A method and system for acquiring information on the surrounding rock and lining structure of an operating tunnel.

Abstract

This invention discloses a method and system for acquiring information on the surrounding rock and lining structure of an operating tunnel, relating to the field of tunnel information recognition technology. The invention acquires images of the tunnel lining structure from different angles and distances across its entire range, extracts characteristic quantities reflecting crack width and direction from the lining images, and uses these characteristic quantities, tunnel lining point cloud data, and a three-dimensional model of the lining to train an image recognition model, thus obtaining a smart recognition model for tunnel lining scanning. Furthermore, it uses a drone to detect multi-dimensional information about the tunnel's surrounding rock, labels the surrounding rock grade based on the multi-dimensional data, and trains a three-dimensional image recognition model based on the surrounding rock grade to obtain a three-dimensional image model of apparent defects in the surrounding rock. Finally, by combining the smart recognition model for tunnel lining scanning and the three-dimensional image model of apparent defects in the surrounding rock, the invention accurately identifies defects and internal crack information in the tunnel's surrounding rock and lining structure, providing a more precise solution for subsequent tunnel construction and repair.

Description

Technical Field

[0001] This invention relates to the field of tunnel information identification technology, and in particular to a method, system, electronic device and storage medium for acquiring information on the surrounding rock and lining structure of an operating tunnel. Background Technology

[0002] The lining in a tunnel refers to the permanent structure that supports and maintains the long-term stability and durability of the tunnel. Its function is to support and maintain the stability of the tunnel, maintain the space required for vehicle operation, prevent the surrounding rock from becoming unstable, and reduce the impact of groundwater. Therefore, tunnel lining must have sufficient strength, durability, and certain resistance to freezing, seepage, and erosion. Tunnel lining is mainly composed of several parts, including the arch, sidewalls, invert, and floor. Drainage ditches are also installed inside the tunnel to drain water. Since the rock mass within the range of gravity stress in a tunnel is usually called the surrounding rock, and the properties of the rock mass vary, different lining forms can be used in different surrounding rocks.

[0003] For tunnels with long operating histories and those requiring subsequent renovation and expansion, it is crucial to quickly, safely, and accurately identify surface defects in the tunnel lining structure, such as cracks, seepage, and spalling, as well as information on surrounding rock fissures, joints, and seepage. Currently, detection methods for operating tunnels are primarily limited to the surface of the lining. For renovated and expanded tunnels, due to structural modifications, it is difficult to simultaneously identify both lining defects and surrounding rock flaws. Furthermore, obtaining such information typically relies on specialized personnel going into the operating tunnel to use scanning equipment for measurement and recording.

[0004] In recent years, with the development of drones, lightweight binocular vision scanning, photogrammetry, and sensor technology, especially the research and development and production of miniaturized and lightweight drones and aerial photography equipment, the application of drone lightweight binocular vision scanning and drone oblique photography in tunnel engineering has been increasing. Drones can be used to scan the information of tunnel lining and surrounding rock. However, the identification of the scanning results mainly relies on a combination of the subjective experience of professionals and preliminary computer identification to detect the information of tunnel lining and surrounding rock. This method reduces the accuracy of the identification of the quality of tunnel lining and surrounding rock, which affects the later construction progress. Summary of the Invention

[0005] This invention provides a method and system for acquiring information on the surrounding rock and lining structure of operating tunnels, which can solve the problem that existing methods reduce the accuracy of identifying the quality of the tunnel lining surrounding rock.

[0006] This invention provides a method for obtaining information on the surrounding rock and lining structure of an operating tunnel, comprising the following steps:

[0007] Acquire 3D point cloud data of the lining of the operating tunnel;

[0008] Acquire multiple images of the tunnel lining at different angles and distances across the entire lining area; extract local brightness information from the multiple lining images and decompose the local brightness information into frequency information and orientation information; obtain surface defect feature data of the tunnel lining containing crack width and orientation features based on the frequency and orientation information.

[0009] Based on 3D point cloud data and apparent defect feature data, a smart recognition model for tunnel lining scanning is generated based on a real-world model of the tunnel lining.

