Soft rock tunnel deformation monitoring method

Abstract

The invention discloses a soft rock tunnel deformation monitoring method, and relates to the technical field of tunnel construction.A target image set of a soft rock tunnel is obtained by carrying out photography and video recording on the soft rock tunnel subjected to excavation construction, and then a three-dimensional digital reference model of the soft rock tunnel is established according to the target image set; continuously carrying out photography and video recording operation on the soft rock tunnel to obtain a current image set corresponding to each acquisition moment, importing the current image set into the unit digital reference model, and comparing and analyzing the current joint trend with the joint trend model to obtain the current displacement of the geologic body; and finally, the deformation monitoring result of the soft rock tunnel is obtained according to the current displacement, so that the deformation condition of the geologic body of the soft rock tunnel can be identified in a photographing and video recording mode, a sensor does not need to be installed in the geologic body of the excavated and formed soft rock tunnel, and the monitoring cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of soft rock tunnel construction, and in particular to a soft rock tunnel deformation monitoring method. Background Art

[0002] During construction, soft rock tunnels, due to their inherent softness, deform after excavation. This deformation can cause the deformed rock to intrude into the soft rock tunnel, reducing the tunnel's clearance and rendering it unusable.

[0003] In existing technology, deformation monitoring of soft rock tunnels typically involves installing multiple sensors within the tunnel. These sensors collect real-time data from the tunnel face, ultimately determining the deformation of the soft rock tunnel based on the collected real-time data. While installing sensors on the surrounding rock within a soft rock tunnel can monitor deformation, the need to install multiple sensors increases monitoring costs in practice. Summary of the Invention

[0004] The main purpose of the present invention is to propose a deformation monitoring method for soft rock tunnels, aiming to solve the technical problem of the existing technology that in actual operation, the monitoring cost increases due to the need to install multiple sensors.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for monitoring deformation of a soft rock tunnel, comprising the following steps:

[0006] photographing and videotaping the soft rock tunnel after excavation to obtain a target image set of the soft rock tunnel;

[0007] Establishing a three-dimensional digital benchmark model of the soft rock tunnel based on the target image set; wherein the three-dimensional digital benchmark model includes a joint strike model of the geological body of the soft rock tunnel;

[0008] Continuously photographing and videotaping the soft rock tunnel to obtain a current image set corresponding to each acquisition moment; wherein each current image in the current image set records the current joint trend of the geological body;

[0009] Importing the current image set into the three-dimensional digital benchmark model and comparing and analyzing the current joint trend with the joint trend model to obtain the current displacement of the geological body;

[0010] A deformation monitoring result of the soft rock tunnel is obtained according to the current displacement.

[0011] In one embodiment, the step of obtaining the deformation monitoring result of the soft rock tunnel according to the current displacement includes:

[0012] generating a current deformation distribution data set of the soft rock tunnel according to the current displacement;

[0013] The deformation monitoring result of the soft rock tunnel is outputted according to the current deformation distribution data set.

[0014] In one embodiment, the step of generating a current deformation distribution dataset of the soft rock tunnel according to the current displacement includes:

[0015] A tunnel cross-section distribution diagram and a deformation trend curve of the soft rock tunnel are generated according to the current displacement, and a current deformation distribution data set is obtained.

[0016] In one embodiment, the step of outputting the deformation monitoring result of the soft rock tunnel according to the current deformation distribution dataset includes:

[0017] outputting a current quantitative analysis result of the soft rock tunnel according to the current deformation distribution data set;

[0018] The deformation monitoring result of the soft rock tunnel is outputted according to the current quantitative analysis result.

[0019] In one embodiment, the step of outputting the current quantitative analysis result of the soft rock tunnel based on the current deformation distribution dataset includes:

[0020] According to the current deformation distribution data set, the current deformation rate, current deformation direction and current dangerous area of ​​the soft rock tunnel are output to obtain the current quantitative analysis result.

[0021] In one embodiment, the three-dimensional digital reference model and the current image set both include spatial coordinate information;

[0022] The step of importing the current image set into the three-dimensional digital benchmark model and comparing and analyzing the current joint trend with the joint trend model to obtain the current displacement of the geological body includes:

[0023] Importing the current image set into the three-dimensional digital reference model, and making the three-dimensional digital reference model correspond to the spatial coordinate information of the current image set one by one;

[0024] The current joint trend is compared and analyzed with the joint trend model to obtain a displacement change data set of the inner wall of the soft rock tunnel and obtain the current displacement of the geological body.

[0025] In one embodiment, the step of comparing and analyzing the current joint trend with the joint trend model to obtain a displacement change dataset of the inner wall of the soft rock tunnel and obtain the current displacement of the geological body includes:

[0026] Comparing and analyzing the current joint trend with the joint trend model, and using pixel-level change detection technology to identify and obtain a current displacement change value of the inner wall of the soft rock tunnel;

[0027] According to the current displacement change value and the collected three-dimensional displacement, a displacement change data set of the inner wall of the soft rock tunnel is obtained to obtain the current displacement of the geological body.

