Soil-rock mixed slope modeling method based on block stone contour library and pre-segmentation strategy

Abstract

The invention provides an earth-rock mixed slope modeling method based on a block stone contour library and a pre-segmentation strategy, and belongs to the field of earth-rock mixed slope model construction. According to the method, a rock block contour library is constructed through an image processing technology, a slope is segmented into a plurality of block-shaped subspaces according to a pre-segmentation strategy, contours in the rock block contour library are copied into the block-shaped subspaces, and finally a soil-rock mixed slope model is constructed. And a rapid and accurate modeling method is provided for stability evaluation of the earth-rock mixed slope.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of soil and rock mixed slope model construction, and particularly relates to a soil and rock mixed slope modeling method based on a block stone contour library and a pre-segmentation strategy. BACKGROUND

[0002] In the field of soil and rock mixed slope model construction, a fast and accurate modeling method can provide high-quality evaluation results for analyzing the stability of soil and rock mixed slopes, and can provide accurate and effective references for slope prevention and control. The existing technology has several significant deficiencies, and technical innovation is urgently needed to improve the efficiency and accuracy of the modeling method. In the past published journal papers, a variety of modeling methods for soil and rock mixed slopes have been disclosed, but each has its own limitations.

[0003] The article entitled "Two-dimensional natural rockfill material generation method based on Minkowski difference and optimized wavefront method and application" (Liu X R, Du L B, Deng Z Y, et al. Two-dimensional natural rockfill material generation method based on Minkowski difference and optimized wavefront method and application[J]. Chinese Journal of Rock Mechanics and Engineering, 2020, 39(09): 1832-1846.) generates contact stacked block stones inside the specified generation domain by using the elliptical replacement method, but this block stone distribution method does not meet the distribution form of block stones inside the slope.

[0004] The article entitled "Soil and rock mixed body slope stability analysis considering block stone long axis inclination and soil and rock contact surface" (Huang X W, Yao Z S, Wang W, et al. Soil and rock mixed body slope stability analysis considering block stone long axis inclination and soil and rock contact surface[J]. Engineering Science and Technology, 2021, 53(01): 47-59.) establishes a soil and rock mixed slope model by using an elliptical shape to replace block stones. This simplified shape ignores the irregular boundary of the block stone, resulting in the neglect of the interlocking effect between rocks, which often leads to inaccurate evaluation of the stability of the slope.

[0005] The article entitled "Numerical investigation of the influence of rock characteristics on the soil-rock mixture (SRM) slopes stability" (Liu S, Wang H, Xu W, et al. Numerical investigation of the influence of rock characteristics on the soil-rock mixture (SRM) slopes stability[J]. KSCE Journal of Civil Engineering, 2020, 24: 3247-3256.) simplifies the rock block into a regular polygon shape, but the rock blocks in the natural SRM are mostly irregular polyhedrons. However, this simplified form underestimates the contribution of the interlocking effect between rocks to the stability of the slope, and this model may lead to incorrect evaluation results when applied to slope stability analysis.

[0006] The article entitled "A new method for constructing finite difference model of soil-rock mixture slope and its stability analysis" (Lianheng Z, Dongliang H, Shuaihao Z, et al. A new method for constructing finite difference model of soil-rock mixture slope and its stability analysis [J]. International Journal of Rock Mechanics and Mining Sciences, 2021, 138: 104605.) uses bounding boxes (i.e. simplified shapes such as rectangles and circles) to replace complex block stone contours for overlap detection to improve modeling speed, but the bounding box algorithm may produce additional iterative calculations in the modeling overlap detection process to achieve the target block stone content in the target generation domain, resulting in reduced modeling speed.

