Intelligent monitoring and early warning method and system for deformation of underground roadway

Abstract

The invention belongs to the technical field of mine safety monitoring and geological disaster early warning, and discloses an intelligent monitoring and early warning method and system for underground roadway deformation, and the method comprises the steps: employing a displacement monitoring anchor rod to monitor and collect displacement data, and employing a binocular camera to collect an artificial marker image; carrying out restoration processing on the artificial marker image by adopting an intelligent image enhancement method based on dark channel prior; performing identification and intelligent sensing sub-pixel-level three-dimensional positioning on the artificial marker in the repaired image to obtain three-dimensional positioning information of the artificial marker; zero-point displacement three-dimensional calibration processing is carried out on the three-dimensional positioning information of the artificial marker according to the displacement data, and the influence of surrounding rock deformation displacement on the position of the artificial marker is corrected; and performing multi-angle fusion intelligent deformation measurement according to the corrected three-dimensional positioning information of the artificial marker, and analyzing a deformation trend to realize monitoring and early warning. According to the invention, the deformation condition of the surrounding rock of the roadway can be accurately monitored in real time, and early warning is given out in time when the deformation exceeds a safety threshold.

Description

Technical Field

[0001] The present invention relates to the technical field of mine safety monitoring and geological disaster early warning, and in particular to an intelligent monitoring and early warning method and system for underground tunnel deformation. Background Art

[0002] Underground tunnels are important passages and work sites for mining, and the stability of their surrounding rocks directly affects the safe production of mines. During the mining and support process of tunnels, the surrounding rocks of the tunnels will be displaced and deformed due to various factors such as mining activities, geological stress, groundwater, and temperature changes. In severe cases, they may cause catastrophic accidents such as surrounding rock instability and collapse, posing a major safety threat to mine workers. Therefore, real-time deformation monitoring of underground tunnels and issuing early warnings when the deformation reaches a dangerous threshold have become necessary means to ensure safe production in mines.

[0003] At present, the deformation monitoring of underground tunnels mainly relies on methods such as manual inspection, fixed sensor monitoring and laser scanning. Manual inspection measures the displacement of the tunnel by visual inspection or using portable equipment, while fixed sensor monitoring installs sensors on the tunnel walls or roof to continuously collect displacement, stress and other data, while laser scanning technology relies on laser ranging to accurately measure tunnel deformation. These methods have been applied to underground tunnel deformation monitoring to a certain extent, especially fixed sensors and laser scanning, which can provide relatively accurate measurement data within a certain range.

[0004] However, existing technologies have significant deficiencies in complex underground tunnel environments. Manual inspections are inefficient and susceptible to human errors, and cannot meet real-time monitoring needs; fixed sensors are limited by installation location and monitoring range, and are prone to blind spots in complex tunnel terrain. At the same time, their sensor accuracy is easily affected by factors such as vibration and humidity; although laser scanning is accurate, it is expensive and the equipment is easily affected by underground environments such as dust and low light. In addition, most existing monitoring systems rely on cloud or background analysis, and data delays occur due to limited underground signal transmission, making it difficult to achieve rapid early warning.

[0005] In view of the existing problems, the present invention proposes an intelligent monitoring and early warning method for underground tunnel deformation, which combines intelligent sensor networks with real-time data processing technology to improve the accuracy, real-time and environmental adaptability of monitoring, thereby effectively solving the shortcomings of the existing technology. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides an intelligent monitoring and early warning method and system for underground tunnel deformation, which can monitor the deformation of tunnel surrounding rock in real time and accurately, and issue an early warning in time when the deformation exceeds the safety threshold, so as to prevent collapse accidents and ensure the safety of mine workers. The present invention aims to solve the problems of poor real-time performance, insufficient precision and poor environmental adaptability of existing monitoring methods, and realizes multi-directional accurate monitoring and rapid response of tunnel deformation in complex underground environments by collecting image restoration, zero-point displacement calibration, intelligent three-dimensional positioning and multi-angle fusion intelligent deformation detection and processing methods for harsh working conditions, thereby improving the safety level of mine production.