[0028] In one embodiment, the step of obtaining a displacement change dataset of the inner wall of the soft rock tunnel based on the current displacement change value and the collected three-dimensional displacement to obtain the current displacement of the geological body includes:

[0029] According to the current displacement change value and the three-dimensional displacement of the geological body calculated by using a stereoscopic vision measurement method, a displacement change data set of the inner wall of the soft rock tunnel is obtained to obtain the current displacement of the geological body.

[0030] In one embodiment, the step of photographing and videotaping the soft rock tunnel after excavation to obtain a target image set of the soft rock tunnel includes:

[0031] The soft rock tunnel that has completed excavation construction is photographed and videotaped according to a preset shooting parameter set to obtain a target image set of the soft rock tunnel; wherein the preset shooting parameter set includes a preset shooting path and preset shooting parameters.

[0032] In one embodiment, the step of continuously photographing and videotaping the soft rock tunnel to obtain a current image set corresponding to each acquisition moment includes:

[0033] The soft rock tunnel is continuously photographed and filmed according to the preset shooting parameter set to obtain an image sequence of the inner wall of the soft rock tunnel corresponding to each acquisition moment, thereby obtaining the current image set.

[0034] When the technical solution of the present invention is used, a target image set of the soft rock tunnel is obtained by photographing and videotaping the soft rock tunnel after excavation construction, and then a three-dimensional digital benchmark model of the soft rock tunnel is established based on the target image set. Subsequently, the soft rock tunnel is photographed and videotaped continuously to obtain a current image set corresponding to each acquisition moment, and the current image set is imported into the unit digital benchmark model and the current joint trend is compared and analyzed with the joint trend model to obtain the current displacement of the geological body. Finally, the deformation monitoring result of the soft rock tunnel is obtained based on the current displacement. The present invention can identify the deformation of the geological body of the soft rock tunnel by photographing and videotaping, without the need to install sensors in the geological body of the excavated soft rock tunnel, thereby reducing the monitoring cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying any creative work.

[0036] Figure 1 Flowchart of the soft rock tunnel deformation monitoring method provided by the present invention;

[0037] Figure 2 for Figure 1 Schematic diagram of the process of step S500 in the example;

[0038] Figure 3 for Figure 2 Schematic diagram of the process of step S520 in the example;

[0039] Figure 4 for Figure 1 Schematic diagram of the process of step S400 in the example;

[0040] Figure 5 for Figure 4 Schematic diagram of the process of step S420 in the example.

[0041] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0043] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture. If the specific posture changes, the directional indications will also change accordingly.

[0044] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features limited to "first" and "second" may explicitly or implicitly include at least one of such features. In addition, if "and / or" or "and / or" appears in the full text, its meaning includes three parallel solutions. Taking "A and / or B" as an example, it includes solution A, solution B, or solutions that satisfy both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of ordinary technicians in this field to implement. When the combination of technical solutions is mutually contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0045] During construction, soft rock tunnels, due to their inherent softness, deform after excavation. This deformation can cause the deformed rock to intrude into the soft rock tunnel, reducing the tunnel's clearance and rendering it unusable.

[0046] In the prior art, when monitoring the deformation of a soft rock tunnel, multiple sensors are usually installed in the soft rock tunnel. The installed sensors are used to collect real-time data on the tunnel face, and the deformation results of the soft rock tunnel are finally determined based on the collected real-time data.

[0047] The applicant has found that installing sensors on the surrounding rock in a soft rock tunnel can monitor the deformation of the soft rock tunnel. However, in actual operation, the need to install multiple sensors increases the monitoring cost.

[0048] Based on this, the present invention proposes a soft rock tunnel deformation monitoring method, whose core inventive concept is: at the initial stage of tunnel excavation, use mobile high-precision photographic equipment to take panoramic photos of the tunnel inner wall; based on multi-view image reconstruction technology, construct a three-dimensional digital benchmark model of the tunnel; select natural feature points on the tunnel inner wall as a reference benchmark for deformation monitoring; regularly use the same photographic equipment to move along the tunnel axis for shooting; use the same shooting parameters and paths to ensure data consistency; obtain a sequence of tunnel inner wall images in the current state; use a deep learning image registration algorithm to align the current image with the benchmark model; identify the displacement changes of the tunnel inner wall through pixel-level change detection technology; use stereo vision measurement technology to calculate the three-dimensional displacement of the deformation; convert the detected pixel changes into actual physical displacement; generate a tunnel section deformation distribution map and deformation trend curve; output quantitative analysis results such as deformation rate, deformation direction, and danger zone.

[0049] It can be seen that the technical solution of the present invention adopts a sensorless design and is completely based on visual technology. There is no need to install any physical sensors in the tunnel, which can avoid the cost of sensor maintenance and replacement, and will not be affected by signal transmission and electromagnetic interference, thereby improving monitoring efficiency.

