A tunnel excavation section overbreak and underbreak detection method, electronic equipment and storage medium

Abstract

This invention provides a method for detecting over- and under-excavation in tunnel excavation sections, comprising: acquiring images of the tunnel excavation section; inputting the design section outline and the excavated section image into an over- and under-excavation detection program; and performing image preprocessing, image segmentation, extraction of the tunnel design section outline, extraction of the tunnel excavated section outline, classification of the design section outline shape, detection of straight lines and calculation of curves in the design section outline, division of curves in the excavated section outline, calculation of over- and under-excavation values, and calculation of the over- and under-excavation area of ​​the excavated section through image processing. This invention obtains the tunnel excavation section outline coordinates and over- and under-excavation data simply by processing the images, which is faster, more economical, and more objective than manual detection. Furthermore, compared with other traditional methods for calculating over- and under-excavation in tunnel excavation sections, this detection method has a shorter image acquisition time, is simpler to operate, and is faster and more convenient for over- and under-excavation calculation.

Description

Technical Field

[0001] This invention relates to the field of civil engineering monitoring technology, and in particular to a method, electronic device and storage medium for detecting over-excavation and under-excavation of tunnel excavation sections based on image recognition. Background Technology

[0002] Tunnel outlines are generally categorized into circular, arched, and rectangular shapes, and thus typically consist of arcs or straight lines. In tunnel engineering, when measuring the over-excavation and under-excavation values ​​of a blasting excavation section, a ruler is generally used to measure the distance from the I-beam perpendicular to the tunnel's designed cross-section to the blasting excavation face. Therefore, for rectangular tunnels, the over-excavation and under-excavation values ​​should be the difference between the blasting excavation face and the tunnel's centerline. For arched tunnels, the over-excavation and under-excavation values ​​can be calculated for both straight and arc sections. The over-excavation and under-excavation values ​​for straight sections can be calculated according to the standards for rectangular tunnels, while the over-excavation and under-excavation values ​​for arc sections should be the difference between the blasting excavation face and the radius of the arc.

[0003] Currently, the over- or under-excavation of tunnel cross sections is generally measured using measuring tapes, total stations, cross-section scanners, or 3D laser scanners. Among them: (1) measuring tapes have low accuracy; (2) total stations have high accuracy but fewer measurement points; (3) cross-section scanners have high accuracy and dense measurement points, but the operation is more complicated and the calculation is prone to errors, and the instruments are more expensive; (4) 3D laser scanners have high accuracy and dense measurement points, but the acquisition time is long, the operation is complicated, and the instruments are expensive.

[0004] In the process of calculating the over-excavation and under-excavation values ​​of tunnel cross sections, the centerline point or mass point of the tunnel cross section is generally used as the reference point. The difference between the distance from the excavated cross section point to the reference point and the distance from the designed cross section to the reference point is calculated. The calculated over-excavation and under-excavation values ​​have certain errors compared with the actual over-excavation and under-excavation values.

[0005] During tunnel construction, the required accuracy for over-excavation and under-excavation of the tunnel cross-section is at the centimeter level. Therefore, the instrument suitable for measuring over-excavation and under-excavation on-site is a cross-section scanner. However, its calculation method for over-excavation and under-excavation has certain problems, resulting in low accuracy. Summary of the Invention

[0006] This invention provides a method for detecting over-excavation and under-excavation in tunnel excavation sections, comprising the following steps:

[0007] Step 1: Acquire images of the tunnel excavation cross-section;

[0008] Step 2: Locate the tunnel design cross-section corresponding to the pile segment in the tunnel excavation cross-section drawing from the tunnel design drawings, and modify and mark the outline in the tunnel design cross-section drawing to obtain the tunnel design cross-section outline drawing; then compare the tunnel design cross-section drawing with the tunnel excavation cross-section. Figure 1 And input it into the over-excavation and under-excavation detection program;

[0009] Step 3: Preprocess the tunnel design cross-section and tunnel excavation cross-section in the over-excavation detection program:

[0010] Bilateral filtering is used to blur and reduce noise in tunnel excavation cross-section diagrams.