[0007] An intelligent monitoring and early warning method for underground tunnel deformation, the method comprising:

[0008] Step 1: Use displacement monitoring anchors to collect displacement data, and use binocular camera data to collect artificial marker images;

[0009] Step 2: Adopting the intelligent image enhancement method based on dark channel prior to perform adaptive local characteristic dynamic adjustment and repair processing on the artificial marker image;

[0010] Step 3: Identify and intelligently perceive sub-pixel 3D positioning of artificial markers in the restored image to obtain 3D positioning information of the artificial markers;

[0011] Step 4: Perform zero-point displacement stereo calibration on the three-dimensional positioning information of the artificial marker according to the displacement data to correct the influence of surrounding rock deformation displacement on the position of the artificial marker;

[0012] Step 5: Perform multi-angle fusion intelligent deformation measurement based on the corrected three-dimensional positioning information of the artificial marker, analyze the deformation trend, and realize monitoring and early warning.

[0013] Preferably, in the step 1,

[0014] Displacement data: including horizontal displacement, vertical displacement, radial displacement, cumulative displacement, relative displacement, instantaneous displacement and cumulative displacement rate;

[0015] Artificial marker image: includes artificial marker image I(x).

[0016] Preferably, in step 2, the method of using an intelligent image enhancement method based on dark channel prior to perform adaptive local characteristic dynamic adjustment and repair processing on the artificial marker image includes:

[0017]

[0018] Where I(x) is the captured image; J(x) is the corrected image; A is the estimated atmospheric light; t adapted(x) is the transmittance adjusted by the adaptive method.

[0019] Preferably, in step 3, identifying and intelligently sensing sub-pixel-level three-dimensional positioning of artificial markers in the restored image, and obtaining three-dimensional positioning information of the artificial markers includes:

[0020]

[0021]

[0022]

[0023] Where f is the focal length of the camera; B is the baseline distance of the binocular camera; d = x Lfinal -x Rfinal is the parallax; x Lfinal 、x Rfinal It represents the x-coordinate of the sub-pixel edge location of the object in the left and right images obtained by the fusion of intelligent edge detection and deep learning; Lfinal ,y Rfinal Represents the y coordinate of the sub-pixel edge positioning of the object in the left and right images obtained by the fusion of intelligent edge detection and deep learning.

[0024] Preferably, in step 4, performing zero-point displacement stereo calibration processing on the three-dimensional positioning information of the artificial marker according to the displacement data, and correcting the influence of surrounding rock deformation displacement on the position of the artificial marker includes:

[0025] P i '=R·P i +T,

[0026] Among them, P i The high-precision three-dimensional coordinates (x tri ,y tri ,z tri );P i ' is the position of the marker after calibration; R is the rotation matrix, which is used to adjust the angle change; T is the translation vector, which represents the overall displacement.

[0027] Preferably, in step 5, performing multi-angle fusion intelligent deformation measurement according to the corrected three-dimensional positioning information of the artificial marker, analyzing the deformation trend, and realizing monitoring and early warning includes:

[0028] The calibrated three-dimensional points (X, Y, Z) of the binocular camera coordinate system are formed into a three-dimensional point set P in the form of:

[0029] P t1 and P t2 is the point cloud collected at different times t1 and t2, and the point cloud P is generatedt1 and P t2 Perform normalization processing;

[0030] The Hausdorff distance algorithm is used to quantify the degree of deformation of the point cloud at different time points, and the normalized point cloud P t1 and P t2 Input the deep neural network respectively to extract the local features h of the points i t1 and h i t2 And the global feature h of the maximum pooling layer global t1 and h global t2 , and concatenate the local features with the corresponding global features to obtain the concatenated features h input ;

[0031] The concatenated features h input Send it to the deep neural network to predict the displacement of each point Where ΔX i ,ΔY i ,ΔZ i It represents the changes in the three dimensions of X, Y, and Z, and realizes real-time monitoring, prediction and early warning of tunnel deformation.