[0010] Multidimensional data of the internal information of the surrounding rock behind the tunnel lining is obtained, the strength of the surrounding rock is identified based on the multidimensional data, and a three-dimensional image model of the apparent defects of the surrounding rock is generated.

[0011] The lining image of the tunnel to be tested is input into the intelligent recognition model of tunnel lining scanning to obtain the apparent defect blocks of the tunnel lining; the multi-dimensional data of the internal information of the surrounding rock behind the tunnel lining is input into the three-dimensional image model of the apparent defects of the surrounding rock to obtain the strength level of the surrounding rock; and the defect information of the tunnel lining and surrounding rock structure is identified based on the apparent defect blocks of the tunnel lining and the strength level of the surrounding rock.

[0012] Preferably, acquiring the three-dimensional point cloud data of the operating tunnel lining includes:

[0013] Multiple drones were used to conduct a full-range scan of the operating tunnel, and at the same time, the lining structure of the operating tunnel was photographed from multiple angles, including vertical, inclined and horizontal angles, to obtain multiple lining images of the operating tunnel lining structure.

[0014] Multiple lining images were stitched together using 3D image modeling software and combined with omnidirectional scanning data to obtain a real-world 3D model of the operating tunnel lining structure, and the 3D point cloud data of the operating tunnel lining in the real-world 3D model was identified.

[0015] Preferably, after acquiring the real-world 3D model and 3D point cloud data of the lining structure, manual annotation is used to mark the lining cracks, seepage, spalling, crack orientation, density, width, extension, water outlet points, and water outflow of the tunnel lining structure.

[0016] Preferably, the step of decomposing local brightness information into frequency information and direction information includes:

[0017] After extracting local brightness information from multiple lining images, the continuous image signals in the multiple lining images are converted into discrete digital signals. Specifically, the continuous image signals are spatially discretized, and the brightness values ​​of the images are simultaneously discretized into integers.

[0018] Multiple lining images are subjected to two-dimensional discrete Fourier transform (2D DFT) to convert the continuous image signals in the lining images from the spatial domain to the frequency domain. After the transformation, each element in the lining image contains frequency information and direction information.

[0019] Based on the frequency and direction information contained in each element, the brightness information of different frequencies and directions in the lining image is obtained. The high-frequency components correspond to the details and texture features of the image, while the low-frequency components correspond to the structure and brightness features of the image.

[0020] Preferably, both the tunnel lining scanning intelligent recognition model and the three-dimensional image model of apparent defects in the surrounding rock are constructed based on deep learning convolutional neural networks (CNNs).

[0021] Preferably, after obtaining the strength grade of the tunnel surrounding rock, drilling is performed on the tunnel surrounding rock, point load strength tests are conducted on the rock obtained from the drilling, the strength grade supplement of the tunnel surrounding rock is obtained, and the strength grade of the tunnel surrounding rock is corrected according to the strength grade supplement of the tunnel surrounding rock.

[0022] This invention also provides an information acquisition system for the surrounding rock and lining structure of operating tunnels, comprising:

[0023] A binocular vision scanning module is used to acquire three-dimensional point cloud data of the lining of the operating tunnel;

[0024] The lining defect identification module is used to acquire multiple lining images of the lining at different angles and distances throughout the entire range of the operating tunnel lining; extract local brightness information from the multiple lining images, and decompose the local brightness information into frequency information and direction information; and obtain the lining appearance defect feature data of the operating tunnel lining, including crack width feature quantity and direction feature quantity, based on the frequency information and direction information.

[0025] The image recognition module is used to generate a smart recognition model of tunnel lining based on a real-world model of the tunnel lining, using 3D point cloud data and surface defect feature data.

[0026] The surrounding rock strength discrimination module is used to acquire multi-dimensional data of the internal information of the surrounding rock behind the tunnel lining, identify the surrounding rock strength based on the multi-dimensional data, and generate a three-dimensional image model of the apparent defects of the surrounding rock.