[0007] In summary, the existing modeling methods of soil-rock mixed slope model generally have the problems of not meeting the natural morphological characteristics of block stone, low modeling efficiency, poor application effect of the model, etc. Therefore, a fast and accurate soil-rock mixed slope modeling method has become a key problem in the field of soil-rock mixed slope model construction. SUMMARY

[0008] The technical problem to be solved by the present application is that the existing modeling methods of soil-rock mixed slope model generally have the problems of not meeting the natural morphological characteristics of block stone, low modeling efficiency, poor application effect of the model, etc. Specifically, the present application proposes a soil-rock mixed slope modeling method based on block stone contour library and pre-segmentation strategy. The method constructs a block stone contour library through image processing technology. At the same time, the slope is segmented into several block subspaces according to the pre-segmentation strategy. Then the contours in the rock block contour library are copied into these block subspaces, and finally the soil-rock mixed slope model is constructed, providing a fast and accurate modeling method for soil-rock mixed slope stability evaluation.

[0009] In order to achieve the purpose of the present application, the present application provides a soil-rock mixed slope modeling method based on block stone contour library and pre-segmentation strategy, comprising the following steps:

[0010] Step 1, the establishment of block stone contour library is completed through block stone contour extraction, morphological feature parameter statistics of block stone contour, block stone contour standardization and classified storage;

[0011] Step 2, the pre-segmentation of the slope area is realized through the acquisition of ellipse generation parameters, the generation of stacked ellipses, ellipse decomposition, von Neumann block generation and merging, and a plurality of block subspaces are obtained;

[0012] Step 3, placing the block stone profile in the block stone profile library obtained in step 1 into the block subspace obtained in step 2 to complete the modeling of the soil and rock mixed slope.

[0013] Preferably, the specific steps of the block stone profile extraction in step 1 include:

[0014] A two-dimensional image of the natural rock block is collected by a camera and recorded as an initial rock block image;

[0015] The initial rock block image is converted into a gray-scale image by using a gray-scale conversion, and the gray-scale image is binarized by setting a gray-scale threshold value, wherein the rock block area is defined as black and the image background area is defined as white;

[0016] Each pixel value in the binarized image is replaced by a weighted average value of its adjacent pixels by using a Gaussian filter, and then a denoising area threshold value is set to eliminate extremely small noises disguised as blocks, and the binarized image after the pixel value replacement and denoising processing is recorded as a rock block image;

[0017] The profile coordinates of the rock block in the rock block image are searched in the binarized pixel matrix by using an edge detection technique, and a block stone original profile line is drawn; then the coordinate points on the block stone original profile line are segmented, i.e. the broken lines between the pixel points are replaced by line segments, so as to reduce the number of coordinate points and realize the extraction of the block stone profile;

[0018] The coordinate points on the profile line of the block stone profile obtained by the above extraction are recorded as control points, and a total of n control points are set, and any one of the control points is recorded as the ith control point, and the coordinates of the ith control point are (x i , y i ), i = 1, 2,..., n; the coordinates of the center point of the block stone profile are recorded as (x0, y0), and the calculation formula is:

[0019]

[0020] Preferably, the pixel value of the black color is 0, the pixel value of the white color is 1, and the denoising area threshold value is set to 1000; the gray-scale threshold value is the best global threshold value of the gray-scale image automatically calculated according to the graythresh function using the Otsu method.

[0021] Preferably, the morphological characteristic parameter statistics of the block stone profile in step 1 are statistics of the following parameters: n control point coordinates (x i , y i ), block stone profile center point coordinates (x0, y0), block stone profile area A, block stone profile equivalent circle diameter D, length L R and length S R, the statistics of the block stone profile axis ratio AR, the angle a between the long axis of the block stone profile and the positive direction of the x axis, wherein, AR = L R / S R .

[0022] Preferably, the process of the block stone profile standardization in step 1 is as follows: first, the center point coordinate (x0, y0) of the block stone profile is moved to the coordinate axis origin (0, 0) position to obtain the translated control point coordinate (x i -x0, y i -y0); then the control point is rotated with the angle a as the rotation angle, so that the long axis of the block stone profile is parallel to the x axis; secondly, the translated control point coordinate is scaled according to the scaling ratio η = 1 / D, and the new control point coordinate is marked as the final control point coordinate (x i , y i ), and the calculation formula is as follows:

[0023]

[0024] The classification storage in step 1 is that the standardized block stone profile is classified and stored according to the block stone profile axis ratio AR.