[0032] The present invention also discloses an intelligent monitoring and early warning system for underground tunnel deformation, the system is used to implement any one of the methods described above, the system comprises: a collection module, a repair module, a positioning module, a calibration module and an early warning module;

[0033] The acquisition module is used to collect displacement data using displacement monitoring anchor rod monitoring, and to collect artificial marker images using binocular camera data;

[0034] The repair module is used to perform adaptive local characteristic dynamic adjustment and repair processing on the artificial marker image by using an intelligent image enhancement method based on dark channel prior;

[0035] The positioning module is used to identify and intelligently perceive sub-pixel-level three-dimensional positioning of artificial markers in the restored image, and obtain three-dimensional positioning information of the artificial markers;

[0036] The calibration module is used to perform zero-point displacement stereoscopic calibration processing on the three-dimensional positioning information of the artificial marker according to the displacement data, and correct the influence of the surrounding rock deformation displacement on the position of the artificial marker;

[0037] The early warning module is used to perform multi-angle fusion intelligent deformation measurement according to the corrected three-dimensional positioning information of the artificial marker, analyze the deformation trend, and realize monitoring and early warning.

[0038] The present invention also discloses an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the methods described above when executing the program.

[0039] The present invention also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed, any one of the methods described above is implemented.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention provides an intelligent monitoring and early warning method for underground tunnel deformation, which significantly improves the accuracy, real-time and comprehensiveness of underground tunnel deformation monitoring by integrating intelligent image restoration technology, zero-point displacement calibration technology, intelligent three-dimensional positioning technology and multi-angle fusion intelligent deformation measurement technology. Intelligent image restoration technology effectively removes environmental interference, enhances image clarity, and ensures the accuracy of deformation monitoring data; zero-point displacement calibration technology eliminates sensor errors and ensures the reliability of displacement data; intelligent three-dimensional positioning technology provides accurate spatial positioning and comprehensively monitors tunnel deformation; multi-angle fusion intelligent measurement technology removes erroneous or low-precision point cloud data by fusing image brightness and texture information from different perspectives, quantifies the degree of deformation, analyzes deformation trends, and achieves monitoring and early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. 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 these drawings without paying creative labor.

[0043] Figure 1 A schematic diagram of an intelligent monitoring and early warning method for underground tunnel deformation according to an embodiment of the present invention;

[0044] Figure 2 Schematic diagram of the structure of an electronic device according to an embodiment of the present invention.

[0046] 1010, processor; 1020, memory; 1030, input / output interface; 1040, communication interface; 1050, bus. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0048] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present disclosure should be understood by people with ordinary skills in the field to which the present disclosure belongs. The "first", "second" and similar words used in the embodiments of the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] Embodiment 1

[0051] like Figure 1 As shown, an embodiment of the present invention provides an intelligent monitoring and early warning method for underground tunnel deformation, comprising the following steps:

[0052] Step 1: Data monitoring and collection, using displacement monitoring anchors to monitor and collect displacement data, and using binocular cameras to collect artificial marker images;

[0053] Step 2: Adaptive local characteristics dynamically adjust image restoration processing, use the intelligent image enhancement method based on dark channel prior to perform intelligent image restoration on images collected in harsh working conditions, reduce interference with images collected by artificial markers under high dust conditions in the tunnel, and improve the recognition accuracy of the positions of artificial markers;

[0054] Step 3: Intelligent 3D positioning: identify the artificial markers in the image after image restoration, and then perform intelligent perception sub-pixel 3D positioning to obtain high-precision 3D positioning information of the artificial markers.

[0055] Step 4: Zero-point displacement stereo calibration: perform zero-point displacement stereo calibration on the three-dimensional positioning information of the artificial marker according to the displacement data to correct the influence of surrounding rock deformation displacement on the position of the artificial marker;

[0056] Step 5: Multi-angle fusion intelligent deformation monitoring, perform multi-angle fusion intelligent deformation measurement based on the three-dimensional positioning information of artificial markers, analyze deformation trends, and realize monitoring and early warning.

[0057] In this embodiment, step 1: data monitoring and collection

[0058] Data monitoring and collection is divided into two parts: displacement monitoring anchor monitoring collects displacement data, and binocular camera data collects artificial marker location information.

[0059] Displacement monitoring anchor monitoring collects displacement data: including horizontal displacement, vertical displacement, radial displacement, cumulative displacement, relative displacement, instantaneous displacement and cumulative displacement rate.