[0050] The present invention provides a soft rock tunnel deformation monitoring method.

[0051] See also Figures 1 to 5 For ease of understanding, the soft rock tunnel deformation monitoring method includes the following steps:

[0052] S100: photographing and videotaping the soft rock tunnel after excavation to obtain a target image set of the soft rock tunnel.

[0053] Specifically, photography and videography operations refer to the collection of images of the inner wall of the tunnel through a camera or video equipment. This can be achieved by using a fixed camera array or a mobile shooting device to obtain complete image data of the tunnel surface.

[0054] S200 . Establishing a three-dimensional digital benchmark model of the soft rock tunnel based on the target image set; wherein the three-dimensional digital benchmark model includes a joint strike model of the geological body of the soft rock tunnel.

[0055] Specifically, the three-dimensional digital benchmark model refers to a tunnel structure model generated by a three-dimensional reconstruction algorithm based on multi-angle images. It can be implemented by combining photogrammetry technology with point cloud data processing. The joint strike model it contains is constructed by extracting rock fracture characteristics through an image recognition algorithm.

[0056] S300: Continuously photograph and videotape the soft rock tunnel to obtain a current image set corresponding to each acquisition moment; wherein each current image in the current image set records the current joint trend of the geological body.

[0057] Specifically, the current image set refers to a sequence of periodically acquired tunnel wall images. Preset capture paths and parameters ensure alignment with the baseline model acquisition conditions, ensuring image spatial coordinates are aligned. Joint strike comparison analysis involves spatially matching the orientation of cracks extracted from the current image with corresponding features in the baseline model. Displacement deviations are calculated using a feature point matching algorithm.

[0058] S400 , importing the current image set into the three-dimensional digital benchmark model and comparing and analyzing the current joint trend with the joint trend model to obtain the current displacement of the geological body.

[0059] S500: Obtain deformation monitoring results of the soft rock tunnel according to the current displacement.

[0060] In this embodiment, full-section images are first acquired immediately after tunnel excavation, and a baseline model incorporating joint orientation features is generated through 3D reconstruction. Subsequent monitoring phases involve repeated image capture at a fixed interval, with each acquired image imported into the baseline model through spatial coordinate registration. The system automatically identifies joint orientations in the current image and compares them pixel-wise with the corresponding fracture features in the baseline model. Rock mass deformation data is then determined by calculating the displacement of these feature points. Deformation monitoring results are dynamically generated based on the spatial distribution and trend of these displacements, without the need for manual intervention or hardware maintenance.

[0061] This solution achieves continuous monitoring through automated image processing and model comparison, which improves the frequency and timeliness of data collection. In addition, the use of the rock mass's own joint trend as a monitoring benchmark has higher environmental adaptability and characteristic stability than manual marking points. This also effectively reduces the hardware investment and operation and maintenance costs of soft rock tunnel deformation monitoring, and realizes contactless, fully automatic continuous monitoring. By using the inherent characteristics of the geological body for displacement calculation, the impact of sensor layout on construction progress is avoided, and the spatial resolution and reliability of monitoring data are improved. The combination of image processing and three-dimensional model comparison technology enables deformation analysis to be carried out on specific geological structural units, providing a more detailed decision-making basis for tunnel stability assessment.

[0062] In this embodiment, a target image set of the soft rock tunnel is obtained by photographing and videotaping the soft rock tunnel after excavation. Then, a three-dimensional digital benchmark model of the soft rock tunnel is established based on the target image set. Subsequently, the soft rock tunnel is photographed and videotaped continuously to obtain a current image set corresponding to each acquisition moment. The current image set is imported into the unit digital benchmark model and the current joint trend is compared and analyzed with the joint trend model to obtain the current displacement of the geological body. Finally, the deformation monitoring result of the soft rock tunnel is obtained based on the current displacement. In this way, the present invention can identify the deformation of the geological body of the soft rock tunnel by photographing and videotaping, without the need to install sensors in the geological body of the excavated soft rock tunnel, thereby reducing the monitoring cost.

[0063] In one embodiment, step S500 includes:

[0064] S510: Generate a current deformation distribution data set of the soft rock tunnel according to the current displacement.

[0065] Specifically, the current deformation distribution dataset refers to the conversion of displacement into structured data containing spatial distribution information through image analysis technology. This can be achieved by using tunnel cross-sectional distribution maps and deformation trend curves. By mapping the displacement into the three-dimensional tunnel model to form visual data, the problem of discretization of traditional sensor data is solved.

[0066] S520: Output the deformation monitoring result of the soft rock tunnel according to the current deformation distribution data set.

[0067] Specifically, deformation monitoring results refer to the quantitative analysis conclusions based on the deformation distribution data set. Specifically, the deformation rate, deformation direction and dangerous area can be automatically output through the algorithm model, and data association analysis can be used to replace manual experience judgment to achieve automatic generation of monitoring results.