[0025] Preferably, the process of obtaining the ellipse generation parameters in step 2 is as follows: first, the following parameters are counted on site: the volume content μ of the block stone inside the slope, the slope area A slope , the gradation number G of the block stone, the proportion w j of each gradation, and the equivalent circle diameter D j corresponding to the particle size of each gradation, wherein j is the serial number of the gradation, j = 1, 2, 3,..., G; secondly, the ellipse is introduced, and the long axis of the ellipse is made consistent with the direction of the long axis of the block stone profile, and the following ellipse generation parameters are calculated: the number N of the ellipses, the length L e of the long axis of the ellipses, and the axis ratio AR of the ellipses, and the calculation formulas are as follows:

[0026] S e = AR / L e ;

[0027] In the formula, S e is the length of the short axis of the ellipse;

[0028] The generation process of the stacked ellipse in step 2 is as follows: first, the boundary of the slope region is divided into independent Back group and Front group, the Front group is the bottom layer of ellipse plus the right boundary, and the Back group is the left boundary plus the current layer of ellipse being generated; a new ellipse is placed using the Front_1 ellipse and the Back_end ellipse as the boundary, the new ellipse is tangent to the two boundaries, the Front_1 and Back_end ellipses are the first and last ellipses of the Front group and the Back group respectively; the new ellipse is incorporated into the end of the Back group, and the Front_1 ellipse is deleted, generating a new Back_end ellipse and Front_1 ellipse to participate in the placement of the next ellipse, and the ellipses in the Front group are gradually reduced in this process; when the Front group is only left with the right boundary, the stacking of the current layer of ellipse is completed, and the right boundary and the layer of ellipse form a new Front group to participate in the stacking of the next layer of ellipse; if the new ellipse intersects with other ellipses, the intersection between the ellipses in the search field circle and the new ellipse is determined according to the intersection; the search field circle has a radius of wherein L new , L Front(1) , L Back(end) are the lengths of the major axis of the new ellipse, the Front_1 ellipse and the Back_end ellipse respectively; until the number of stacked ellipses reaches the number N of ellipses, the generation of the stacked ellipses in the slope region is completed.

[0029] The decomposition of the ellipse in step 2 is to decompose the stacked ellipse in the slope region into an inscribed circle, specifically, first draw an inscribed circle Q0 with the center of the ellipse as the center and 1 / 2S e as the radius, the two intersection points of the inscribed circle Q0 and the x-axis are (x1, 0) and (-x1, 0), then draw decomposition circles Q1 and Q e with (x1, 0) and (-x1, 0) as the centers and 1 / 2S -1 as the radius, respectively, wherein one of the intersection points of the decomposition circle Q1 and the x-axis is (x2, 0), and one of the intersection points of the decomposition circle Q -1 and the x-axis is (-x2, 0), |x2| > |x1|; then draw decomposition circles Q2 and Q -2 with (x2, 0) and (-x2, 0) as the centers and 1 / 2S e as the radius, respectively, and so on, 1 ellipse is decomposed into 2Y+1 decomposition circles along the positive / negative direction of the x-axis;

[0030] The von Noe block generation and merging in step 2 is: based on the center coordinates of 2Y+1 decomposition circles in an ellipse, and taking the radius of the 2Y+1 decomposition circles as the weight, each ellipse is converted into 2Y+1 gapless weighted voronoi blocks, and the weighted voronoi block is recorded as the sub-attribute voronoi block of the ellipse; the sub-attribute voronoi blocks of the stacked ellipses in the slope region are merged by connecting the common control points of the adjacent two sub-attribute voronoi blocks of the ellipses, the slope space is divided into a plurality of block sub-spaces, and the pre-segmentation of the slope region is realized.

[0031] Preferably, the implementation process of step 3 is as follows:

[0032] Firstly, the following parameters are obtained one by one by statistics: the axial ratio AR of the block sub-space r , the included angle β with the x-axis direction, the area A s , and the contour control point coordinates (x s , y s ) of the block sub-space and other characteristic parameters;

[0033] Secondly, the block stone contour with similar AR to AR r is selected from the block stone contour library, and finally placed in the block sub-space after rotation and scaling, and when all the block stone contours in the block sub-space are placed, the modeling of the soil and rock mixed slope is realized.