[0060] Binocular camera data collects artificial marker images: including artificial marker image I(x).

[0061] In this embodiment, step 2: adaptive local characteristics dynamically adjust image restoration processing

[0062] An intelligent image enhancement method based on dark channel prior is used to perform denoising and enhancement processing on images collected in complex environments with high dust in the tunnel, reduce environmental interference on the image, and improve the recognition accuracy of the position of artificial markers. In order to improve the image dust removal effect, an adaptive local characteristic dynamic adjustment algorithm is used, and the calculation formula is:

[0063]

[0064] Where I(x) is the captured image; J(x) is the corrected image; A is the estimated atmospheric light, which is the color value of the area most severely covered by dust in the captured image. The brightest pixels in the top k% of the dark channel image are taken as candidates (here k is set to 0.1), and the corresponding maximum brightness in the original image is selected as the atmospheric light A; t adapted (x) is the transmittance adjusted by the adaptive method, and the calculation formula is

[0065] t adapted(x) =t(x)·α(x) (2)

[0066] Among them, t(x) is the original refractive index calculated by the dark channel, α(x) is the adaptive adjustment factor, and the calculation formula of t(x) is:

[0067]

[0068] Among them, ω is the adjustment coefficient, which is used to maintain a certain dust composition and avoid excessive treatment; A c is the atmospheric light component in color channel c; I c (y) is the intensity value of the image in color channel c, and Ω(x) is the local window centered on x, which is set to a 15×15 window here.

[0069] The calculation formula of α(x) in formula (2) is:

[0070]

[0071] Among them, μ is the local mean of the image, the window size is set consistent with Ω(x), σ is the local standard deviation of the image, and β is a constant used to adjust the sensitivity of the adaptive factor.

[0072] At this point, the restored image J(x) is obtained, which improves the recognition accuracy of the position of the artificial landmark.

[0073] In this embodiment, step three: sub-pixel positioning algorithm integrating intelligent edge detection and deep learning

[0074] For the artificial markers in the image J(x) after image restoration, a sub-pixel positioning algorithm that integrates intelligent edge detection and deep learning is used to obtain high-precision three-dimensional positioning information of the artificial markers, effectively realize the recognition and positioning of subtle targets, break through the pixel-level limitation, and achieve sub-pixel accuracy. The specific method is as follows:

[0075] First, an intelligent edge detection algorithm is designed to detect the preliminary edge contours in the image. The edge data generated in this stage will be used as the input of the deep learning model to improve the model's sensitivity to details. A deep network is used to extract edge features at different scales, and a feature map F is generated through operations such as convolution and pooling in the network layer, where each pixel represents the probability that the point in the feature map is an edge. The feature map F is a probability density map of the edge. After training, the network can learn sub-pixel position information so that the output of the feature map can reflect the accurate edge distribution. Based on the output feature map F of the deep network, the interpolation strategy can be used to accurately locate the target point at the sub-pixel level, and the sub-pixel edge probability p(x,y) at different positions can be obtained. The calculation formula is:

[0076]

[0077] Among them, p(i,j) is the edge probability value of the pixel output of the feature map F at a certain position (x,y), and ω(i,j) is the interpolation weight, which is calculated by the gradient direction and intensity in the deep network feature output. The small offset value to the edge is calculated through the deep network model to generate the correction factor (Δx, Δy). After obtaining the preliminary edge position (x', y'), the correction factor is applied to the positioning result to achieve sub-pixel precision edge positioning:

[0078] (x final ,y final )=(x'+Δx,y'+Δy) (6)

[0079] Finally, the calculated pixel values ​​are used to obtain high-precision three-dimensional coordinates (X, Y, Z) using the sub-pixel precision disparity value and the triangulation formula:

[0080]

[0081]

[0082]

[0083] Where f is the focal length of the camera; B is the baseline distance of the binocular camera; d = x Lfinal -x Rfinal is the parallax; x Lfinal 、x Rfinal It represents the x-coordinate and y-coordinate of the sub-pixel edge location of the object in the left and right images obtained by the fusion of intelligent edge detection and deep learning. Lfinal ,y Rfinal Represents the y coordinate of the sub-pixel edge positioning of the object in the left and right images obtained by the fusion of intelligent edge detection and deep learning.