[0027] The information discrimination module is used to input the lining image of the tunnel lining to be tested into the intelligent recognition model of tunnel lining scanning to obtain the apparent defect blocks of the tunnel lining; input the multi-dimensional data of the internal information of the surrounding rock behind the tunnel lining to be tested into the three-dimensional image model of the apparent defects of the surrounding rock to obtain the strength level of the surrounding rock; and identify the defect information of the tunnel lining and surrounding rock structure based on the apparent defect blocks of the tunnel lining and the strength level of the surrounding rock.

[0028] This invention also provides an electronic device, including a memory and a processor;

[0029] The memory is used to store computer programs;

[0030] When the processor executes the computer program stored in the memory, it implements the steps of the method for acquiring information on the surrounding rock and lining structure of an operating tunnel as described above.

[0031] This invention also provides a computer storage medium for storing a computer program, which, when executed by a processor, implements the steps of the method for acquiring information on the surrounding rock and lining structure of an operating tunnel as described above.

[0032] This invention provides a method and system for acquiring information on the surrounding rock and lining structure of operating tunnels. Compared with the prior art, its advantages are as follows:

[0033] This invention acquires images of the tunnel lining structure from different angles and distances across its entire range, extracts characteristic quantities reflecting crack width and direction from these images, and uses these characteristic quantities, tunnel lining point cloud data, and a 3D model of the lining to train an image recognition model, thus obtaining a smart recognition model for tunnel lining scanning. Furthermore, it uses a drone to detect multi-dimensional information about the tunnel's surrounding rock, labels the surrounding rock grade based on the multi-dimensional data, and trains a 3D image recognition model based on the surrounding rock grade to obtain a 3D image model of apparent defects in the surrounding rock. Finally, by combining the smart recognition model for tunnel lining scanning and the 3D image model of apparent defects in the surrounding rock, it accurately identifies defects and internal cracks in the tunnel's surrounding rock and lining structure. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the overall process of a method for obtaining information on the surrounding rock and lining structure of an operating tunnel, provided by an embodiment of the present invention.

[0035] Figure 2 A schematic diagram illustrating the system principle of a method and system for acquiring information on the surrounding rock and lining structure of an operating tunnel, provided in an embodiment of the present invention;

[0036] Figure 3 A schematic diagram illustrating the system principle breakdown of a method and system for acquiring information on the surrounding rock and lining structure of an operating tunnel, provided in an embodiment of the present invention.

[0037] Figure 4 This is a schematic diagram of an electronic device framework for an information acquisition method and system for the surrounding rock and lining structure of an operating tunnel, provided in an embodiment of the present invention. Detailed Implementation

[0038] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0039] See Figure 2 This invention provides a method and system for acquiring information on the surrounding rock and lining structure of an operating tunnel, including an unmanned aerial vehicle platform 100, a lightweight binocular vision scanning module 200, an oblique photography module 300, an image recognition module 400, a surrounding rock information acquisition module 500, an intelligent diagnosis module 600, and a three-dimensional visualization module 700.

[0040] The UAV platform 100 is used to control at least one UAV to perform flight missions and attitude control actions in the target tunnel according to a preset planned route, so as to photograph the tunnel lining in the target tunnel. The UAV platform 100 can control at least one UAV to perform flight missions and attitude control actions in the target tunnel according to a planned route, so as to photograph the tunnel lining in the target tunnel. In addition, the UAV platform 100 can control multiple UAVs to perform flight missions and attitude control actions in the target tunnel according to a planned route, thereby improving the accuracy of photographing the tunnel lining and thus improving the level of intelligence in tunnel lining scanning and surrounding rock information acquisition.

[0041] like Figure 3 As shown, the UAV platform 100 includes a data storage and transmission module 101 and a UAV flight control module 102. The data storage and transmission module 101 is used to store point cloud data and image data. The data storage and transmission module 101 of the UAV platform 100 can store high-definition point cloud data and image data collected and recorded in real time by the lightweight binocular vision scanning module 200 and the oblique photography module 300. The storage capacity meets the data acquisition requirements of a complete tunnel lining scanning operation, and at the same time meets the requirements for real-time lossless transmission of data under tunnel conditions, ensuring that the main control room outside the tunnel can control the UAV platform 100. 0. Synchronization of control; the UAV flight control module 102 is used to control at least one UAV to perform flight missions and attitude control actions; the UAV flight control module 102 set in the UAV platform 100 can control at least one UAV to perform flight missions and attitude control actions. The UAV flight control module 102 can share information from the global navigation satellite module 201, the inertial navigation module 202 and the binocular vision scanning module 203. At the same time, based on the UAV platform 100's own autonomous route planning function and autonomous navigation function, the autonomous flight and attitude control of the UAV platform 100 under tunnel conditions can be realized.