[0034] Preferably, the scaling ratio of the block stone contour is After the block stone contour is placed in the block sub-space, the contour control point coordinates are updated to (x inew , y inew ), and the calculation formula is:

[0035]

[0036] Wherein, k is the number of contour control points of the block sub-space, s is the contour control point serial number of the block sub-space, s=1, 2, 3,..., k.

[0037] Compared with the prior art, the beneficial effects of the present application are as follows:

[0038] 1. The block stone contour library constructed by the present application extracts the natural contour of the block stone through image processing technology, retains the natural characteristics of the block stone contour, and provides the block stone contour conforming to the actual situation for constructing the soil and rock mixed slope.

[0039] 2. The pre-segmentation strategy used in the present application includes the acquisition of ellipse generation parameters, the generation of stacked ellipses, ellipse decomposition, von Noe block generation and merging, and the slope is divided into a plurality of sub-spaces through the pre-segmentation strategy.

[0040] The generation parameter of the ellipse is obtained by adopting the statistical parameters of the morphological characteristics of the block stones in the slope, and the generation parameter of the ellipse is obtained by transformation, so that the model can match the rock block volume content, diameter, inclination, axis ratio and other predefined parameters with high precision, so that the statistical characteristics of the rock block can be more comprehensively considered.

[0041] The generation process of the stacked ellipse adopts indirect replacement of complex rock block contour for overlap detection, which not only retains the natural shape characteristics of the rock block, but also avoids the high calculation requirement of the algorithm for complex contour, significantly improves the construction efficiency of the soil and stone mixed slope model. Meanwhile, compared with the traditional bounding box (circular or rectangular simplified shape) method, the ellipse is closer to the natural rock shape, which can ensure efficient overlap detection while more truly reflecting the geometric characteristics of the rock block. In addition, the overlap search circle is constructed in the ellipse stacking process, which reduces the overlap detection range and optimizes the calculation performance.

[0042] 3. The profile in the block stone profile library is placed in the pre-segmented block space by rotation and scaling, and the volume content of the block stone in the constructed soil and stone mixed slope model is accurately controlled.

[0043] 4. The present application considers the morphological statistical characteristics of the block stone, so that the soil and stone mixed slope model with special rock block distribution (such as directional arrangement) can be constructed, which breaks through the limitation of the conventional model which only depends on the volume content, and has wider applicability. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 The flow chart for the soil and stone mixed slope model construction.

[0045] Figure 2 The schematic diagram for block stone profile extraction.

[0046] Figure 3 The schematic diagram for block stone morphological characteristic parameters.

[0047] Figure 4 The schematic diagram for the classification and storage process of the block stone profile library.

[0048] Figure 5 The schematic diagram for the generation process of the stacked ellipse.

[0049] Figure 6 The schematic diagram for the generation result of the ellipse.

[0050] Figure 7 The schematic diagram for the ellipse decomposition.

[0051] Figure 8 The schematic diagram for the generation and merging of Von Neumann block.

[0052] Figure 9 The schematic diagram for the placement process of the block stone profile.

[0053] Figure 10 Fig. 1 is a schematic diagram of modeling results of a soil and rock mixed slope. DETAILED DESCRIPTION

[0054] The application will be described in detail below with reference to the accompanying drawings.

[0055] The application provides a soil and rock mixed slope modeling method based on a block stone contour library and a pre-segmentation strategy, and a modeling process thereof is shown in Fig. 2. Figure 1 The method constructs a block stone contour library through image processing technology, and simultaneously segments the slope into a plurality of block subspaces according to a pre-segmentation strategy; then, the contours in the block stone contour library are copied into the block subspaces, and finally a soil and rock mixed slope model is constructed, specifically, comprising the following steps.

[0056] Step 1: Establishing a block stone contour library through block stone contour extraction, block stone contour morphological feature parameter statistics, block stone contour standardization and classified storage.

