A method for calculating the degree of potential slip area of a rock excavation face

Abstract

The application discloses a rock excavation face slip potential area degree calculation method, and belongs to the field of slip potential area evaluation. The method comprises the following steps: obtaining a fracture image of an excavation face; obtaining start point coordinates and end point coordinates of all fracture traces; linearizing each fracture trace according to the start point coordinates and the end point coordinates of each fracture trace, and obtaining a linearized image of the fractures of the excavation face; performing skeleton extraction on the linearized image of the fractures of the excavation face, and obtaining a trace skeleton image; sequentially traversing pixel points of the trace skeleton image, and obtaining a potential slip area degree index of the pixel points; normalizing the potential slip area degree index of each pixel point, obtaining a potential slip area degree matrix, and mapping the potential slip area degree matrix to a pseudo-color space, thereby completing the calculation and visualization of the rock excavation face slip potential area degree. The application solves the problem that the existing potential slip area evaluation does not consider the rock fracture trace and the precision is insufficient based on a multi-factor index calculation method.

Description

Technical Field

[0001] The invention belongs to the field of slip potential area assessment, and in particular relates to a method for calculating the extent of slip potential area of ​​a rock excavation surface. Background Art

[0002] Rock fissures are a common structural feature of the excavation face during tunnel excavation. When construction encounters unfavorable geological conditions such as karst, faults, and fracture zones, various construction steps often cause external loads to cause cracks to develop, slip, and even break in the excavation face. This can lead to serious engineering accidents such as mud and water inrush and large-scale face collapse, resulting in unnecessary loss of life and property during tunnel excavation. Therefore, fully utilizing the excavation face crack image dataset and assessing the extent of potential slip areas during tunnel construction based on this information is crucial for the smooth progress of tunnel construction.

[0003] A thorough understanding of the impact of cracks on the strength and deformation behavior of rock masses is crucial for determining the stability of underground excavations in rock engineering. Furthermore, the geometric distribution of cracks significantly influences rock strength and deformation properties. Currently, methods for identifying cracks in underground excavation surfaces are broadly categorized into contact and non-contact measurement. Contact measurement primarily relies on manual measurement and sketching by experienced geologists using tools such as vernier calipers and geological compasses. However, this method not only exposes workers to high-risk environments but also relies heavily on subjective experience. Consequently, this method has significant drawbacks. Non-contact measurement primarily relies on manual recognition using traditional image processing algorithms and intelligent recognition using deep learning algorithms based on machine vision to extract cracks from captured images of tunnel excavation surfaces. Traditional image processing recognition algorithms primarily rely on edge detection algorithms such as the Canny operator and the Sobel operator. Their drawback is that they are sensitive to lighting and noise, requiring manual threshold setting to process image data in varying conditions. In contrast, deep learning algorithms based on machine vision, such as convolutional neural networks (CNNs) and image segmentation algorithms (e.g., U-Net, DeeplabV3, etc.), can automatically learn complex features in images, thereby enabling automated and intelligent recognition of gaps with clear semantic meaning. However, these algorithms typically require large amounts of image data training to ensure high recognition accuracy. Notably, deep learning algorithms for detecting 3D gaps, such as 3D CNNs and PointNet, have also achieved varying degrees of development in recent years, but their effectiveness remains suboptimal, primarily focusing on the classification of single objects or the segmentation of regular objects.

[0004] In terms of potential slip zone assessment in rock, the slip and collapse of rock structural surfaces are often controlled by a complex combination of various discontinuities, and the discontinuities of these discontinuities are often closely related to discrete folds and cracks on the rock surface. Currently, the determination of the extent of potential slip zones in rock mostly focuses on natural rock fractures generated under natural conditions such as steep rock slopes in complex mountainous areas and freeze-thaw cycles in cold regions. Determination methods include geometric determination based on three-dimensional fracture models and fracture evolution determination based on mechanical principles. However, there is little research on the determination of the extent of potential slip zones in artificial fracture networks on excavation surfaces. Furthermore, for rock excavation surfaces that are constantly cycling in rock engineering, the potential for slip and collapse is closely related to the fracture traces on the surface. Therefore, proposing a potential slip zone assessment based on fracture traces on the excavation surface is crucial for the safe construction of rock engineering. Summary of the Invention

