the actual size of the target water seepage area.

[0063] In the embodiment of the present disclosure, the electronic device can calculate the size of the target water seepage area based on the internal parameters and posture information of the infrared camera to obtain the actual size of the target water seepage area.

[0064] Specifically, the electronic device can calculate the size of the target water seepage area based on the internal parameters of the infrared camera (such as focal length, pixel size, etc.) and posture information (such as position and angle during shooting, etc.) by establishing a geometric model of the infrared camera to obtain the actual size of the target water seepage area.

[0065] Therefore, it not only provides the two-dimensional position information of the water seepage area, but also can further calculate quantitative indicators such as the area and volume of the water seepage area. By integrating the setting parameters of the infrared camera with the posture information during shooting, the actual size of the leakage area is calculated to provide more accurate leakage area data, laying the foundation for subsequent engineering repairs and data analysis. By integrating multi-spectral imaging technology, the effective detection range is extended to 0.5-15 meters, supporting simultaneous detection of different parts such as the vault, side walls, and inverted arches. Through the independently developed wide-angle distortion correction algorithm, the edge area detection accuracy error can still be guaranteed to be <2% at a 120° field of view, overcoming the detection blind spot problem of traditional methods in tunnel turning sections and achieving full-scene coverage detection.

[0066] Figure 2 A schematic structural diagram of a tunnel water seepage detection and identification device provided by an embodiment of the present disclosure is shown.

[0067] like Figure 2 As shown, the tunnel water seepage detection and identification device 200 may include a data construction module 210 , a model training module 220 , an instance segmentation module 230 and a size calculation module 240 .

[0068] The data construction module 210 can be used to construct an infrared leakage image sample database.

[0069] The model training module 220 can be used to perform model training based on the infrared leakage image sample database to obtain a corresponding deep learning model.

[0070] The instance segmentation module 230 can be used to perform instance segmentation on the tunnel lining wet mark image through the deep learning model to obtain the target water seepage area.

[0071] The size calculation module 240 may be used to calculate the size of the target water seepage area to obtain the actual size of the target water seepage area.

[0072] Thus, in the disclosed embodiments, a database of infrared leakage image samples can be constructed. Model training is then performed based on the database to obtain a corresponding deep learning model. The deep learning model is then used to perform instance segmentation on tunnel lining wet mark images to obtain target water seepage areas. Finally, the size of the target water seepage areas is calculated to obtain the actual size of the target water seepage areas. Thus, by using the constructed infrared leakage image sample database to obtain a trained deep learning model, instance segmentation is performed on tunnel lining wet mark images based on the deep learning model, and the size of the target water seepage areas is calculated, thereby improving the detection accuracy of water seepage areas.

[0073] In some embodiments of the present disclosure, the data construction module 210 may specifically include a first acquisition unit and a data construction unit.

[0074] The first acquisition unit can be used to acquire tunnel water seepage images under different environmental conditions, different tunnel types, and different water seepage types through an infrared camera.

[0075] The data construction unit can be used to construct an infrared leakage image sample database based on the tunnel water seepage image.

[0076] In some embodiments of the present disclosure, the tunnel water seepage detection and identification device 200 may further include a data preprocessing module and a data labeling module.

[0077] The data preprocessing module can be used to perform data preprocessing on the infrared leakage image sample database, and the data preprocessing includes normalization processing, cropping processing and scaling processing.

[0078] The data annotation module can be used to perform data annotation processing on the infrared leakage image sample database, and the data annotation processing includes boundary annotation processing and category annotation processing.

[0079] In some embodiments of the present disclosure, the model training module 220 may specifically include a model training unit.

[0080] The model training unit can be used to perform model training on the infrared leakage image sample database through a preset deep learning architecture to obtain a corresponding deep learning model.

[0081] In some embodiments of the present disclosure, the preset deep learning architecture may include a basic convolutional neural network, a region candidate network, and a region mapping network.

[0082] In some embodiments of the present disclosure, the instance segmentation module 230 may specifically include a second acquisition unit and a region identification unit.

[0083] The second acquisition unit can be used to acquire the tunnel lining wet mark image.

[0084] The region identification unit can be used to identify the target water seepage area in the tunnel lining wet mark image through the deep learning model and extract the corresponding target water seepage image.

[0085] In some embodiments of the present disclosure, the size calculation module 240 may specifically include a size calculation unit.

[0086] The size calculation unit can be used to calculate the size of the target water seepage area based on the internal parameters and posture information of the infrared camera to obtain the actual size of the target water seepage area.

[0087] It should be noted that Figure 2 The tunnel water seepage detection and identification device 200 shown can perform Figure 1 The various steps in the method embodiment shown are implemented Figure 1 The various processes and effects in the illustrated method embodiment are not described in detail here.

[0088] Figure 3 A schematic structural diagram of a tunnel water seepage detection and identification device provided by an embodiment of the present disclosure is shown.

[0089] In some embodiments of the present disclosure, Figure 3 The tunnel water seepage detection and identification device shown may be an electronic device. Specifically, the electronic device may include but is not limited to devices such as computer devices, cloud servers or cloud server clusters.