[0011] The pixel dimensions of the tunnel excavation cross-section and the tunnel design cross-section were calculated using pixel calculation and size comparison methods, respectively.

[0012] Mark the target locations in the tunnel excavation cross-section diagram and the known engineering coordinate points in the tunnel design cross-section diagram;

[0013] Input the engineering coordinates of the target point location corresponding to the tunnel excavation cross-section and the engineering coordinates of the tunnel design cross-section, and match the tunnel excavation cross-section and the tunnel design cross-section according to the engineering coordinates of the two.

[0014] Step 4: Use the HSV color space image segmentation algorithm to process the tunnel excavation cross-section and the tunnel design cross-section to obtain the binarized image of the tunnel excavation cross-section outline, the binarized image of the tunnel design cross-section outline, and the binarized image of the circular outline of the connection point of the tunnel design cross-section outline curve.

[0015] Step 5: Extraction of tunnel design cross-sectional profile and tunnel excavation cross-sectional profile:

[0016] Based on the binarized images of the tunnel design cross-section profile and the circular profile, the tunnel design cross-section profile and the circular profile are extracted to obtain the design cross-section profile with a line width of one pixel, the circular profile with a line width of one pixel, and the number of curve nodes.

[0017] Extracting the profile of the excavation section based on the binarized image of the tunnel excavation section profile, and extracting the profile of the tunnel design section based on the binarized image of the tunnel design section profile and the binarized image of the circular profile;

[0018] Step 6: Classify the contour shapes in the design cross-section contour drawing and determine the number of arcs that make up the tunnel contour based on the number of circular contours.

[0019] Step 7: Perform straight line detection and contour curve calculation on the tunnel design cross-section outline.

[0020] Step 8: Divide the contour curves of the tunnel excavation cross-section outline diagram;

[0021] Step 9: Calculate the over-excavation and under-excavation values ​​for the cross-section;

[0022] Step 10: Calculate the over-excavation and under-excavation areas. The specific method is as follows: Based on the actual size of the pixels, calculate the real coordinates of the excavation cross-section outline pixels and the design cross-section outline pixels, and draw them on the same image according to the corresponding coordinates. Set the real size of this image to 5*5mm. Considering that there may be discontinuous outline pixels due to the original real size being smaller than the real size of the merged image, interpolation is used to supplement the tunnel outline pixels, thereby forming a continuous binary image of the tunnel design cross-section outline and the tunnel excavation cross-section outline. Then, the closed outline is found according to the connected component algorithm, the number of closed outline pixels is counted, and the area of ​​the closed outline is calculated. Then, the over-excavation and under-excavation values ​​are calculated for any point on the closed outline to determine whether the closed outline is over-excavated or under-excavated. The over-excavation area and under-excavation area of ​​the tunnel cross-section are statistically analyzed.

[0023] Optionally, the specific process of processing the tunnel excavation cross-section and the tunnel design cross-section using a color segmentation algorithm in step four is as follows:

[0024] S4.1 Convert the RGB values ​​of the image pixels of the tunnel excavation cross-section and the tunnel design cross-section to HSV values, and set the range of H, S, and V values ​​of the color of the outline.

[0025] S4.2 Extract the pixels corresponding to the H, S, and V values ​​from the tunnel excavation cross-section and tunnel design cross-section, and assign the corresponding pixels to 0, which is black; assign pixels in other color ranges to 255, which is white.

[0026] S4.3 Extract the color gamut of red and blue pixels from the tunnel design section image to obtain a binarized image of the tunnel outline design and a binarized image of the circular outline connecting points of the tunnel design section outline curves; and extract the color gamut of red pixels from the tunnel excavation section image to obtain a binarized image of the tunnel excavation section outline.