[0057] In this embodiment, the specific steps of block stone contour extraction include:

[0058] Collecting a two-dimensional image of a natural rock block by a digital camera, and recording it as an initial rock block image;

[0059] Converting the initial rock block image into a gray-scale image by using a gray-scale conversion, and setting a gray-scale threshold to binarize the gray-scale image, wherein a rock block region is defined as black and an image background region is defined as white;

[0060] Replacing each pixel value in the binarized image with a weighted average value of adjacent pixels by using a Gaussian filter, then setting a denoising area threshold to eliminate extremely small noises disguised as blocks, and recording the binarized image after pixel value replacement and denoising processing as a rock block image;

[0061] Searching for contour coordinates of the rock block in the rock block image in the binarized pixel matrix by using an edge detection technology, and drawing a block stone original contour line; then segmenting the coordinate points on the block stone original contour line, i.e. replacing the broken line between pixel points with a line segment, so as to reduce the number of coordinate points and realize block stone contour extraction;

[0062] Recording the coordinate points on the contour line of the block stone contour obtained through the above extraction as control points, and setting a total of n control points, recording any one control point as the ith control point, and the coordinates of the ith control point are (x i , y i ), i=1, 2,..., n; recording the coordinates of the center point of the block stone contour as (x0, y0), and the calculation formula is:

[0063]

[0064] The pixel value of black is 0, the pixel value of white is 1, and the denoising area threshold is set to 1000; the gray threshold is the optimal global threshold of the gray image automatically calculated by the Otsu method according to the graythresh function.

[0065] The morphological feature parameters of the block stone profile are statistics of the following parameters: n control point coordinates (x i ,y i ), block stone profile center point coordinates (x0, y0), block stone profile area A, block stone profile equivalent circle diameter D, length L R and the length S R of the short axis of the block stone profile axis ratio AR, the included angle a of the long axis of the block stone profile and the positive direction of the x axis, wherein, AR=L R / S R .

[0066] The standardization process of the block stone profile is as follows: first, move the center point coordinates (x0, y0) of the block stone profile to the coordinate axis origin (0, 0) position, and mark the translated control point coordinates as (x i -x0, y i -y0); then rotate the control point with the included angle a as the rotation angle, so that the long axis of the block stone profile is parallel to the x axis; secondly, scale the translated control point coordinates according to the scaling ratio η = 1 / D, and mark the new control point coordinates as the final control point coordinates (x′ i , y′ i ), and the calculation formula is:

[0067]

[0068] The classification storage is to classify and store the standardized block stone profile according to the block stone profile axis ratio AR.

[0069] Figure 2 is a block stone profile extraction schematic diagram, Figure 3 is a block stone morphological feature parameter schematic diagram, Figure 4 is a block stone profile library classification storage process schematic diagram.

[0070] Step 2, the pre-segmentation of the slope area is realized by obtaining the ellipse generation parameters, generating the stacked ellipse, ellipse decomposition, von Neumann block generation and merging, and a plurality of block subspaces are obtained.

[0071] In this embodiment, the process of obtaining the ellipse generation parameters is as follows: first, the following parameters are counted on site: the volume content μ of the block stone inside the slope, the slope area A slope , the block stone grading number G, and the proportion wj , the equivalent circle diameter D corresponding to the size of each graded block stone j , where j is the serial number of the grading, j = 1, 2, 3,..., G; secondly, an ellipse is introduced, and the major axis of the ellipse is consistent with the direction of the block stone profile major axis, and the following ellipse generation parameters are calculated: the number of ellipses N, the length of the ellipse major axis L e and the axis ratio AR of the ellipse, and their calculation formulas are respectively:

[0072]

[0073] In the formula, S e is the length of the ellipse minor axis.

[0074] In this embodiment, the ellipse generation parameters are shown in Table 1. Wherein, A slope = 105.27m 2 .