[0042] In addition, the UAV flight control module 102 can transmit data information to the main control room in real time through the data storage and transmission module 101 based on the inertial navigation module 202, the binocular vision scanning module 203 and the data acquisition module 301, so as to realize the flight of the UAV under manual remote control and realize the UAV platform 100 to perform fine scanning and shooting of the tunnel lining.

[0043] The lightweight binocular vision scanning module 200 is used to collect point cloud data of tunnel lining when at least one UAV is performing flight missions and attitude control actions, so as to improve the feasibility of intelligent tunnel lining scanning; it includes a global navigation satellite module 201, an inertial navigation module 202 and a binocular vision scanning module 203.

[0044] The Global Navigation Satellite Module 201 and the Inertial Navigation Module 202 are used to calculate the real-time pose of the UAV platform 100. Based on the UAV platform 100's autonomous route planning and navigation functions, they enable autonomous flight and pose control of the UAV platform 100 under tunnel conditions, improving the intelligence level of tunnel lining scanning. The binocular vision scanning module 203 is used to acquire point cloud data of the tunnel lining. It can quickly acquire three-dimensional coordinate point cloud data of the tunnel lining and rapidly obtain a high-precision, high-resolution digital surface model of the tunnel lining, thereby improving the accuracy of tunnel data acquisition.

[0045] The oblique photography module 300 is used to acquire image data of the tunnel lining and generate a real-world 3D model of the lining, thereby obtaining more complete and accurate surface information of the tunnel lining and improving the accuracy of tunnel lining scanning and recognition; it includes a data acquisition module 301, an infrared imaging module 302 and a 3D imaging module 303.

[0046] The data acquisition module 301 includes multiple cameras to simultaneously acquire image data of the tunnel lining from multiple different angles. Specifically, it acquires tunnel lining image data from different angles such as vertical and inclined, improving the completeness and accuracy of obtaining information about the tunnel lining surface. In addition, the data acquisition module 301 can adjust the distance between the drone and the tunnel sidewall, as well as the up, down, left, and right orientation, according to the actual situation of the tunnel lining, through the drone flight control module 102. This, combined with multiple high-definition cameras, allows for the acquisition of omnidirectional image information of the tunnel lining, avoiding the obstruction of joints and cracks in the recessed areas by the protruding parts of the lining, thereby improving the accuracy and reliability of obtaining information about the actual situation of the tunnel lining.

[0047] The infrared imaging module 302 is used to perform infrared imaging of tunnel lining. It is equipped with multiple infrared cameras and can simultaneously collect infrared image data of tunnel lining from different angles such as vertical and tilt, thereby improving the completeness and accuracy of obtaining information about the surface of tunnel lining.

[0048] The 3D imaging module 303 can automatically identify tunnel lining images taken from different angles and orientations by the data acquisition module 301 and the infrared imaging module 302, and use 3D image modeling software to stitch and optimize the tunnel lining images to generate a real-world 3D model of the lining that is consistent with the actual tunnel lining state and has clear concave and convex surface elements, thereby improving the accuracy of tunnel lining information.

[0049] The image recognition module 400 is used to obtain a smart recognition model of the tunnel lining based on the actual 3D model of the lining, thereby improving the intelligence level of tunnel lining scanning and improving the accuracy of lining appearance information recognition; it includes a sample calibration module 401 and a lining defect recognition module 402.