[0084] In this embodiment, step 4: zero-point displacement stereo calibration

[0085] The displacement data is used to calibrate the position of the artificial markers in stereo. First, the anchor fixing position is defined as the origin of the coordinate system, and the relative position of each anchor is recorded. Then, the starting position of the marker is set as the reference position, and the relative position of the marker is recorded in the coordinate system.

[0086] The displacement data collected by the anchor rod is used as input, and the three-dimensional displacement of the marker relative to the anchor rod is calculated through the spatial geometric transformation model. Using the displacement data collected from multiple angles, the calibration equation is established through triangulation or least squares method, and the calculation formula is:

[0087] D' i =R·P i +T (10)

[0088] Among them, P i is the point source, and is the high-precision three-dimensional coordinate (x tri ,y tri ,z tri ), belongs to the landmark point set P i ∈P, that is; D i ' is the calibrated marker position, i.e., the target point, which belongs to the target point set D i ′∈D′; R is the rotation matrix, which is used to adjust the angle change; T is the translation vector, which represents the overall displacement. Using the multi-directional displacement data collected by the monitoring anchor, the three-dimensional displacement vector P of the marker is calculated. i ', the calculation formula is as follows:

[0089] P i ′=D i ′-P anchor (11)

[0090] Among them, P anchor is the anchor position (x anc ,y anc ,z anc ).

[0091] Finally, the iterative closest point algorithm is used to optimize the displacement calibration results of the markers to obtain the calibrated marker positions, that is, to minimize the weighted mean square error E(R,T) of the two point sets, thereby optimizing the rotation matrix R and the translation vector T:

[0092]

[0093] The optimization process is:

[0094] (1) Calculate the closest point match for each point source P i ∈P, first find the target point set D i The closest point D in ′∈D′ i * , satisfying the formula:

[0095]

[0096] (2) Calculate the centroid of the source point set and the centroid of the target point set respectively:

[0097]

[0098]

[0099] (3) Remove the influence of the centroid and centralize the point set:

[0100]

[0101]

[0102] (4) Calculate the covariance matrix using the decentralized point set:

[0103]

[0104] (5) Perform singular value decomposition on the covariance matrix:

[0105]

[0106] Among them, U is a left singular vector matrix, ∑ is a diagonal matrix, V is a right singular vector matrix, V T represents transpose. At this time, the rotation matrix R can be calculated by the following formula:

[0107] R=V·U T (20)

[0108] (6) Calculate the translation vector T using the center of mass relationship:

[0109]

[0110] (7) Update the source point set according to the calculated rotation matrix R and translation vector T:

[0111] P i new =R·P i +T (22)

[0112] (8) If the change of the error function E(R,T) is less than the preset threshold or the maximum number of iterations is reached, the iteration is stopped, otherwise return to (1).

[0113] In this embodiment, step 5: multi-angle fusion intelligent deformation monitoring

[0114] After obtaining the calibrated sub-pixel landmark positions, an intelligent deformation prediction model based on a neural network is used to monitor the deformation degree between clouds. The specific method is as follows:

[0115] The calibrated three-dimensional points (X, Y, Z) of the binocular camera coordinate system are formed into a three-dimensional point set P in the form of: P t1 ={p i =(X i ,Y i ,Z i )} and P t2 = {q j =(X j ,Y j ,Z j )} is the point cloud collected at different times t1 and t2, and the generated point cloud is Pt1 and P t2 Normalization processing, using the Hausdorff distance algorithm to quantify the degree of point cloud deformation at different time points, and normalized point cloud and Input the deep neural network respectively. The specific structure and algorithm processing of the neural network are as follows:

[0116] First, for each point p i Constructing a domain through the K-nearest neighbor algorithm:

[0117] N(p i )={p i1 ,p i2 ,...,p ik} (twenty three)

[0118] Among them, p ij Yes i field point.