[0068] When generating the current deformation distribution dataset, the system matches the displacement data with the spatial coordinates of the 3D tunnel model to create a distribution map reflecting the degree of deformation across each tunnel section. This is combined with time series data to generate a deformation trend curve, resulting in a dataset encompassing both spatial and temporal dimensions. When outputting deformation monitoring results, key parameters are extracted based on the distribution map and trend curve. For example, deformation rate is calculated from the slope of the curve, and hazardous areas are identified through spatial differences in the distribution map. Ultimately, monitoring conclusions are output, including quantitative indicators.

[0069] In this embodiment, a structured data set is generated through image analysis, which directly associates spatial coordinates with displacement, enabling automatic data integration and visual expression without the need for additional hardware deployment. At the same time, deformation features are automatically identified through an algorithm model to improve monitoring accuracy and response speed.

[0070] This application solves the problem of high sensor deployment costs. It generates an intuitive deformation distribution dataset through image processing and data analysis, and outputs quantitative monitoring results based on data-driven output, avoiding human experience errors. It can quickly identify tunnel deformation trends and risk areas, providing a reliable basis for engineering decision-making.

[0071] In one embodiment, step S510 includes:

[0072] A tunnel cross-section distribution diagram and a deformation trend curve of the soft rock tunnel are generated according to the current displacement, and a current deformation distribution data set is obtained.

[0073] Specifically, a tunnel cross-section distribution map is a 2D projection of the displacements of sections at different locations, converted into a 3D model using point cloud data registration and projection algorithms. This map visually demonstrates the spatial deformation differences among tunnel sections. A deformation trend curve is a statistical curve constructed based on time-series displacement data. This curve can be implemented using polynomial fitting or sliding average algorithms, revealing the regularity of rock mass deformation over time.

[0074] After obtaining the current displacement, the discrete displacement data is mapped to the surface of the tunnel's three-dimensional model using a spatial interpolation algorithm to generate a continuous deformation field covering the entire cross-section. The cross-sectional distribution diagram extracts deformation profiles at specific locations through orthogonal projection. For example, a cross-sectional diagram is generated every 5 meters to form a visualization of the deformation distribution along the longitudinal direction of the tunnel. The deformation trend curve extracts the displacement time series of key monitoring points, such as selecting the three points with the largest displacement, and uses the least squares method to fit their displacement change rate curves. Future deformation trends are then predicted based on historical data. The combination of the cross-sectional distribution diagram and the deformation trend curve constructs a multidimensional dataset that includes spatial distribution characteristics and temporal evolution laws, forming a complete description of the overall deformation state of the tunnel.

[0075] In this embodiment, through the combination of image processing and data analysis technology, without the need to add hardware equipment, the visual reconstruction of the deformation state of the entire section and the quantitative analysis of the evolution law in the time dimension are achieved, breaking through the technical limitations of traditional point monitoring in terms of spatial coverage and time predictability.

[0076] This application effectively reduces the hardware cost of the monitoring system, while realizing spatial visualization and temporal prediction analysis of the overall deformation state of the tunnel, providing decision support data containing multi-dimensional information for engineering safety assessment.

[0077] In one embodiment, step S520 includes:

[0078] S521. Outputting a current quantitative analysis result of the soft rock tunnel according to the current deformation distribution data set.

[0079] Specifically, the current deformation distribution dataset refers to a data set containing tunnel cross-section distribution maps and deformation trend curves. Specifically, three-dimensional modeling software can be used to spatially reconstruct image data to reflect the deformation degree and time variation pattern of each tunnel section.

[0080] S522: Output the deformation monitoring result of the soft rock tunnel according to the current quantitative analysis result.

[0081] Specifically, current quantitative analysis results refer to quantitative indicators extracted from deformation distribution data through mathematical methods. Numerical analysis methods can be used to calculate deformation rate, deformation direction, and hazardous areas, transforming complex deformation characteristics into comparable numerical parameters. Deformation monitoring results refer to the final conclusions reached after secondary data processing. These can be presented using visual charts combined with warning level classification to guide construction decisions.

[0082] First, a distributed dataset containing tunnel cross-sectional geometry and deformation trends is fed into the analysis system. A spatial data parsing algorithm is then used to extract the displacement characteristics of each monitoring point. Numerical differentiation methods are then used to calculate the deformation rate, combined with vector analysis to determine the principal deformation direction. Simultaneously, a stress distribution model is used to identify potential risk areas, generating a current quantitative analysis result encompassing multiple quantitative indicators. Finally, the quantitative analysis results are compared with pre-set safety thresholds, and data fusion techniques are used to generate deformation monitoring results with warning levels, such as red, yellow, and blue, to indicate risk areas.