[0050] The sample calibration module 401 is used to annotate the surface defects of the tunnel lining. Based on the generated three-dimensional coordinate point cloud data and the real-scene three-dimensional model of the tunnel lining, professional geologists manually annotate the orientation, density, width, and extension of cracks, seepage, spalling, and fissures in the lining, as well as the water outlet points and water volume in the lining. The sample calibration module 401 can dynamically update the sample library. When the surface information of the tunnel lining changes, the sample library is expanded in a timely manner by professional personnel through manual annotation, and the weight of new samples is increased to train and test the model. This avoids a sudden drop in the recognition accuracy of the tunnel lining scanning intelligent recognition model due to sudden changes in the surface information of the lining, thereby improving the accuracy and reliability of the tunnel lining scanning intelligent recognition model.

[0051] The tunnel lining defect recognition module 402 applies Weiblet transform to the photographic image, decomposing the local brightness information of the photographic image into frequency and direction components. It detects feature quantities reflecting crack width and direction from the captured image. Based on the apparent defects of the lining, point cloud data, and the actual 3D model of the lining, it trains an image recognition model to obtain a smart recognition model for tunnel lining scanning. It can also train an image recognition model based on the corresponding images of the apparent defects of the lining, point cloud data, and 3D model labeled by the sample calibration module 401. After the model training and testing are completed, actual tunnel lining photographic images are used as verification samples to verify the recognition accuracy and obtain a smart recognition model for tunnel lining scanning.

[0052] The decomposition of local brightness information is as follows: after extracting local brightness information from multiple lining images, the continuous image signals in the multiple lining images are converted into discrete digital signals. Specifically, the continuous image signals are spatially discretized, and the brightness values ​​of the images are simultaneously discretized into integers. The multiple lining images are subjected to a two-dimensional discrete Fourier transform (2D DFT) to transform the continuous image signals in the lining images from the spatial domain to the frequency domain. After the transformation, each element in the lining image contains frequency information and direction information.

[0053] By utilizing the frequency and orientation information contained in each element, brightness information of different frequencies and orientations in the lining image is obtained. The high-frequency components correspond to the details and texture features of the image, while the low-frequency components correspond to the structure and brightness features of the image.

[0054] The surrounding rock information acquisition module 500 is used to detect the internal information of the tunnel surrounding rock when at least one UAV performs the flight mission and attitude control actions, thereby improving the accuracy of strength identification of the tunnel surrounding rock under multiple disturbances and improving the safety of tunnel construction; it includes an airborne ground-penetrating radar module 501, a feature input module 502, a surrounding rock strength identification module 503, and a surrounding rock strength correction module 504.

[0055] The airborne ground-penetrating radar module 501 is used to transmit radar waves and receive signals to detect the internal information of the surrounding rock of the operating tunnel. At the same time, it selects geological parameters that meet preset conditions to generate input parameters for the surrounding rock strength identification module 503.

[0056] The feature input module 502 can select geological parameters that meet certain conditions as input parameters for the surrounding rock strength identification module 503 in the following steps, based on the surrounding rock classification method used in the tunnel and the results of identifying apparent defects in the lining by the tunnel airborne ground-penetrating radar. This reduces the cost of manual operation and improves the efficiency of the measurement data of the surrounding rock scanning results.

[0057] The surrounding rock strength identification module 503 is used to train and test the neural network model based on the input parameters of the surrounding rock strength identification module 503, combined with a preset deep learning neural network algorithm and the actual surrounding rock strength annotation results, to obtain a three-dimensional image model of the apparent defects of the surrounding rock. It can also train and test the neural network model based on the input parameters determined by the feature input module 502, combined with the deep learning neural network algorithm and the actual surrounding rock grade annotation results, thereby obtaining a three-dimensional image model of the apparent defects of the surrounding rock. This improves work efficiency and construction safety, and is convenient to operate with simple steps, making it easy for workers to learn and master.

[0058] The surrounding rock strength correction module 504 is used to perform point load strength tests on the drilled rock based on the on-site drilling conditions to obtain supplementary parameters for surrounding rock strength. This is used to correct the identification results of the three-dimensional image model of apparent defects in the surrounding rock. By performing point load strength tests on the drilled rock based on the on-site drilling conditions to supplement the surrounding rock strength parameters and correct the identification results of the three-dimensional image model of apparent defects in the surrounding rock, the module effectively improves the accuracy of surrounding rock strength identification and enhances the safety of tunnel construction.