[0119] Then calculate the coordinate difference of each field point relative to the center point and construct the neighborhood feature f ij :

[0120] f ij ={p ij -p i ,p i} (twenty four)

[0121] Among them, p i Represents the absolute coordinates of the center point, retaining the global position information, p ij -p i is the relative coordinate of the neighborhood, representing the local geometric structure, such as obtaining the neighborhood feature matrix F i ={f ij |j=1,2,...,K}, K means the neighborhood matrix contains K numbers. Then the convolutional neural network is used to perform local feature h on the neighborhood features. i extract:

[0122] h i =ReLU(CNN(F i )) (25)

[0123] CNN stands for convolutional neural network, and ReLU(x) = max(0,x) is the ReLU activation function. Use maximum pooling to aggregate domain features and obtain the local features of each center point:

[0124] h i =MaxPool(h i ) (26)

[0125] Through the above process, we can get Pt1 and P t2 The local feature h i t1 and h i t2 .

[0126] Then the global feature extraction is performed, and the local features h of the point cloud are fused through multi-pooling operation. i Aggregated into a single global feature h global :

[0127] h global =[MaxPool(h i ),AvgPool(h i )] (27)

[0128] Among them, AvgPool represents average pooling, that is, finding the average value for each feature dimension. i t1 and h i t2 Substituting into the above formula, we can obtain the global features h global t1 and h global t2 , the local features h after splicing i t1 and h i t2 and the global feature h global t1 and h global t2 The deep neural network processes the features of each point in the input point cloud, including the concatenation of local features and global features. input =[h i ,h global ], and obtain the comprehensive feature representation of the point. Then use a multi-layer one-dimensional convolutional network to extract the comprehensive features of the point layer by layer. Each layer of the convolutional network further improves the feature expression ability through a nonlinear activation function. In the deep structure of the convolutional network, the point cloud features are reduced and mapped layer by layer to extract more abstract feature representations. Through the output layer of the convolutional network, the three-dimensional displacement of each point is predicted, including the change ΔX in the three dimensions of X, Y, and Z. i ,ΔY i ,ΔZ i . Get the displacement information of all points in the point cloud as the basic data for deformation monitoring, and predict and output the displacement of each point Realize real-time monitoring, prediction and early warning of tunnel deformation.

[0129] The technical solution of the present invention,

[0130] 1. The present invention proposes an image restoration processing method with adaptive local characteristic dynamic adjustment, which is specifically aimed at the impact of high dust on image acquisition in complex tunnels. This method combines the dark channel prior algorithm and designs a method for adaptive local dynamic adjustment of parameters, which can effectively reduce the blur of images in high dust environments, thereby improving image clarity and usability. By intelligently adjusting the restoration strategy of local image areas, this method adapts to dynamic changes in complex tunnel environments while ensuring image quality, greatly improving the stability and accuracy of image restoration processing.

[0131] 2. The zero-point displacement stereo calibration method proposed in this invention calibrates the three-dimensional coordinate system of the binocular camera by using the data collected by the displacement monitoring anchor rod, which significantly improves the accuracy of deformation monitoring of the surrounding rock of the tunnel for the special scene of the underground tunnel. This method effectively combines the monitoring data with the visual information, accurately calibrates the three-dimensional spatial positioning of the camera, ensures the high accuracy and real-time performance of deformation monitoring, and provides more reliable technical support for the safety monitoring of underground tunnels.

[0132] 3. The sub-pixel positioning algorithm that integrates intelligent edge detection and deep learning proposed in the present invention significantly improves positioning accuracy by combining traditional edge detection technology and advanced deep learning methods. Based on edge information, the algorithm uses deep neural networks to extract and fuse features, accurately capturing sub-pixel details in the image, thereby achieving high-precision positioning. When processing images in complex environments, the algorithm can adaptively identify and optimize edge features, effectively reduce noise interference, improve positioning stability and robustness, and provide strong technical support for refined three-dimensional positioning.

[0133] It should be noted that the method of the embodiment of the present disclosure can be performed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present disclosure, and the multiple devices will interact with each other to complete the described method.