[0075] Table 1 Block stone shape characteristic parameter statistical table

[0076]

[0077] The generation process of the stacked ellipse is as follows: first, the boundary of the slope area is divided into independent Back group and Front group, the Front group is the bottom layer of the ellipse plus the right boundary, and the Back group is the left boundary plus the current layer of the ellipse being generated; the new ellipse is placed using the Front_1 ellipse and the Back_end ellipse as the boundary, the new ellipse is tangent to the two boundaries, and the Front_1 and Back_end ellipses are the first and last ellipses of the Front group and the Back group respectively; the new ellipse is added to the end of the Back group, and the Front_1 ellipse is deleted at the same time, to generate a new Back_end ellipse and Front_1 ellipse to participate in the placement of the next ellipse, and the Front group ellipses are gradually reduced in this process; when the Front group is left with only the right boundary, the ellipse stacking of the current layer is completed, and the layer of ellipses and the right boundary form a new Front group to participate in the stacking of the next layer of ellipses; if the generated new ellipse intersects with other ellipses, the intersection between the ellipse in the search field circle and the new ellipse is determined according to the intersection; the search field circle radius wherein L new , L Front(1) , L Back(end) are the lengths of the major axes of the new ellipse, the Front_1 ellipse and the Back_end ellipse respectively; until the number of stacked ellipses reaches the number of ellipses N, the generation of the stacked ellipses in the slope area is completed.

[0078] The ellipse decomposition is to decompose the stacked ellipses in the slope area into inscribed circles, specifically, first taking the center of the ellipse as the center of the circle, 1 / 2S eDraw an inscribed circle Q0 with a radius of 1 / 2S. The two intersection points of the inscribed circle Q0 and the x-axis are (x1,0) and (-X1,0). Then, take (x1,0) and (-x1,0) as the center and 1 / 2S as the center. e Draw the decomposition circle Q1 and decomposition circle Q for the radius -1 , where one of the intersection points of the decomposition circle Q1 and the x-axis is (x2,0), and the decomposition circle Q -1 One of the intersection points with the x-axis is (-x2,0), |x2|>|x1|; then take (x2,0) and (-x2,0) as the center, 1 / 2S e Draw decomposition circle Q2 and decomposition circle Q for radius -2 , and so on, decompose 1 ellipse into 2Y+1 decomposition circles along the positive / negative direction of the x-axis.

[0079] The von Neumann blocks are generated and merged as follows: based on the center coordinates of 2Y+1 decomposition circles in an ellipse and with the radius of the 2Y+1 decomposition circles as the weight, each ellipse is converted into 2Y+1 gapless weighted voronoi blocks, and the weighted voronoi blocks are recorded as the sub-Voronoi blocks of the ellipse; by connecting the common control points of the sub-Voronoi blocks of two adjacent ellipses, the sub-Voronoi blocks of the stacked ellipses in the slope area are merged, and the slope space is divided into multiple block subspaces, thereby realizing the pre-segmentation of the slope area.

[0080] Figure 5 Schematic diagram of the generation process of stacked ellipses. Figure 6 This is a schematic diagram of the generated ellipse. Figure 7 is a schematic diagram of ellipse decomposition, Figure 8 Schematic diagram of von Neumann block generation and merging. Figure 8 In the equation, A and B are the common control points of two adjacent elliptical sub-Voronoi blocks, and a, b, c, and d are non-common control points. The slope space is divided into several block-shaped subspaces by connecting the common control points.

[0081] Step 3: Place the rock contours in the rock contour library obtained in step 1 into the block subspace obtained in step 2 to complete the soil-rock mixed slope modeling.

[0082] In this embodiment, the implementation process of step 3 is as follows:

[0083] First, the following parameters are obtained one by one by statistics: the axis ratio AR of the block subspace r , angle β with the x-axis, area A s and the coordinates of the contour control points of the block subspace (x s ,y s ) and other characteristic parameters;

[0084] Secondly, select AR and AR r Similar block stone profile, through rotation, scaling and finally placed in the block subspace, when all block stone profile placed in the block subspace, realize the modeling of the soil and rock mixed slope.

[0085] The scaling ratio of the block stone profile

[0086] After the block stone profile is placed in the block subspace, the profile control point coordinates are updated to (x inew , y inew ), and the calculation formula is:

[0087]

[0088] Where k is the number of profile control points of the block subspace, s is the profile control point number of the block subspace, s = 1, 2, 3,..., k.