[0005] In response to the above-mentioned deficiencies in the prior art, the present invention provides a method for calculating the potential slip zone extent of a rock excavation surface, which solves the problem that the existing potential slip zone assessment does not take rock fracture traces into consideration and is therefore not accurate enough.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is: a method for calculating the potential area of ​​slippage of a rock excavation surface, comprising the following steps:

[0007] S1. Obtaining crack images of the excavation surface;

[0008] S2. Establish a rectangular coordinate system based on the edge of the crack image of the excavation surface, and obtain the starting point coordinates and end point coordinates of all crack traces;

[0009] S3. Setting a distance threshold, and obtaining key nodes of each fracture trace based on the starting point coordinates and the end point coordinates of each fracture trace based on the distance threshold, and linearizing each fracture trace based on the key nodes of each fracture trace to obtain a linearized fracture image of the excavation surface;

[0010] S4, performing skeleton extraction on the linearized image of the excavation surface crack to obtain a trace skeleton image;

[0011] S5. sequentially traverse the pixel points of the trace skeleton image, and obtain a potential slip region degree index of the pixel point based on the product of the length and slope of the straightened crack trace segment within the grid area centered on the pixel point according to a preset grid width;

[0012] S6. Normalize the potential slip area degree index of each pixel point to obtain a potential slip area degree matrix, and map the potential slip area degree matrix to a pseudo-color space to complete the calculation and visualization of the potential slip area degree of the rock excavation surface.

[0013] Furthermore, the step S3 is specifically as follows:

[0014] S301, setting a distance threshold;

[0015] S302, obtaining the starting point coordinates and the ending point coordinates of the current crack trace;

[0016] S303, obtaining a straight line function of the straight line where the starting point coordinates and the end point coordinates are located;

[0017] S304, calculating the distance between all pixel points on the current crack trace and the straight line function based on the straight line function of the starting point coordinates and the end point coordinates;

[0018] S305: Based on the distance between each pixel point on the current crack trace and the straight line function, the pixel point with the largest distance is selected as a candidate key node point;

[0019] S306: Determine whether the distance between the key node candidate point and the straight line function is less than the distance threshold. If so, directly use the end point of the current crack trace as the key node. Otherwise, use the key node candidate point closest to the starting point of the current crack trace as the key node, and use the key node as the new starting point of the current crack trace. Return to step S303 until the end point of the current crack trace is output as the key node.

[0020] S307. Connect the starting point of the current crack trace and each key node in sequence to complete the linearization of the current crack trace, and return to step S302 to perform the linearization of the next crack trace until the linearization of all crack traces on the excavation surface crack image is completed to obtain the linearized image of the excavation surface crack.

[0021] Furthermore, the step S5 is specifically as follows:

[0022] S501, setting grid width;

[0023] S502, obtaining the current pixel point according to the trace skeleton image;

[0024] S503, dividing the grid with the current pixel as the center according to the grid width to obtain the grid area of ​​the current pixel;

[0025] S504: Obtain a straightened line segment of the crack trace within the grid area of ​​the current pixel to obtain the crack line segment of the current pixel;

[0026] S505: Obtain the number of pixels of the crack segment of the current pixel point, and obtain the length of the crack segment of the current pixel point;

[0027] S506, obtaining the slope of the crack line segment at the current pixel point;

[0028] S507: Obtain a potential slip region degree index of the current pixel point according to the product of the length of the crack line segment of the current pixel point and the slope of the crack line segment of the current pixel point;

[0029] S508: Return to step S502 to obtain the potential slip area degree index of the next pixel point.

[0030] Furthermore, the expression of the potential slip region degree index of the current pixel point in step S507 is:

[0031]

[0032] Among them, ω now is the potential slip area degree indicator of the current pixel; is the penalty coefficient of the mth crack line segment of the current pixel; m is the crack line segment number of the current pixel; l m is the length of the mth crack line segment at the current pixel; k m is the slope of the mth crack line segment at the current pixel; |·| is the absolute value; a and b are the specific values ​​of the penalty coefficients determined according to the actual project.

[0033] Furthermore, the S6 is specifically as follows:

[0034] S601, normalizing the potential slip region degree index of each pixel point to obtain the normalized potential slip region degree index of each pixel point;

[0035] S602. Establish an empty matrix based on the size of the trace skeleton image, and arrange the normalized potential slip area degree index of each pixel based on the pixel coordinates to obtain a potential slip area degree matrix. Map the potential slip area degree matrix to a pseudo-color space of [0-255] to complete the calculation and visualization of the potential slip area degree of the rock excavation surface.