[0134] It should be noted that some embodiments of the present disclosure are described above. Other embodiments are within the scope of the attached claims. In some cases, it should be understood that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention. The actions or steps recorded in the claims can be performed in an order different from that in the above embodiment and still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0135] Embodiment 2

[0136] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present invention further discloses an intelligent monitoring and early warning system for underground tunnel deformation, the system is used to implement any of the methods described, and the system includes: a collection module, a repair module, a positioning module, a calibration module and an early warning module;

[0137] The acquisition module is used to collect displacement data using displacement monitoring anchor rods and to collect images of artificial markers using binocular camera data;

[0138] The repair module is used to perform adaptive local characteristic dynamic adjustment and repair processing on the artificial marker image by using an intelligent image enhancement method based on dark channel prior;

[0139] The positioning module is used to identify and intelligently perceive sub-pixel-level 3D positioning of artificial markers in the restored image, and obtain 3D positioning information of artificial markers;

[0140] The calibration module is used to perform zero-point displacement stereoscopic calibration processing on the three-dimensional positioning information of the artificial marker according to the displacement data, and correct the influence of surrounding rock deformation displacement on the position of the artificial marker;

[0141] The early warning module is used to perform multi-angle fusion intelligent deformation measurement based on the corrected three-dimensional positioning information of artificial markers, analyze deformation trends, and achieve monitoring and early warning.

[0142] The system of the above embodiment is used to implement a corresponding intelligent monitoring and early warning method for underground tunnel deformation in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0143] It should be noted that the above-mentioned intelligent monitoring and early warning system for underground tunnel deformation is embodied in the form of functional units. The term "module" here can be implemented in the form of software and / or hardware, and is not specifically limited to this.

[0144] For example, a "module" may be a software program, a hardware circuit, or a combination of the two that implements the above functions. The hardware circuit may include an application specific integrated circuit (ASIC), an electronic circuit, a processor (such as a shared processor, a dedicated processor, or a group processor, etc.) and a memory for executing one or more software or firmware programs, a combined logic circuit, and / or other suitable components that support the described functions.

[0145] Embodiment 3

[0146] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, an intelligent monitoring and early warning method for underground tunnel deformation as described in any of the above embodiments is implemented.

[0147] Figure 2 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.

[0148] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0149] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0150] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0151] The communication interface 1040 is used to connect a communication module (not shown in the figure) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB (Universal Serial Bus), network cable, etc.), or through a wireless mode (such as mobile network, WIFI (Wireless Fidelity), Bluetooth, etc.).

[0152] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0153] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0154] The system of the above embodiment is used to implement a corresponding intelligent monitoring and early warning method for underground tunnel deformation in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0155] Embodiment 4

[0156] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present disclosure also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute an intelligent monitoring and early warning method for underground tunnel deformation as described in any of the above embodiments.

[0157] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0158] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute an intelligent monitoring and early warning method for underground tunnel deformation as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0159] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present disclosure (including the claims) is limited to these examples. Based on the concept of the present disclosure, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present disclosure as described above, which are not provided in detail for the sake of simplicity.

[0160] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present disclosure difficult to understand, the known power / ground connections to the integrated circuit (IC) chips and other components may or may not be shown in the provided figures. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present disclosure difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present disclosure will be implemented (that is, these details should be fully within the scope of understanding of those skilled in the art). Where specific details (e.g., circuits) are set forth to describe exemplary embodiments of the present disclosure, it is apparent to those skilled in the art that the embodiments of the present disclosure can be implemented without these specific details or with changes in these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0161] Although the present disclosure has been described in conjunction with specific embodiments of the present disclosure, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0162] Therefore, the units of each example described in the embodiments of the present application can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present application.

[0163] The embodiments of the present disclosure are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. An intelligent monitoring and early warning method for underground tunnel deformation, characterized in that: The method comprises: Step 1: Use displacement monitoring anchors to collect displacement data, and use binocular camera data to collect artificial marker images; Step 2: Adopting the intelligent image enhancement method based on dark channel prior to perform adaptive local characteristic dynamic adjustment and repair processing on the artificial marker image; Step 3: Identify and intelligently perceive sub-pixel 3D positioning of artificial markers in the restored image to obtain 3D positioning information of the artificial markers; Step 4: Perform zero-point displacement stereo calibration on the three-dimensional positioning information of the artificial marker according to the displacement data to correct the influence of surrounding rock deformation displacement on the position of the artificial marker; Step 5: Perform multi-angle fusion intelligent deformation measurement based on the corrected three-dimensional positioning information of the artificial marker, analyze the deformation trend, and realize monitoring and early warning.