[0089] Figure 9 The block stone profile placement process is shown in the figure, Figure 10 The modeling result of the soil and rock mixed slope is shown in the figure.

Claims

1. A soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy, characterized in that: The following steps are involved: Step 1: Establishing a stone contour database by extracting stone contours, counting morphological characteristic parameters of stone contours, standardizing stone contours, and classifying and storing them; Step 2: Pre-segment the slope area by obtaining ellipse generation parameters, generating stacked ellipses, decomposing ellipses, generating and merging von Neumann blocks to obtain multiple block subspaces. Step 3: Place the rock contours in the rock contour library obtained in step 1 into the block subspace obtained in step 2 to complete the soil-rock mixed slope modeling.

2. The soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy according to claim 1 is characterized in that: The specific steps of extracting the stone outline in step 1 include: A two-dimensional image of the natural rock block is collected by a camera and recorded as an initial image of the rock block; Converting the initial rock image into a grayscale image by grayscale conversion, and setting a grayscale threshold to binarize the grayscale image, wherein the rock area is defined as black and the image background area is defined as white; Use Gaussian filtering to replace each pixel value in the binary image with the weighted average of its adjacent pixels. Then set a denoising area threshold to eliminate the tiny noise that disguises itself as a block. The binary image after pixel value replacement and denoising is recorded as a rock block image. Edge detection technology is used to search for the outline coordinates of the rock block in the binary pixel matrix and draw the original outline of the rock block. Then, the coordinate points on the original outline of the rock block are segmented, that is, line segments are used instead of broken lines between pixel points to reduce the number of coordinate points and realize the rock block outline extraction. The coordinate points on the contour line of the block stone obtained by the above extraction are recorded as control points. Suppose there are n control points in total, and any one of them is recorded as the i-th control point. The coordinates of the i-th control point are (x i ,y i ), i = 1, 2, ..., n; the coordinates of the center point of the stone outline are marked as (x0, y0), and the calculation formula is:

3. The soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy according to claim 2 is characterized in that: The black pixel value is 0, the white pixel value is 1, and the denoising area threshold is set to 1000; the grayscale threshold is the optimal global threshold of the grayscale image automatically calculated using the Otsu method based on the graythresh function.

4. The soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy according to claim 2 is characterized in that: The morphological characteristic parameter statistics of the block stone outline in step 1 are the statistics of the following parameters: n control point coordinates (x i ,y i ), the coordinates of the center point of the stone outline (x0, y0), the area of ​​the stone outline A, the equivalent circle diameter D of the stone outline, and the length of the long axis of the stone outline L R The length of the minor axis S R , the statistics of the block outline axis ratio AR, the angle a between the block outline long axis and the positive direction of the x-axis: Among them, AR=L R / S R .

5. The soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy according to claim 4 is characterized in that: The process of standardizing the stone contour in step 1 is as follows: First, the center point coordinates (x0, y0) of the stone contour are moved to the origin of the coordinate axis (0, 0), and the coordinates of the control point after translation (x i -x0,y i -y0); then rotate the control point with the included angle a as the rotation angle so that the long axis of the stone outline is parallel to the x-axis; then scale the coordinates of the control point after translation according to the scaling ratio η = 1 / D, and mark the new control point coordinates as the final control point coordinates (x′ i , y′ i ), which is calculated as follows: The classified storage in step 1 is to classify and store the standardized block stone contours according to the block stone contour axis ratio AR.