[0083] In this embodiment, a deformation distribution dataset is constructed from image data, fully reflecting the overall deformation state of the tunnel. Mathematical analysis methods are used to extract multidimensional quantitative parameters, addressing the limited monitoring accuracy inherent in traditional methods due to insufficient data dimensionality. Furthermore, secondary data processing transforms raw data into directly applicable monitoring conclusions, avoiding the subjective errors of manual interpretation. A phased data processing process transforms complex spatial deformation information into quantifiable engineering parameters, providing a reliable basis for construction safety assessments. This also reduces the hardware costs associated with traditional sensor deployment, improving the cost-effectiveness and maintainability of the monitoring system.

[0084] In one embodiment, step S521 includes:

[0085] According to the current deformation distribution data set, the current deformation rate, current deformation direction and current dangerous area of ​​the soft rock tunnel are output to obtain the current quantitative analysis result.

[0086] Specifically, the current deformation distribution dataset refers to structured data containing spatial displacement information generated by comparing an image sequence with a three-dimensional benchmark model. This can be achieved by superimposing analysis of tunnel cross-section distribution maps and deformation trend curves to reflect the spatial distribution characteristics of rock mass displacement. The current deformation rate refers to the rate of change of rock mass displacement per unit time. This can be achieved by fitting first-order derivative calculations of time series data to dynamically assess the urgency of deformation development. The current deformation direction refers to the spatial orientation of the rock mass displacement vector. This can be achieved by using vector decomposition and synthesis methods of a three-dimensional coordinate system to determine the dominant direction of geological body displacement. The current danger zone refers to a local area where the displacement or deformation rate exceeds a preset threshold. This can be achieved by using a clustering algorithm combined with displacement gradient detection to identify potential locations of structural instability.

[0087] By spatially registering a continuously acquired image sequence with a three-dimensional digital benchmark model, displacement change data for each monitoring point is extracted to form a deformation distribution dataset. The deformation rate is calculated by calculating the difference in displacement between adjacent time intervals, outputting the value in units of, for example, millimeters per day. The deformation direction is determined by decomposing the three-dimensional components of the displacement vector and determining the azimuth corresponding to the maximum displacement component. A density clustering algorithm is used to identify hazardous areas, classifying spatially contiguous monitoring points with displacement exceeding a safety threshold as high-risk areas. This allows for dynamic quantitative assessment of the deformation state of the entire cross-section without relying on physical sensors.

[0088] It should be noted that the calculation equations used in the calculation method to obtain the deformation rate in this embodiment are all existing technologies, and are not improved or designed in this embodiment, so they will not be described here one by one. However, it can be illustrated that the algorithm used in this embodiment is preferably an artificial intelligence algorithm, and specifically, a BP neural network algorithm can be selected to implement

[0089] In this example, by automatically calculating deformation rate and direction, the dynamic characteristics of rock mass displacement can be identified in real time, providing data support for support decisions. Cluster analysis can also be used to locate dangerous areas, providing early warning of potential landslide risks and improving tunnel construction safety.

[0090] In one embodiment, the three-dimensional digital reference model and the current image set both include spatial coordinate information;

[0091] Step S400 includes:

[0092] S410, importing the current image set into the three-dimensional digital reference model, and making the three-dimensional digital reference model correspond to the spatial coordinate information of the current image set one by one;

[0093] S420 , comparing and analyzing the current joint trend with the joint trend model to obtain a displacement change dataset of the inner wall of the soft rock tunnel and obtain a current displacement of the geological body.

[0094] Specifically, a three-dimensional digital benchmark model refers to a three-dimensional digital model of the geological body's joint orientation established using initial photographic data. This can be achieved using laser scanning or oblique photography technology, and is used to provide a benchmark reference for deformation analysis. Spatial coordinate information refers to the three-dimensional position data corresponding to each pixel in the model and image. This can be achieved using a global positioning system or total station measurement technology to ensure that the model matches real-time data in spatial dimensions. The displacement change dataset refers to the quantitative results of the tunnel wall displacement obtained through image comparison and analysis. This can be achieved using a feature point matching algorithm or optical flow method, and is used to characterize the degree of deformation of the geological body in three-dimensional space.

[0095] By importing the current image set into the 3D digital benchmark model and aligning their spatial coordinate information one-to-one, comparison errors caused by coordinate system deviations are eliminated. Based on forced alignment of spatial coordinates, the current joint strikes acquired in real time are compared pixel-by-pixel with the joint strike model in the benchmark model, and image differencing techniques are used to identify changes in joint surface position. A 3D point cloud registration algorithm is then used to calculate the spatial offsets of corresponding feature points, generating a displacement change dataset containing displacement direction and magnitude. Finally, an integral operation is used to determine the overall displacement of the geological volume.

[0096] This application effectively reduces the hardware cost of soft rock tunnel deformation monitoring and avoids construction interference caused by sensor deployment. Through spatial coordinate matching and full-section image analysis, it reduces equipment investment while improving the accuracy of displacement calculations and achieving comprehensive coverage monitoring of the tunnel wall deformation state.