[0027] Optionally, the specific process of extracting the tunnel design cross-section profile and circular profile based on the binary image of the tunnel design cross-section profile and the binary image of the circular profile in step five is as follows:

[0028] S5.1 First, a 3×3 sliding pane is constructed to traverse the binarized image. The eight-neighborhood algorithm is used to detect the surrounding neighborhood of pixels with a pixel value of 0, determine the contour boundary and remove single pixels. Then, the TWO-PASS algorithm is used to mark the connected components to determine the connectivity between contours, obtain the design section contour boundary map and the circular contour boundary map, and determine the number of curve nodes N according to the number of marked circular contours.

[0029] S5.2. The contour boundary map and the circular contour boundary map are iteratively processed using the erosion algorithm and opening operation until the pixel width of the contour boundary map and the circular contour boundary map is a single pixel. The zero-order moment and the first-order moment of the circular contour are calculated by counting the number of pixels and pixel coordinates in the contour, thereby obtaining the centroid coordinates of the design section contour and the centroid coordinates of the circular contour.

[0030] Optionally, the specific process of classifying the contour shapes in the design cross-sectional contour drawing in step six is ​​as follows:

[0031] S6.1 Construct a cross-sectional profile database for tunnels of four types and calculate the Hu invariant moment for different cross-sectional types. Specifically, the different cross-sectional types are divided into four categories: circular or arched, rectangular, horseshoe-shaped, and elliptical. Among them: the profile of a circular or arched tunnel consists of a single circle or multiple arcs, without straight segments; the profile of a rectangular tunnel consists only of straight segments; the profile of a horseshoe-shaped tunnel consists of straight segments and arc segments, with arc segments only in the upper or middle part of the profile; the profile of an elliptical tunnel consists of straight segments and arc segments, with straight segments only in the middle part of the profile.

[0032] S6.2 Calculate the Hu invariant moment of the design cross-section profile, find the tunnel cross-section type with the smallest difference in Hu invariant moment value in the profile cross-section database, and the design cross-section type is this cross-section type.

[0033] Optionally, the specific method for detecting straight lines in the design section contour map in step seven is as follows: detect straight lines in the design section contour map by using the cumulative probability Hough transform algorithm, and then output the coordinates of two nodes of the straight line segment to determine the range of the straight line segment. Construct a set of pixel points for each straight line segment and a set of curve segments based on the pixel points within the straight line segment. Then, calculate the straight line formula based on the coordinates of the endpoints of the straight line, and construct the corresponding parallel straight line at the centroid of the design section.

[0034] Optionally, the specific method for calculating the contour curve of the design cross-section in step seven is as follows:

[0035] S7.1. Let the coordinates of the centroid of the circular profile be the curve nodes;

[0036] S7.2. Further divide the set of curve segments according to the curve construction rules to obtain the pixel points of each curve segment;

[0037] S7.3. Using the coordinates of the first side endpoint (x1, y1), the other side endpoint (x2, y2), and the midpoint (x3, y3) of the curve, calculate the center coordinates (x0, y0) of the arc curve and the corresponding arc pixel radius r0; the specific formula is as follows:

[0038] In the formula, (x1,y1) are the column and row numbers of the endpoints on one side of the curve, (x1,y1) are the column and row numbers of the endpoints on the other side of the curve, (x3,y3) are the column and row numbers of the center of the curve, (x0,y0) are the column and row numbers of the center of the arc corresponding to the curve, and r0 is the pixel radius of the arc corresponding to the curve.

[0039] Optionally, the specific process of dividing the contour curve of the tunnel excavation cross-section in step eight is as follows:

[0040] S8.1 Traverse the binary image of the tunnel excavation section outline, extract the black pixels with a pixel value of 255, obtain the row and column number of each black pixel, and then calculate the coordinate value of each point.

[0041] S8.2. Based on the design contour curve and the coordinates of the endpoints of the straight lines, the set of contour line segments of the excavation section is delineated according to the boundary division criteria. The division starts from the lower left endpoint of the tunnel contour, and the following rules are established:

[0042] ① First, divide the set of pixels of the excavation cross-section outline of the straight line segment, and then divide the set of pixels of the excavation cross-section outline of the curved line segment.