[0036] Furthermore, the expression of the potential slip region degree index after normalization of each pixel point in step S601 is:

[0037]

[0038] Among them, ω r ' is the potential slip area degree index of the r-th pixel after normalization; ω r is the potential slip area degree index of the r-th pixel; ω min is the minimum value of the potential slip area degree index of each pixel; ω max It is the maximum value of the potential slip area degree index of each pixel.

[0039] The beneficial effects of the present invention are as follows: the present invention mainly extracts and quantifies the key features and distribution of the rock excavation surface based on deep learning and traditional image algorithms, and proposes an index calculation method for the degree of fragmentation of the cyclic excavation surface and the potential slip area, which provides an important basis for the characterization and prediction of different rock structural surfaces during the excavation process; at the same time, in the process of rock tunnel excavation, accurately grasping the structural characteristics of the excavation surface is crucial to the safe implementation of tunnel construction; the present invention evaluates the structural characteristics of the excavation surface by finely characterizing and analyzing parameters such as the length and inclination of the crack trace, and calculating the degree of fragmentation and potential slip area of ​​the rock excavation surface, which provides an important reference for the extraction of fine rock structural characteristics and ensuring the safe construction of tunnels. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Flow chart of the method of the present invention.

[0041] Figure 2 Schematic diagram of the straightening of the crack trace in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0043] like Figure 1 As shown, in one embodiment of the present invention, a method for calculating the potential area degree of slippage of a rock excavation face includes the following steps:

[0044] S1. Obtaining crack images of the excavation surface;

[0045] S2. Establish a rectangular coordinate system based on the edge of the crack image of the excavation surface, and obtain the starting point coordinates and end point coordinates of all crack traces;

[0046] S3. Setting a distance threshold, and obtaining key nodes of each fracture trace based on the starting point coordinates and the end point coordinates of each fracture trace based on the distance threshold, and linearizing each fracture trace based on the key nodes of each fracture trace to obtain a linearized fracture image of the excavation surface;

[0047] S4, performing skeleton extraction on the linearized image of the excavation surface crack to obtain a trace skeleton image;

[0048] S5. sequentially traverse the pixel points of the trace skeleton image, and obtain a potential slip region degree index of the pixel point based on the product of the length and slope of the straightened crack trace segment within the grid area centered on the pixel point according to a preset grid width;

[0049] S6, normalize the potential slip region degree index of each pixel point to obtain a potential slip region degree matrix, and map the potential slip region degree matrix to a pseudo-color space to complete the calculation and visualization of the rock excavation surface slip potential region degree.

[0050] In this embodiment, the skeleton module of python is used to extract the skeleton of the excavation surface crack straightening image to obtain a trace skeleton image.

[0051] The step S3 is specifically:

[0052] S301, setting a distance threshold;

[0053] S302, obtaining the start point coordinates and end point coordinates of the current crack trace;

[0054] S303, obtaining the straight line function of the straight line where the start point coordinates and the end point coordinates are located;

[0055] S304, calculating the distance between all pixel points on the current crack trace and the straight line function according to the straight line function of the straight line where the start point coordinates and the end point coordinates are located;

[0056] S305, taking the pixel point with the maximum distance as the key node candidate point according to the distance between each pixel point on the current crack trace and the straight line function;

[0057] S306, judging whether the distance between the key node candidate point and the straight line function is less than the distance threshold, if yes, directly taking the end point of the current crack trace as the key node, otherwise, taking the key node candidate point closest to the start point of the current crack trace as the key node, and taking the key node as the new start point of the current crack trace, returning to step S303 until the end point of the current crack trace is output as the key node;

[0058] S307, connecting the start point of the current crack trace and each key node in turn to complete the straightening of the current crack trace, and returning to step S302 to perform the straightening of the next crack trace until the straightening of all crack traces on the excavation surface crack image is completed, obtaining the excavation surface crack straightening image.