[0097] In one embodiment, step S420 includes:

[0098] S421, comparing and analyzing the current joint trend with the joint trend model, and using pixel-level change detection technology to identify and obtain a current displacement change value of the inner wall of the soft rock tunnel;

[0099] S422: Obtain a displacement change data set of the inner wall of the soft rock tunnel according to the current displacement change value and the collected three-dimensional displacement, and obtain the current displacement of the geological body.

[0100] Specifically, pixel-level change detection technology refers to identifying subtle two-dimensional plane displacement changes by comparing the current image with the image differences at the corresponding position in the benchmark model pixel by pixel. This can be achieved by image grayscale value analysis or feature point matching algorithm to capture small offsets of joint strike in the plane direction. Three-dimensional displacement refers to the three-dimensional position change of the geological body in the spatial coordinate system obtained by stereo vision measurement method. This can be achieved by binocular camera or multi-view image reconstruction technology to supplement the displacement information in the vertical direction. The displacement change data set refers to the multi-dimensional displacement data set formed by integrating the two-dimensional plane displacement change value and the three-dimensional space displacement. Specifically, the displacement information of different dimensions can be associated and superimposed through the data fusion algorithm to comprehensively characterize the deformation state of the tunnel inner wall.

[0101] After the three-dimensional digital benchmark model is constructed, pixel-by-pixel comparisons are performed between the joint strikes in the current image and the benchmark model using pixel-level change detection technology to identify changes in planar displacement. For example, if a joint strike shifts laterally in the image, the amount of displacement can be quantified. Simultaneously, stereo vision measurement methods are used to obtain the actual displacement of the geological body in three-dimensional space. For example, a binocular camera is used to capture multiple angles of the same area and calculate the differences in three-dimensional coordinates. After the planar displacement change values ​​and the three-dimensional displacement values ​​are synchronously collected, a data fusion algorithm is used to generate a complete dataset containing both two-dimensional and three-dimensional displacement information. This quantifies both the local deformation and the overall displacement of the tunnel wall, avoiding errors that can result from single-dimensional measurements.

[0102] In this embodiment, non-contact monitoring is achieved through image processing technology, and the complementarity of two-dimensional and three-dimensional displacement data is combined to reduce the hardware deployment cost and improve the accuracy and spatial resolution of displacement detection. For example, the sensor in the prior art can only obtain the displacement data of discrete points, while this solution can cover the entire tunnel inner wall surface through pixel-level analysis. This enables the present application to achieve high-precision quantitative monitoring of the displacement changes of the inner wall of a soft rock tunnel, and a complete displacement data set can be obtained without relying on densely deployed physical sensors. Through the coordinated analysis of two-dimensional plane displacement and three-dimensional spatial displacement, local deformation areas and overall displacement trends can be accurately identified, solving the monitoring blind spot problem caused by the single data dimension or insufficient spatial coverage in traditional methods, and providing reliable data support for tunnel stability assessment.

[0103] In one embodiment, step S422 includes:

[0104] According to the current displacement change value and the three-dimensional displacement of the geological body calculated by using a stereoscopic vision measurement method, a displacement change data set of the inner wall of the soft rock tunnel is obtained to obtain the current displacement of the geological body.

[0105] Specifically, the current displacement change value refers to the deformation difference of the inner wall surface of the soft rock tunnel during the continuous monitoring period, which is identified by pixel-level change detection technology. Specifically, it can be achieved by image grayscale contrast or feature point matching algorithm to capture subtle local displacement changes. The stereo vision measurement method refers to the technology of solving the three-dimensional coordinates of the same geological body through multi-view images. Specifically, it can be achieved by using a binocular camera system or a multi-eye vision reconstruction algorithm, and the three-dimensional displacement of the geological body in space is obtained by parallax calculation. The displacement change dataset refers to a collection of displacement vector information at different positions on the inner wall of the soft rock tunnel in three-dimensional space. Specifically, it can be generated by spatial coordinate transformation and displacement vector superposition, and is used to quantify the displacement direction and size of each monitoring point.

[0106] During monitoring of the inner walls of soft rock tunnels, fixed-view cameras are deployed to continuously capture images of the tunnel surface. Pixel-level change detection technology is used to extract local deformation characteristics of joint orientations in images taken at adjacent time points, generating current displacement change values. Simultaneously, stereo vision measurement methods are used to perform three-dimensional reconstruction of multi-view images of the same area, calculating the absolute displacement of the geological body in a spatial coordinate system. The two data sources are fused, and the pixel-level deformation differences are mapped into a three-dimensional spatial coordinate system, generating a displacement change dataset that includes displacement direction, magnitude, and spatial distribution. This dataset achieves full-dimensional quantification of displacement vectors by eliminating monocular projection errors and combining the spatial resolution capabilities of stereo vision.