[0043] ② For the division of the range of a straight line segment, the intersection of the straight line perpendicular to the node and the outline of the excavation section is generally calculated. Two nodes will form two intersection points on both sides, and the range formed by the intersection points is the set of pixel points of the straight line segment.

[0044] ③ Considering that there are two intersecting straight line segments in the tunnel outline, for over-excavation, the division is generally done according to rule ①. The remaining part of the two straight line segments is divided at the intersection node and then independently regarded as the set of that node. For under-excavation, the intersection point of the two straight line segments, the intersection point of the straight line with the slope of the vertical line segment length / horizontal line length and the excavation section outline is calculated. The intersection point and the intersection point of the other node of the straight line segment according to rule ② form the set of pixel points of the straight line segment.

[0045] ④ For the range of the curve segment, the intersection of the straight line between the curve segment node and the center of the circle and the outline of the excavation section is generally calculated. The two nodes of the curve segment form two intersection points on both sides. The range formed by the intersection points is the set of pixels of the curve segment.

[0046] ⑤ Considering that there is a straight segment and a curved segment in the tunnel outline, according to rule ①, the range of the straight segment is first divided, and the range of the curved segment is the set of the intersection points of the straight segment and the excavation section outline and the intersection points of the curved segment and the excavation section outline calculated by rule ④.

[0047] ⑥ Considering that there are two intersecting curve segments in the tunnel outline, the intersection point of the previous curve segment with the excavation section outline and the intersection point of the other node of this curve segment with the excavation section outline calculated according to rule ④ are combined to form the pixel set of the excavation section outline curve segment of this curve segment.

[0048] Optionally, the specific process for calculating the over-excavation and under-excavation values ​​of the cross-section in step nine is as follows:

[0049] S9.1. Based on the tunnel excavation cross-section contour pixels of the straight segments, curved segments, and nodes divided in step eight, calculate the over-excavation and under-excavation values ​​for different ranges. The specific calculation formula is as follows:

[0050] For a straight segment, the over-excavation / under-excavation value d = d2 - d1, where: d1 is the distance between the straight line and the parallel line passing through the centroid of the design section, and d2 is the perpendicular distance between a point and the parallel line passing through the centroid of the design section.

[0051] For a curve segment, the over- or under-excavation value d = r - r0; for a node, the over-excavation value d = [(x - x1)]. 2 +(y-y1) 2 ] 0.5 ;

[0052] S9.2 Determine whether there is over-excavation in the tunnel cross-section. If d is greater than 0, it is over-excavation; if d is less than 0, it is under-excavation. Wherein:

[0053]

[0054] In the formula, α is the actual pixel size in cm; l0 is the distance between the designed straight line segment and the parallel straight line passing through the centroid of the designed cross-section. i r0 is the perpendicular distance from a pixel on the excavation section outline to a parallel straight line passing through the centroid of the designed section; r0 is the radius of the arc corresponding to the curve segment. i x is the distance from a pixel on the excavation cross-section outline to the center of the arc corresponding to the curve segment; i ,y i Let m and n be the column and row numbers of a pixel in the excavation cross-section outline, where m and n are the column and row numbers of the node.

[0055] The present invention also provides an electronic device for use in the tunnel excavation section over-excavation and under-excavation detection method as described above, the electronic device comprising one or more processors;

[0056] Storage device for storing one or more programs;

[0057] When the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method.

[0058] The present invention also provides a storage medium for use in an electronic device as described above, which stores a computer program that implements all steps of the over-excavation and under-excavation calculation method.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] This invention provides a method for detecting over-excavation and under-excavation in tunnel excavation sections. By processing images, the tunnel cross-section contour coordinates and over-excavation / under-excavation data can be obtained, which is faster, more economical, and more objective than manual inspection. Furthermore, compared to other traditional over-excavation / under-excavation calculation methods, this detection method has a shorter image acquisition time, is simpler to operate, and the over-excavation / under-excavation calculation process is faster and more convenient.