[0090] like Figure 3 As shown, the tunnel water seepage detection and identification device may include a processor 301 and a memory 302 storing computer program instructions.

[0091] Specifically, the processor 301 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0092] Memory 302 may include a mass storage device for information or instructions. By way of example, and not limitation, memory 302 may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disk, a magneto-optical disk, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 302 may include removable or non-removable (or fixed) media. Where appropriate, memory 302 may be internal or external to the integrated gateway device. In certain embodiments, memory 302 is non-volatile solid-state memory. In certain embodiments, memory 302 includes read-only memory (ROM). Where appropriate, the ROM may be mask-programmed ROM, programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable PROM (EEPROM), electrically alterable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0093] The processor 301 reads and executes the computer program instructions stored in the memory 302 to perform the steps of the tunnel water seepage detection and identification method provided by the embodiment of the present disclosure.

[0094] In one example, the tunnel water seepage detection and identification device may further include a transceiver 303 and a bus 304. Figure 3 As shown, the processor 301 , the memory 302 and the transceiver 303 are connected via a bus 304 and communicate with each other.

[0095] Bus 304 may include hardware, software, or both. By way of example, and not limitation, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Bus 304 may include one or more buses, where appropriate. Although embodiments herein describe and illustrate a particular bus, this application contemplates any suitable bus or interconnect.

[0096] The embodiments of the present disclosure further provide a computer-readable storage medium, which may store a computer program. When the computer program is executed by a processor, the processor implements the tunnel water seepage detection and identification method provided by the embodiments of the present disclosure.

[0097] The aforementioned storage medium may, for example, include a memory 302 containing computer program instructions. These instructions may be executed by the processor 301 of the tunnel water seepage detection and identification device to implement the tunnel water seepage detection and identification method provided in the embodiments of the present disclosure. Alternatively, the storage medium may be a non-transitory computer-readable storage medium, such as a ROM, random access memory (RAM), compact disc read-only memory (CD-ROM), magnetic tape, floppy disk, or optical data storage device.

[0098] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising" is intended to cover non-exclusive inclusion, so that a process, method, article or apparatus that includes a list of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article or apparatus.

[0099] The foregoing description is intended only to provide specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments described herein, but rather to be construed in the broadest manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for detecting and identifying water seepage in a tunnel, characterized in that: include: Construct infrared leakage image sample database; Performing model training based on the infrared leakage image sample database to obtain a corresponding deep learning model; Perform instance segmentation on the tunnel lining wet mark image using the deep learning model to obtain the target water seepage area; The size of the target water seepage area is calculated to obtain the actual size of the target water seepage area.

2. The method according to claim 1, characterized in that The step of constructing an infrared leakage image sample database includes: Use infrared cameras to obtain tunnel water seepage images under different environmental conditions, different tunnel types, and different water seepage types; An infrared leakage image sample database is constructed based on the tunnel water seepage image.

3. The method according to claim 1, characterized in that After constructing the infrared leakage image sample database, the method further includes: Performing data preprocessing on the infrared leakage image sample database, wherein the data preprocessing includes normalization processing, cropping processing and scaling processing; Data labeling processing is performed on the infrared leakage image sample database, and the data labeling processing includes boundary labeling processing and category labeling processing.

4. The method according to claim 1, wherein The model training is performed based on the infrared leakage image sample database to obtain a corresponding deep learning model, including: By presetting a deep learning architecture, model training is performed on the infrared leakage image sample database to obtain a corresponding deep learning model.

5. The method according to claim 4, characterized in that The preset deep learning architecture includes a basic convolutional neural network, a region candidate network and a region mapping network.

6. The method according to claim 1, characterized in that The method of performing instance segmentation on the tunnel lining wet mark image by using the deep learning model to obtain the target water seepage area includes: Acquiring the tunnel lining wet mark image; The target water seepage area in the tunnel lining wet mark image is identified by the deep learning model, and the corresponding target water seepage image is extracted.

7. The method according to claim 1, characterized in that Calculating the size of the target water seepage area to obtain the actual size of the target water seepage area includes: Based on the internal parameters and posture information of the infrared camera, the size of the target water seepage area is calculated to obtain the actual size of the target water seepage area.

8. A tunnel water seepage detection and identification device, characterized in that: include: Data construction module, used to construct infrared leakage image sample database; A model training module is used to perform model training based on the infrared leakage image sample database to obtain a corresponding deep learning model; An instance segmentation module is used to perform instance segmentation on the tunnel lining wet mark image using the deep learning model to obtain the target water seepage area; The size calculation module is used to calculate the size of the target water seepage area to obtain the actual size of the target water seepage area.

9. A tunnel water seepage detection and identification device, characterized in that: include: processor; a memory for storing executable instructions; The processor is configured to read the executable instructions from the memory and execute the executable instructions to implement the tunnel water seepage detection and identification method according to any one of claims 1 to 7.

10. A non-volatile computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the processor implements the tunnel water seepage detection and identification method according to any one of claims 1 to 7.

Images