[0107] In this embodiment, the three-dimensional displacement is directly calculated based on the image data through a stereo vision measurement method, without the need to deploy physical sensors, thus reducing hardware dependence and installation costs. At the same time, combined with pixel-level change detection technology, it is possible to capture subtle deformations in the entire field in a non-contact manner, and to improve the spatial resolution of the displacement data through three-dimensional displacement calculation. This allows the present application to solve the problems of high cost and low spatial resolution caused by traditional sensor deployment, and to achieve three-dimensional full-field monitoring of the deformation of the inner wall of a soft rock tunnel. Through the coordinated processing of stereo vision measurement and pixel-level change detection, while reducing hardware investment, the dimension and accuracy of the displacement data are improved, providing a quantitative basis containing spatial vector information for tunnel stability assessment.

[0108] In one embodiment, step S100 includes:

[0109] The soft rock tunnel that has completed excavation construction is photographed and videotaped according to a preset shooting parameter set to obtain a target image set of the soft rock tunnel; wherein the preset shooting parameter set includes a preset shooting path and preset shooting parameters.

[0110] Specifically, a preset shooting parameter set refers to a set of pre-defined image acquisition rules, which can be implemented using fixed route planning and standardized parameter configurations. For example, a laser scanner or drone-mounted camera captures images at evenly spaced points along the tunnel axis. This parameter set standardizes the initial image acquisition process and ensures spatial consistency among images captured at different time points. A preset shooting path refers to the movement trajectory planning of the image acquisition device, which can be implemented using a sequence of three-dimensional coordinate points or a gridded coverage scheme, such as arranging multiple fixed shooting points along the circumference of the tunnel wall. This path ensures complete image coverage and avoids model matching errors due to differences in viewing angles. Preset shooting parameters refer to the device operating parameters during image acquisition, which can be implemented using a fixed combination of focal length, aperture value, and exposure time, such as setting the resolution to at least 20 megapixels and the illumination compensation mode to automatic. These parameters maintain consistent image quality and provide a stable data foundation for subsequent model reconstruction.

[0111] After tunnel excavation is complete, a mobile platform equipped with cameras moves along a pre-set path, capturing multi-angle images at each spatial coordinate point according to preset parameters. The spatial coordinate information of all images is recorded in real time using an inertial navigation system or laser rangefinder and mapped to a three-dimensional digital reference model. The resulting target image set has uniform spatial coverage and imaging quality, enabling subsequent comparative analysis to eliminate errors caused by varying imaging conditions through pixel-level alignment, thereby accurately identifying changes in joint orientation.

[0112] In some specific embodiments, the preset capture path can generate a circular scanning trajectory based on the tunnel cross-section shape, for example, by using a track-type mobile platform to run along a circular guide rail on the tunnel wall, triggering a capture at intervals of a certain distance. The preset capture parameters may include white balance mode, image storage format, and timestamp synchronization rules, such as saving images in RAW format and recording GPS time information.

[0113] This application can reduce the number of sensors installed, lowering equipment procurement and maintenance costs, while also ensuring the accuracy of subsequent model matching through a standardized image acquisition process. Because the initial image set has a uniform spatial reference and imaging quality, the detection error of joint orientation changes is controlled within the pixel level, thereby improving the reliability of deformation monitoring results.

[0114] In one embodiment, step S300 includes:

[0115] The soft rock tunnel is continuously photographed and filmed according to the preset shooting parameter set to obtain an image sequence of the inner wall of the soft rock tunnel corresponding to each acquisition moment, thereby obtaining the current image set.

[0116] Specifically, an image sequence refers to a collection of inner wall images arranged in chronological order, which can be obtained through timed photography or triggered photography to form an image data chain with temporal correlation for dynamic analysis of subsequent displacement changes.

[0117] After tunnel excavation is complete, initial images of the tunnel's interior walls are captured using a preset path and parameters to establish a baseline model. During subsequent monitoring, the same capture path and parameters are used for periodic image acquisition, generating a time-stamped sequence of interior wall images. By reusing the preset parameter set, data deviations caused by differences in shooting angle, lighting conditions, or equipment position are eliminated, ensuring the alignment of images in the spatial coordinate system at each moment. The changes in the direction of the interior wall joints recorded in the image sequence can be compared with the 3D model to extract displacements, thus avoiding monitoring blind spots caused by insufficient local sensor deployment.

[0118] In this embodiment, automated image acquisition is achieved through a preset parameter set, eliminating the need to deploy a physical sensor network. Overall deformation trends can be analyzed directly from image sequences. While existing sensors only capture data at discrete points, this solution uses continuous images covering the entire tunnel interior, enabling continuous spatial monitoring and improving data integrity.

[0119] This application addresses the existing problem of increased monitoring costs caused by installing multiple sensors, while ensuring consistent and continuous image acquisition through standardized shooting parameters. Image sequence-based monitoring replaces traditional sensor networks, reducing hardware investment and maintenance costs. It also eliminates blind spots through full area coverage and improves displacement detection accuracy.