2. The method according to claim 1, characterized in that In the step 1, Displacement data: including horizontal displacement, vertical displacement, radial displacement, cumulative displacement, relative displacement, instantaneous displacement and cumulative displacement rate; Artificial marker image: includes artificial marker image I(x).

3. The method according to claim 1, characterized in that In the step 2, the method of using the dark channel prior-based intelligent image enhancement method to perform adaptive local characteristic dynamic adjustment and repair processing on the artificial marker image includes: Where I(x) is the captured image; J(x) is the corrected image; A is the estimated atmospheric light; t adapted (x) is the transmittance adjusted by the adaptive method.

4. The method according to claim 1, characterized in that In the step 3, the artificial markers in the restored image are identified and intelligently sensed for sub-pixel 3D positioning, and the 3D positioning information of the artificial markers is obtained, including: Where f is the focal length of the camera; B is the baseline distance of the binocular camera; d = x Lfinal -x Rfinal is the parallax; x Lfinal 、x Rfinal It represents the x-coordinate of the sub-pixel edge location of the object in the left and right images obtained by the fusion of intelligent edge detection and deep learning; Lfinal ,y Rfinal Represents the y coordinate of the sub-pixel edge positioning of the object in the left and right images obtained by the fusion of intelligent edge detection and deep learning.

5. The method according to claim 1, characterized in that In the fourth step, the three-dimensional positioning information of the artificial marker is calibrated for zero-point displacement according to the displacement data, and the influence of the surrounding rock deformation displacement on the position of the artificial marker is corrected, including: P i '=R·P i +T, Among them, P i The high-precision three-dimensional coordinates (x tri ,y tri ,z tri );P i ' is the position of the marker after calibration; R is the rotation matrix, which is used to adjust the angle change; T is the translation vector, which represents the overall displacement.

6. The method according to claim 1, characterized in that In the step 5, multi-angle fusion intelligent deformation measurement is performed according to the corrected three-dimensional positioning information of the artificial marker, and deformation trend is analyzed to achieve monitoring and early warning, including: The calibrated three-dimensional points (X, Y, Z) of the binocular camera coordinate system are formed into a three-dimensional point set P in the form of: P t1 and P t2 is the point cloud collected at different times t1 and t2, and the point cloud P is generated t1 and P t2 Perform normalization processing; The Hausdorff distance algorithm is used to quantify the degree of deformation of the point cloud at different time points, and the normalized point cloud P t1 and P t2 Input the deep neural network respectively to extract the local features h of the points i t1 and h i t2 And the global feature h of the maximum pooling layer global t1 and h global t2 , and concatenate the local features with the corresponding global features to obtain the concatenated features h input ; The concatenated features h input Send it to the deep neural network to predict the displacement of each point Where ΔX i ,ΔY i ,ΔZ i It represents the changes in the three dimensions of X, Y, and Z, and realizes real-time monitoring, prediction and early warning of tunnel deformation.

7. An intelligent monitoring and early warning system for underground tunnel deformation, the system is used to implement the method described in any one of claims 1 to 6, characterized in that: The system includes: a collection module, a repair module, a positioning module, a calibration module and an early warning module; The acquisition module is used to collect displacement data using displacement monitoring anchor rod monitoring, and to collect artificial marker images using binocular camera data; The repair module is used to perform adaptive local characteristic dynamic adjustment and repair processing on the artificial marker image by using an intelligent image enhancement method based on dark channel prior; The positioning module is used to identify and intelligently perceive sub-pixel-level three-dimensional positioning of artificial markers in the restored image, and obtain three-dimensional positioning information of the artificial markers; The calibration module is used to perform zero-point displacement stereoscopic calibration processing on the three-dimensional positioning information of the artificial marker according to the displacement data, and correct the influence of the surrounding rock deformation displacement on the position of the artificial marker; The early warning module is used to perform multi-angle fusion intelligent deformation measurement according to the corrected three-dimensional positioning information of the artificial marker, analyze the deformation trend, and realize monitoring and early warning.

8. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed, the method according to any one of claims 1 to 6 is implemented.

Images