[0061] In addition to the objectives, features, and advantages described above, the present invention has other objectives, features, and advantages. The invention will now be described in further detail with reference to the figures. Attached Figure Description

[0062] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0063] Figure 1 This is a schematic diagram of the overall process of a method for detecting over-excavation and under-excavation of a tunnel excavation section in this embodiment;

[0064] Figure 2 This is a schematic diagram of the node and endpoint calculation process in this embodiment;

[0065] Figure 3 This is a schematic diagram of the straight segment range division under the under-excavation situation in this embodiment;

[0066] Figure 4 This is a schematic diagram of the straight segment range division under the over-excavation situation in this embodiment;

[0067] Figure 5 This is a schematic diagram illustrating the division of straight and curved segments in the case of under-excavation in this embodiment;

[0068] Figure 6 This is a schematic diagram illustrating the division of straight and curved segments in the case of over-excavation in this embodiment;

[0069] Figure 7 This is a schematic diagram illustrating the division of the curve segment range under the under-excavation condition in this embodiment;

[0070] Figure 8 This is a schematic diagram illustrating the division of the curve segment range in the case of over-excavation in this embodiment;

[0071] Figure 9 This is a schematic diagram of the tunnel design cross-section in this embodiment;

[0072] Figure 10 This is a cross-sectional outline of the design with a line width of one pixel in this embodiment;

[0073] Figure 11 This is a schematic diagram of the tunnel excavation cross-section with enhanced color contrast in this embodiment;

[0074] Figure 12 This is a binarized image of the tunnel design cross-section profile in this embodiment;

[0075] Figure 13 This is a cross-sectional outline of the tunnel excavation in this embodiment;

[0076] Figure 14 This is a list of pixel row and column numbers for the cross-sectional outline diagram in this embodiment;

[0077] Figure 15 This is a list of pixel row and column numbers for the excavation outline in this embodiment;

[0078] Figure 16 This is a schematic diagram of node coordinate calculation in this embodiment;

[0079] Figure 17 This is a schematic diagram of the calculation of arc parameters in this embodiment;

[0080] Figure 18 This is the list of over-mining and under-mining values ​​in this embodiment;

[0081] Figure 19 This is a schematic diagram of image overlay calculation in this embodiment;

[0082] Figure 20 This is a schematic diagram of over-excavation area detection in this embodiment;

[0083] Figure 21 This is a schematic diagram illustrating the calculation of the over-excavated area in this embodiment;

[0084] Figure 22 This is a schematic diagram of the under-excavation area detection in this embodiment;

[0085] Figure 23 This is a schematic diagram of the calculation of the area of ​​the under-excavated region in this embodiment. Detailed Implementation

[0086] The following are specific embodiments of the present invention, which are described in conjunction with the accompanying drawings. However, the present invention is not limited to these embodiments.

[0087] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used herein in the specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0088] This embodiment:

[0089] The present invention provides a method for detecting over-excavation and under-excavation in tunnel excavation sections, see [link to relevant documentation]. Figure 1 As shown, it includes the following steps:

[0090] Step 1: Collect tunnel excavation cross-section diagrams.

[0091] Optionally, the specific process for acquiring the tunnel excavation cross-section diagram in step one is as follows:

[0092] S1.1 Select a cross section within the tunnel and install a light strip transmitter in the middle of the cross section according to the laser ranging module 1 / 2;

[0093] S1.2 Adjust the tripod and knob in the light strip transmitter to center the circular bubble of the light strip transmitter and make the compass direction of the light strip transmitter consistent with the azimuth angle of the tunnel design axis, and mark the ground target point emitted by the laser ranging module 3 (the specific process can refer to the existing technology).

[0094] S1.3. Set up the acquisition bracket along the transmission line of the laser ranging module of the light strip emitter, and adjust the acquisition bracket so that the laser point of the light strip emitter falls into the target of the bracket, and record the distance gap between the tripod and the acquisition bracket.