[0059] This invention can check whether the data storage and transmission module and the flight control module of the UAV platform are working properly, check whether the lightweight binocular vision scanning module and the oblique photography module on the UAV platform are working properly, set the flight path of the UAV platform, rely on the UAV platform for autonomous navigation flight, combine the real-time synchronous images captured and transmitted by the data acquisition module and the data storage and transmission module, and use the manual remote control assisted flight method to ensure that the UAV safely reaches the tunnel.

[0060] Next, the pose information of the UAV platform can be determined based on the global navigation satellite module and the inertial navigation module. The three-dimensional coordinate point cloud data of the lining can be quickly acquired through the binocular vision scanning module to generate a high-precision, high-resolution digital surface model of the tunnel lining.

[0061] Secondly, based on the actual situation of the tunnel lining, the distance between the drone and the tunnel lining, as well as its orientation (up, down, left, right) can be adjusted through the drone flight control module. Multiple high-definition cameras can be used to acquire all-around image information of the lining and generate a real-world 3D model of the lining.

[0062] Furthermore, the digital surface model and the 3D model of the lining can be input into the tunnel lining scanning intelligent recognition model to obtain the tunnel lining defect recognition results; at the same time, the tunnel surrounding rock strength level can be obtained based on the 3D image model of the apparent defects of the surrounding rock, and combined with the surrounding rock strength parameters obtained by the correction module, the tunnel surrounding rock strength identification results can be obtained.

[0063] This invention utilizes a drone platform to control a drone to perform flight missions and attitude control maneuvers within a target tunnel along a planned route. A lightweight binocular vision scanning module collects point cloud data of the tunnel lining, while an oblique photography module collects impact data of the tunnel lining, generating a realistic 3D model of the lining. An image recognition module then obtains a smart recognition model of the tunnel lining scan. Furthermore, a surrounding rock information acquisition module uses airborne ground-penetrating radar to detect internal information of the surrounding rock and, based on the processing results, obtains surface information of the surrounding rock, resulting in a 3D image model of surface defects. This significantly reduces manual operation costs, improves the accuracy of surrounding rock strength assessment, and enhances the safety of tunnel construction. Therefore, it solves the problems of related technologies that rely on manual identification of tunnel lining defects and internal surrounding rock information, which increases labor costs, makes it difficult to ensure the safety of workers, reduces the accuracy of surrounding rock strength assessment, and lowers the safety of tunnel reconstruction and expansion construction.

[0064] like Figure 1 As shown, the specific method includes the following steps:

[0065] Step 1: Perform flight missions and attitude control maneuvers in the target tunnel according to the preset planned route in order to photograph the tunnel lining in the target tunnel.

[0066] Step 2: Collect point cloud data of the tunnel lining.

[0067] Step 3: Collect the impact data of the tunnel lining, generate a real-world 3D model of the lining, and obtain the intelligent recognition model of the tunnel lining scanning.

[0068] Step 4: Obtain multi-dimensional data on the internal information of the tunnel surrounding rock, intelligently identify the strength of the surrounding rock based on the processing results, and generate a three-dimensional image model of the apparent defects of the surrounding rock.

[0069] Step 5: Perform intelligent diagnosis, identification, and summary of apparent defects in the lining and surrounding rock.

[0070] Step Six: Combine the intelligent identification model of tunnel lining scanning with the three-dimensional image model of apparent defects in the surrounding rock to realize a three-dimensional visualized transparent geological model that includes both the appearance information of the tunnel lining and the information of the internal surrounding rock.

[0071] This invention controls a drone via a drone platform to perform flight missions and attitude control actions in a target tunnel according to a planned route. A lightweight binocular vision scanning module collects point cloud data of the tunnel lining, and an oblique photography module photographs the tunnel lining within the target tunnel. The collected image data of the tunnel lining generates a realistic 3D model of the lining, resulting in a smart recognition model for tunnel lining scanning based on this model. A surrounding rock information acquisition module performs 3D detection and data collection on the surrounding rock behind the lining to generate a 3D image model of apparent defects in the surrounding rock. Combining the smart recognition model for tunnel lining scanning and the 3D image model of apparent defects in the surrounding rock creates a 3D visualized transparent geological model that integrates lining surface information and internal surrounding rock information. This significantly reduces manual operation costs, improves lining accuracy, and enhances tunnel construction safety. It addresses the problems of manually identifying tunnel lining scanning and surrounding rock grades, which increases manual operation costs, makes it difficult to ensure the safety of workers, reduces the accuracy of surrounding rock strength identification, and lowers tunnel construction safety.