[0059] In this embodiment, as shown in Figure 2 , when calculating the crack curve shown in (a) of Figure 2 , the two ends of the curve are defined as the start node and the end node, and the line function W1 is formed by connecting the two ends; the distance from any position between the two end nodes to the line function W1 is calculated to obtain a set of Euclidean distances d i , and the maximum value D max, according to the manually set threshold D, the key nodes are identified and recorded. Figure 2 As shown in (b), the key node obtained is used as the next starting node to repeat the above calculation. Inevitably, in the actual calculation process, the following will occur Figure 2 If the distances of all nodes from the starting node to the end node of the line function W3 shown in (c) are less than the manually set threshold D, the end node of the line function is directly output as the key node to complete the approximate linearization process. Figure 2 [X0, Y0] is the starting point coordinate of the crack trace; [X n ,Y n ] is the coordinate of the end point of the crack trace; [X i ,Y i ] is the coordinate of the i-th pixel point on the crack trace; [X k ,Y k ] is the coordinate of the key node on the crack trace; [X0', Y0'] is the starting coordinate of the second round of linearization of the crack trace; [X k ',Y k '] is the coordinate of the key node of the second round of linearization on the crack trace; D' max is the maximum distance of the second round of linearization of the crack trace; [X0”, Y0”] is the starting coordinate of the third round of linearization of the crack trace; [X k ”,Y k ”] are the coordinates of the key nodes of the third round of linearization on the crack trace; D” max is the maximum distance of the third round of straightening of the crack trace.

[0060] The step S5 is specifically as follows:

[0061] S501, setting grid width;

[0062] S502, obtaining the current pixel point according to the trace skeleton image;

[0063] S503, dividing the grid with the current pixel as the center according to the grid width to obtain the grid area of ​​the current pixel;

[0064] S504: Obtain a straightened line segment of the crack trace within the grid area of ​​the current pixel to obtain the crack line segment of the current pixel;

[0065] S505: Obtain the number of pixels of the crack segment of the current pixel point, and obtain the length of the crack segment of the current pixel point;

[0066] S506, obtaining the slope of the crack line segment at the current pixel point;

[0067] S507: Obtain a potential slip region degree index of the current pixel point according to the product of the length of the crack line segment of the current pixel point and the slope of the crack line segment of the current pixel point;

[0068] S508: Return to step S502 to obtain the potential slip area degree index of the next pixel point.

[0069] The expression of the potential slip region degree index of the current pixel point in step S507 is:

[0070]

[0071] Among them, ω now is the potential slip area degree indicator of the current pixel; is the penalty coefficient of the mth crack line segment of the current pixel; m is the crack line segment number of the current pixel; l m is the length of the mth crack line segment at the current pixel; k m is the slope of the mth crack line segment at the current pixel; |·| is the absolute value; a and b are the specific values ​​of the penalty coefficients determined according to the actual project.

[0072] In this embodiment, in actual engineering, the value of the inclination angle is a key task for the potential slip area assessment, and the inclination angle has obvious symmetry. For example, inclination angles of 45° and 135° are equivalent in the actual potential slip area assessment because they both represent slope inclination angles in the same direction. This symmetry means that when performing the potential slip area assessment, it is only necessary to consider the absolute value of the inclination angle when the crack trace is approximated as a straight line, without distinguishing its specific tilt direction. Therefore, in order to simplify the risk assessment process and improve computational efficiency, the present invention adopts a method of statistically calculating the absolute value of the crack trace approximating a straight line. Through this method, not only can the main inclination information of the crack trace be accurately captured, but also the unnecessary computational complexity introduced due to the symmetry of the inclination angle can be eliminated.

[0073] The S6 is specifically:

[0074] S601, normalizing the potential slip region degree index of each pixel point to obtain the normalized potential slip region degree index of each pixel point;

[0075] S602. Establish an empty matrix based on the size of the trace skeleton image, and arrange the normalized potential slip area degree index of each pixel based on the pixel coordinates to obtain a potential slip area degree matrix. Map the potential slip area degree matrix to a pseudo-color space of [0-255] to complete the calculation and visualization of the potential slip area degree of the rock excavation surface.

[0076] The expression of the potential slip region degree index after normalization of each pixel point in step S601 is:

[0077]

[0078] Among them, ω r ' is the potential slip area degree index of the r-th pixel after normalization; ω r is the potential slip area degree index of the r-th pixel; ω min is the minimum value of the potential slip area degree index of each pixel; ω max It is the maximum value of the potential slip area degree index of each pixel.