[0095] S1.4 Place the camera (such as a mobile phone) on the acquisition bracket, and after turning off all light sources in front of the tunnel, take a picture of the tunnel excavation cross-section (see...). Figure 11 (as shown);

[0096] S1.5. Remove the light strip transmitter, acquisition bracket, and imaging equipment from the tunnel;

[0097] S1.6. Use a total station to determine the coordinates of the ground target at the location of the light strip transmitter and the acquisition support.

[0098] Optionally, the light strip emitter is selected as a light strip emitter with a light strip width of 2mm-5mm and a red color.

[0099] Step 2: Locate the tunnel design cross-section drawing that corresponds to the tunnel excavation cross-section drawing from the tunnel design drawings (see...). Figure 9(As shown), the outline color in the tunnel design cross-section drawing is changed to red, and small blue circles are marked at the connection points of the outline curves in the tunnel design cross-section drawing to obtain the modified tunnel design cross-section drawing; the tunnel design cross-section drawing is then compared with the tunnel excavation cross-section. Figure 1 The data is then input into the over- or under-excavation detection program. Preferably, the over- or under-excavation detection program is existing technology.

[0100] Step 3: In the over-excavation and under-excavation detection program, bilateral filtering is used to perform blurring and noise reduction on the tunnel excavation cross-section diagram;

[0101] The pixel dimensions of the tunnel excavation cross-section and the tunnel design cross-section were calculated using pixel calculation and size comparison methods, respectively.

[0102] Mark the target locations in the tunnel excavation cross-section diagram and the known engineering coordinate points in the tunnel design cross-section diagram;

[0103] Input the engineering coordinates of the target point location corresponding to the tunnel excavation cross-section diagram and the engineering coordinates of the tunnel design cross-section image, and match the tunnel excavation cross-section diagram with the tunnel design cross-section diagram based on the two engineering coordinates.

[0104] Optionally, the specific method for calculating the actual pixel size of the tunnel excavation cross-section and the tunnel design cross-section using the size comparison method is as follows: Select any two points in the tunnel excavation cross-section and the tunnel design cross-section respectively, calculate the pixel coordinate difference based on the pixel coordinates of the two points, and the actual size of the pixel point = the actual distance between the two points / the pixel distance difference.

[0105] Step 4: In the over-excavation and under-excavation detection program, the HSV color space image segmentation algorithm is used to process the tunnel excavation cross-section and the tunnel design cross-section to obtain a binary image of the tunnel excavation cross-section contour and a binary image of the tunnel design cross-section contour (see [link to tunnel design cross-section contour binary image]). Figure 12 (as shown) and the binarized diagram of the circular outline connecting points of the tunnel design cross-section contour curve.

[0106] Optionally, the specific process of processing the tunnel excavation cross-section and tunnel design cross-section using the HSV color space image segmentation algorithm is as follows:

[0107] S4.1 Convert the RGB values ​​of the image pixels of the tunnel excavation cross-section and the tunnel design cross-section to HSV values, and set the range of H, S, and V values ​​of the color of the outline.

[0108] S4.2 Extract the pixels corresponding to the H, S, and V values ​​from the tunnel excavation cross-section and tunnel design cross-section, and assign the corresponding pixels to 0, which is black; assign pixels in other color ranges to 255, which is white.

[0109] S4.3 Extract the color gamut of red and blue pixels from the tunnel design section image to obtain the binarized image of the tunnel outline design (i.e., the binarized image of the tunnel design section outline) and the binarized image of the circular outline connecting points of the tunnel design section outline curves; and extract the color gamut of red pixels from the tunnel excavation section image to obtain the binarized image of the tunnel excavation section outline.

[0110] Step 5: Based on the binarized image of the tunnel design cross-section profile and the binarized image of the circular profile, extract the tunnel design cross-section profile and the circular profile to obtain a design cross-section profile image with a line width of one pixel (see...). Figure 10 (as shown), a circular outline with a line width of one pixel and the number of curve nodes;

[0111] The profile of the excavation section is extracted based on the binarized image of the tunnel excavation section profile, and the profile of the tunnel design section is extracted based on the binarized image of the tunnel design section profile and the binarized image of the circular profile.