[0072] This invention utilizes image recognition technology and deep learning neural network algorithms to create a three-dimensional image model of apparent defects in surrounding rock, enabling standardized identification of tunnel surrounding rock grades. This saves significant manpower, improves work efficiency, ensures construction safety, and is easy to operate with simple steps, making it convenient for workers to learn and master. It also reduces on-site preparation and debugging time, saving time and facilitating widespread adoption.

[0073] like Figure 4 The diagram shown is a structural schematic of an electronic device provided in an embodiment of the present invention, comprising:

[0074] Memory 1, processor 2, and computer program stored on memory 1 and capable of running on processor 2.

[0075] When processor 2 executes the program, it implements the steps of the method for obtaining information on the surrounding rock of the lining of an operating tunnel provided in the above embodiments.

[0076] Electronic devices also include:

[0077] Communication interface 3 is used for communication between memory 1 and processor 2.

[0078] Memory 1 is used to store computer programs that can run on processor 2.

[0079] Memory 1 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0080] If memory 1, processor 2, and communication interface 3 are implemented independently, then communication interface 3, memory 1, and processor 2 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0081] If memory 1, processor 2, and communication interface 3 are integrated on a single chip, then memory 1, processor 2, and communication interface 3 can communicate with each other through their internal interfaces.

[0082] Processor 2 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0083] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for acquiring information on the surrounding rock and lining structure of an operating tunnel, characterized in that, Includes the following steps: Acquire 3D point cloud data of the lining of the operating tunnel; Acquire multiple images of the tunnel lining from different angles and distances across the entire range of the operational tunnel lining; Local brightness information is extracted from multiple lining images and decomposed into frequency and orientation information. Based on the frequency and orientation information, surface defect feature data of the operating tunnel lining, including crack width and orientation features, are obtained. Based on 3D point cloud data and apparent defect feature data, a smart recognition model for tunnel lining scanning based on a real-world model of tunnel lining is generated. Multidimensional data of the internal information of the surrounding rock behind the tunnel lining is obtained, the strength of the surrounding rock is identified based on the multidimensional data, and a three-dimensional image model of the apparent defects of the surrounding rock is generated. The lining image of the tunnel to be tested is input into the tunnel lining scanning intelligent recognition model to obtain the tunnel lining apparent defect blocks; the multi-dimensional data of the internal information of the surrounding rock behind the tunnel lining is input into the three-dimensional image model of the apparent defects of the surrounding rock to obtain the strength level of the tunnel surrounding rock. And based on the apparent defect blocks of the tunnel lining and the strength grade of the surrounding rock, identify the defect information of the lining and surrounding rock structure of the operating tunnel to be tested; The process of decomposing local brightness information into frequency information and direction information includes: After extracting local brightness information from multiple lining images, the continuous image signals in the multiple lining images are converted into discrete digital signals. Specifically, the continuous image signals are spatially discretized, and the brightness values ​​of the images are simultaneously discretized into integers. Multiple lining images are subjected to two-dimensional discrete Fourier transform (2D DFT) to convert the continuous image signals in the lining images from the spatial domain to the frequency domain. After the transformation, each element in the lining image contains frequency information and direction information. Based on the frequency and direction information contained in each element, the brightness information of different frequencies and directions in the lining image is obtained. The high-frequency components correspond to the details and texture features of the image, while the low-frequency components correspond to the structure and brightness features of the image. The tunnel lining scanning intelligent recognition model and the three-dimensional image model of apparent defects in the surrounding rock are both constructed based on deep learning convolutional neural networks (CNNs).