6. The soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy according to claim 5, characterized in that: The process of obtaining the ellipse generation parameters in step 2 is as follows: First, the following parameters are counted on site: the volume content of the rocks inside the slope μ, the slope area A slope , the number of gradations of the block stone G, the proportion of each gradation w j , the equivalent circle diameter D corresponding to the particle size of each level of block stone j , where j is the gradation number, j = 1, 2, 3, ..., G; secondly, introduce the ellipse, and make the major axis of the ellipse consistent with the major axis of the stone outline, and calculate the following ellipse generation parameters: the number of ellipses N, the length of the major axis of the ellipse L e and the axis ratio AR of the ellipse, which are calculated as follows: S e =AR / L e ; Where S e is the length of the minor axis of the ellipse; The generation process of the stacked ellipses described in step 2 is as follows: first, the boundary of the slope area is divided into mutually independent Back and Front groups. The Front group is the bottom layer of ellipses plus the right boundary, and the Back group is the left boundary plus the current layer of ellipses being generated. A new ellipse is placed using the Front_1 ellipse and the Back_end ellipse as boundaries. The new ellipse is tangent to the two boundaries. The Front_1 and Back_end ellipses are the first and last ellipses of the Front and Back groups, respectively. The new ellipse is merged into the end of the Back group, and the Front_1 ellipse is deleted. A new Back_end ellipse is generated and participates in the placement of the next ellipse with the Front_1 ellipse. During this process, the number of ellipses in the Front group gradually decreases. When only the right boundary is left in the Front group, the stacking of the ellipses in the current layer is completed. The ellipse in this layer and the right boundary form a new Front group and participate in the stacking of the next layer of ellipses. If the generated new ellipse intersects with other ellipses, the intersection determination is performed based on the ellipse within the intersection search area circle and the new ellipse; The radius of the search area circle Among them L new , L Front(1) , L Back(end) are the major axis lengths of the new ellipse, the Front_1 ellipse, and the Back_end ellipse, respectively; until the number of stacked ellipses reaches the number of ellipses N, the generation of stacked ellipses in the slope area is completed; The decomposition of the ellipse in step 2 is to decompose the ellipse stacked in the slope area into an inscribed circle. Specifically, first, take the center of the ellipse as the center of the circle, 1 / 2S e Draw an inscribed circle Q0 with a radius of 1 / 2S. The two intersection points of the inscribed circle Q0 and the x-axis are (x1, 0) and (-x1, 0). Then, take (x1, 0) and (-x1, 0) as the center and 1 / 2S as the center. e Draw the decomposition circle Q1 and decomposition circle Q for the radius -1 , where one of the intersection points of the decomposition circle Q1 and the x-axis is (x2,0), and the decomposition circle Q -1 One of the intersection points with the x-axis is (-x2,0), |x2|>|x1|; then take (x2,0) and (-x2,0) as the center, 1 / 2S e Draw decomposition circle Q2 and decomposition circle Q for radius -2 , and so on, decompose 1 ellipse into 2Y decomposition circles along the positive / negative direction of the x-axis; The generation and merging of the von Neumann blocks described in step 2 are as follows: based on the center coordinates of the 2Y+1 decomposition circles in an ellipse and with the radius of the 2Y+1 decomposition circles as the weight, each ellipse is converted into 2Y+1 gapless weighted Voronoi blocks, and the weighted Voronoi blocks are recorded as the sub-Voronoi blocks of the ellipse; by connecting the common control points of the sub-Voronoi blocks of two adjacent ellipses, the sub-Voronoi blocks of the stacked ellipses in the slope area are merged, and the slope space is divided into multiple block subspaces to achieve pre-segmentation of the slope area.

7. The soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy according to claim 6, characterized in that: The implementation process of step 3 is as follows: First, the following parameters are obtained one by one by statistics: the axis ratio AR of the block subspace r , angle β with the x-axis, area A s and the coordinates of the contour control points of the block subspace (x s ,y s ) and other characteristic parameters; Next, select AR and AR from the block stone contour library. r Similar rock contours are rotated and scaled and finally placed in the block subspace. When all the rock contours in the block subspace are placed, the soil-rock mixed slope modeling is achieved.

8. The soil-rock mixed slope modeling method based on a rock contour library and a pre-segmentation strategy according to claim 7 is characterized in that: Scaling ratio of the stone outline After the stone contour is placed in the block subspace, the coordinates of the contour control points are updated to (xinew, yinew), which is calculated as follows: Wherein, k is the number of contour control points of the block subspace, s is the sequence number of the contour control points of the block subspace, s=1, 2, 3, ..., k.

Images