[0120] The above description is merely an exemplary embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformation made by utilizing the contents of the present invention's description and drawings under the technical concept of the present invention, or directly / indirectly applied in other related technical fields, is included in the patent protection scope of the present invention.

Claims

1. A method for monitoring deformation of a soft rock tunnel, characterized in that: The following steps are involved: photographing and videotaping the soft rock tunnel after excavation to obtain a target image set of the soft rock tunnel; Establishing a three-dimensional digital benchmark model of the soft rock tunnel based on the target image set; wherein the three-dimensional digital benchmark model includes a joint strike model of the geological body of the soft rock tunnel; Continuously photographing and videotaping the soft rock tunnel to obtain a current image set corresponding to each acquisition moment; wherein each current image in the current image set records the current joint trend of the geological body; Importing the current image set into the three-dimensional digital benchmark model and comparing and analyzing the current joint trend with the joint trend model to obtain the current displacement of the geological body; A deformation monitoring result of the soft rock tunnel is obtained according to the current displacement.

2. The soft rock tunnel deformation monitoring method according to claim 1, characterized in that: The step of obtaining the deformation monitoring result of the soft rock tunnel according to the current displacement includes: generating a current deformation distribution data set of the soft rock tunnel according to the current displacement; The deformation monitoring result of the soft rock tunnel is outputted according to the current deformation distribution data set.

3. The soft rock tunnel deformation monitoring method according to claim 2, characterized in that: The step of generating a current deformation distribution data set of the soft rock tunnel according to the current displacement comprises: A tunnel cross-section distribution diagram and a deformation trend curve of the soft rock tunnel are generated according to the current displacement, and a current deformation distribution data set is obtained.

4. The soft rock tunnel deformation monitoring method according to claim 3, characterized in that: The step of outputting the deformation monitoring result of the soft rock tunnel according to the current deformation distribution data set includes: outputting a current quantitative analysis result of the soft rock tunnel according to the current deformation distribution data set; The deformation monitoring result of the soft rock tunnel is outputted according to the current quantitative analysis result.

5. The soft rock tunnel deformation monitoring method according to claim 4, characterized in that: The step of outputting the current quantitative analysis result of the soft rock tunnel according to the current deformation distribution data set includes: According to the current deformation distribution data set, the current deformation rate, current deformation direction and current dangerous area of ​​the soft rock tunnel are output to obtain the current quantitative analysis result.

6. The soft rock tunnel deformation monitoring method according to any one of claims 1 to 5, characterized in that: The three-dimensional digital reference model and the current image set both include spatial coordinate information; The step of importing the current image set into the three-dimensional digital benchmark model and comparing and analyzing the current joint trend with the joint trend model to obtain the current displacement of the geological body includes: Importing the current image set into the three-dimensional digital reference model, and making the three-dimensional digital reference model correspond to the spatial coordinate information of the current image set one by one; The current joint trend is compared and analyzed with the joint trend model to obtain a displacement change data set of the inner wall of the soft rock tunnel and obtain the current displacement of the geological body.

7. The soft rock tunnel deformation monitoring method according to claim 6, characterized in that: The step of comparing and analyzing the current joint trend with the joint trend model to obtain a displacement change dataset of the inner wall of the soft rock tunnel and obtain the current displacement of the geological body includes: Comparing and analyzing the current joint trend with the joint trend model, and using pixel-level change detection technology to identify and obtain a current displacement change value of the inner wall of the soft rock tunnel; According to the current displacement change value and the collected three-dimensional displacement, a displacement change data set of the inner wall of the soft rock tunnel is obtained to obtain the current displacement of the geological body.

8. The soft rock tunnel deformation monitoring method according to claim 7, characterized in that: The step of obtaining a displacement change data set of the inner wall of the soft rock tunnel based on the current displacement change value and the collected three-dimensional displacement to obtain the current displacement of the geological body includes: According to the current displacement change value and the three-dimensional displacement of the geological body calculated by using a stereoscopic vision measurement method, a displacement change data set of the inner wall of the soft rock tunnel is obtained to obtain the current displacement of the geological body.

9. The soft rock tunnel deformation monitoring method according to claim 8, characterized in that: The step of photographing and videotaping the soft rock tunnel after excavation to obtain a target image set of the soft rock tunnel includes: The soft rock tunnel that has completed excavation construction is photographed and videotaped according to a preset shooting parameter set to obtain a target image set of the soft rock tunnel; wherein the preset shooting parameter set includes a preset shooting path and preset shooting parameters.

10. The soft rock tunnel deformation monitoring method according to claim 9, characterized in that: The step of continuously photographing and videotaping the soft rock tunnel to obtain a current image set corresponding to each acquisition moment includes: The soft rock tunnel is continuously photographed and filmed according to the preset shooting parameter set to obtain an image sequence of the inner wall of the soft rock tunnel corresponding to each acquisition moment, thereby obtaining the current image set.

Images