2. The method for acquiring information on the surrounding rock and lining structure of an operating tunnel according to claim 1, characterized in that, The acquisition of the three-dimensional point cloud data of the operating tunnel lining includes: Multiple drones were used to conduct a full-range scan of the operating tunnel, and at the same time, the lining structure of the operating tunnel was photographed from multiple angles, including vertical, inclined and horizontal angles, to obtain multiple lining images of the operating tunnel lining structure. Multiple lining images were stitched together using 3D image modeling software and combined with omnidirectional scanning data to obtain a real-world 3D model of the operating tunnel lining structure, and the 3D point cloud data of the operating tunnel lining in the real-world 3D model was identified.

3. The method for acquiring information on the surrounding rock and lining structure of an operating tunnel according to claim 2, characterized in that, After acquiring the real-world 3D model and 3D point cloud data of the lining structure, manual annotation is used to mark the lining cracks, seepage, spalling, crack orientation, density, width, extension, water outlet points, and water outflow of the tunnel lining structure.

4. The method for acquiring information on the surrounding rock and lining structure of an operating tunnel according to claim 1, characterized in that, After obtaining the strength grade of the tunnel surrounding rock, boreholes are drilled into the tunnel surrounding rock, and point load strength tests are performed on the rock obtained from the boreholes to obtain the strength grade supplement of the tunnel surrounding rock. The strength grade of the tunnel surrounding rock is then corrected based on the strength grade supplement of the tunnel surrounding rock.

5. An information acquisition system for the surrounding rock and lining structure of an operating tunnel, characterized in that, include: A binocular vision scanning module is used to acquire three-dimensional point cloud data of the lining of the operating tunnel; The lining defect identification module is used to acquire multiple lining images of the lining at different angles and distances throughout the entire range of the operating tunnel lining; extract local brightness information from the multiple lining images, and decompose the local brightness information into frequency information and direction information; and obtain the lining appearance defect feature data of the operating tunnel lining, including crack width feature quantity and direction feature quantity, based on the frequency information and direction information. The image recognition module is used to generate a smart recognition model of tunnel lining based on a real-world model of the tunnel lining, using 3D point cloud data and surface defect feature data. The surrounding rock strength discrimination module is used to acquire multi-dimensional data of the internal information of the surrounding rock behind the tunnel lining, identify the surrounding rock strength based on the multi-dimensional data, and generate a three-dimensional image model of the apparent defects of the surrounding rock. The information discrimination module is used to input the lining image of the tunnel lining to be tested into the tunnel lining scanning intelligent recognition model to obtain the tunnel lining apparent defect blocks; and to input the multi-dimensional data of the internal information of the surrounding rock behind the tunnel lining to be tested into the three-dimensional image model of the apparent defects of the surrounding rock to obtain the strength level of the tunnel surrounding rock. And based on the apparent defect blocks of the tunnel lining and the strength grade of the surrounding rock, identify the defect information of the lining and surrounding rock structure of the operating tunnel to be tested; The process of decomposing local brightness information into frequency information and direction information includes: After extracting local brightness information from multiple lining images, the continuous image signals in the multiple lining images are converted into discrete digital signals. Specifically, the continuous image signals are spatially discretized, and the brightness values ​​of the images are simultaneously discretized into integers. Multiple lining images are subjected to two-dimensional discrete Fourier transform (2D DFT) to convert the continuous image signals in the lining images from the spatial domain to the frequency domain. After the transformation, each element in the lining image contains frequency information and direction information. Based on the frequency and direction information contained in each element, the brightness information of different frequencies and directions in the lining image is obtained. The high-frequency components correspond to the details and texture features of the image, while the low-frequency components correspond to the structure and brightness features of the image. The tunnel lining scanning intelligent recognition model and the three-dimensional image model of apparent defects in the surrounding rock are both constructed based on deep learning convolutional neural networks (CNNs).

6. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer programs; When the processor executes the computer program stored in the memory, it implements the steps of the method for acquiring information on the surrounding rock and lining structure of an operating tunnel as described in any one of claims 1 to 4.

7. A computer storage medium, characterized in that, Used to store a computer program, which, when executed by a processor, implements the steps of a method for acquiring information on the surrounding rock and lining structure of an operating tunnel as described in any one of claims 1 to 4.

Images