Claims

1. A method for calculating the potential area of ​​slippage of a rock excavation face, characterized in that: The following steps are involved: S1. Obtaining crack images of the excavation surface; S2. Establish a rectangular coordinate system based on the edge of the crack image of the excavation surface, and obtain the starting point coordinates and end point coordinates of all crack traces; S3. Setting a distance threshold, and obtaining key nodes of each fracture trace based on the starting point coordinates and the end point coordinates of each fracture trace based on the distance threshold, and linearizing each fracture trace based on the key nodes of each fracture trace to obtain a linearized fracture image of the excavation surface; S4, performing skeleton extraction on the linearized image of the excavation surface crack to obtain a trace skeleton image; Step S5, sequentially traversing the pixel points of the trace skeleton image, and obtaining a potential slip region degree index of the pixel point based on the product of the length and slope of the straightened segment of the crack trace within the grid area centered on the pixel point according to a preset grid width; the specific steps of step S5 are as follows: S501, setting grid width; S502, obtaining the current pixel point according to the trace skeleton image; S503, dividing the grid with the current pixel as the center according to the grid width to obtain the grid area of ​​the current pixel; S504: Obtain a straightened line segment of the crack trace within the grid area of ​​the current pixel to obtain the crack line segment of the current pixel; S505: Obtain the number of pixels of the crack segment of the current pixel point, and obtain the length of the crack segment of the current pixel point; S506, obtaining the slope of the crack line segment at the current pixel point; S507: Obtain a potential slip region degree index of the current pixel point according to the product of the length of the crack line segment of the current pixel point and the slope of the crack line segment of the current pixel point; S508, returning to step S502, obtaining the potential slip area degree index of the next pixel point; S6. Normalize the potential slip area degree index of each pixel point to obtain a potential slip area degree matrix, and map the potential slip area degree matrix to a pseudo-color space to complete the calculation and visualization of the potential slip area degree of the rock excavation surface.

2. The method for calculating the potential area of ​​slippage of a rock excavation face according to claim 1, characterized in that: The step S3 is specifically as follows: S301, setting a distance threshold; S302, obtaining the starting point coordinates and the ending point coordinates of the current crack trace; S303, obtaining a straight line function of the straight line where the starting point coordinates and the end point coordinates are located; S304, calculating the distance between all pixel points on the current crack trace and the straight line function based on the straight line function of the starting point coordinates and the end point coordinates; S305: Based on the distance between each pixel point on the current crack trace and the straight line function, the pixel point with the largest distance is selected as a candidate key node point; S306: Determine whether the distance between the key node candidate point and the straight line function is less than the distance threshold. If so, directly use the end point of the current crack trace as the key node. Otherwise, use the key node candidate point closest to the starting point of the current crack trace as the key node, and use the key node as the new starting point of the current crack trace. Return to step S303 until the end point of the current crack trace is output as the key node. S307. Connect the starting point of the current crack trace and each key node in sequence to complete the linearization of the current crack trace, and return to step S302 to perform the linearization of the next crack trace until the linearization of all crack traces on the excavation surface crack image is completed to obtain the linearized image of the excavation surface crack.

3. The method for calculating the potential area of ​​slippage of a rock excavation face according to claim 1, characterized in that: The expression of the potential slip region degree index of the current pixel point in step S507 is: in, is the potential slip area degree indicator of the current pixel; The current pixel The penalty coefficient of the crack line segment; The crack segment number for the current pixel; The current pixel The length of the crack line segment; The current pixel The slope of the crack line segment; is the absolute value; and These are specific values ​​of penalty coefficients determined based on actual project conditions.

4. The method for calculating the potential area of ​​slippage of a rock excavation face according to claim 1, characterized in that: The S6 is specifically: S601, normalizing the potential slip region degree index of each pixel point to obtain the normalized potential slip region degree index of each pixel point; S602. Establish an empty matrix based on the size of the trace skeleton image, and arrange the normalized potential slip area degree index of each pixel based on the pixel coordinates to obtain a potential slip area degree matrix. Map the potential slip area degree matrix to a pseudo-color space of [0-255] to complete the calculation and visualization of the potential slip area degree of the rock excavation surface.

5. The method for calculating the potential area of ​​slippage of a rock excavation face according to claim 4, characterized in that: The expression of the potential slip region degree index after normalization of each pixel point in step S601 is: in, For the The potential slip area degree index after normalization of pixel points; For the The potential slip area degree index of each pixel; is the minimum value of the potential slip area degree index of each pixel; It is the maximum value of the potential slip area degree index of each pixel.

Images