[0112] Optionally, the specific process for extracting the tunnel design cross-section profile and circular profile based on the binary image of the tunnel design cross-section profile and the binary image of the circular profile is as follows:

[0113] S5.1 First, a 3×3 sliding pane traverses the binarized image. The eight-neighborhood algorithm is used to detect the surrounding neighborhood of pixels with a value of 0, determining the contour boundaries and removing single pixels. Then, the TWO-PASS algorithm is used to label connected components, determining the connectivity between contours, and obtaining the design cross-section contour boundary map and the circular contour boundary map. The number of curve nodes N is determined based on the number of labeled circular contours (see [link to documentation]). Figure 16 (as shown);

[0114] S5.2. The tunnel design section profile boundary map and circular profile boundary map are iteratively processed using the erosion algorithm and opening operation until the pixel width of the profile boundary map and the circular profile boundary map is a single pixel. The zero-order moment and the first-order moment of the circular profile are calculated by statistically analyzing the number of pixels and pixel coordinates within the profile (profile boundary map and circular profile boundary map), thereby obtaining the centroid coordinates of the design section profile and the circular profile.

[0115] Optionally, the specific process for extracting the excavation cross-section contour based on the binarized image of the tunnel excavation cross-section contour is as follows:

[0116] S5.3, see also Figure 2 As shown, the Zhang-Suen algorithm is used to refine the contour line to form a single-pixel width tunnel excavation section contour line. The eight-neighborhood algorithm is used to detect the endpoints and nodes of the contour line, that is, to detect the pixel values ​​of the eight neighborhoods of the center pixel. Black is set to 1 and white to 0. If the neighborhood value is 1, it is an endpoint; if the neighborhood value is 3, it is a node.

[0117] S5.4 Calculate the distance between a node and its two adjacent endpoints, and remove the curve containing the endpoint with the shortest distance;

[0118] S5.5 Calculate the distance between two adjacent endpoints, and connect the endpoints whose distances are within a set threshold range (the threshold range is set according to the actual pixel size, generally set as 10cm / number of actual pixels) with straight lines to form the extracted tunnel excavation cross-sectional outline (see...). Figure 13 (as shown); where: the area connected beyond the threshold indicates severe over- or under-excavation or no relevant area image was collected. Repeat step one to find the corresponding point and take a photo of the area again.

[0119] Step 6: Classify the contour shapes in the design cross-section contour drawing and determine the number of arcs that make up the tunnel contour based on the number of circular contours.

[0120] Optionally, the specific process for classifying the contour shapes in the design cross-sectional contour drawing is as follows:

[0121] S6.1 Construct a database of cross-sectional profiles for tunnels of four types, and calculate the Hu invariant moments for different cross-sectional types (specifically, circular or arched, rectangular, horseshoe-shaped, and elliptical; where: circular or arched tunnel profiles consist of a single circle or multiple arcs with no straight segments; rectangular tunnel profiles consist only of straight segments; horseshoe-shaped tunnel profiles consist of straight segments and arc segments, with arc segments only in the upper or middle part of the profile; and elliptical tunnel profiles consist of straight segments and arc segments, with straight segments only in the middle part of the profile).

[0122] S6.2 Calculate the Hu invariant moment of the design cross-section profile, and find the tunnel cross-section type with the smallest difference in Hu invariant moment value in the profile cross-section database. This cross-section type is the design cross-section type.

[0123] Optionally, the specific method for determining the number of arcs constituting the tunnel outline based on the number of circular outlines is as follows: for tunnels initially judged to be circular or arched and horseshoe-shaped, the number of arcs is N-1; for tunnels initially judged to be elliptical, the number of arcs is N-2.

[0124] Step 7: Perform straight line detection on the design cross-section outline of the tunnel that is determined to be elliptical, and calculate the outline curve of the design cross-section outline.

[0125] Optionally, the specific method for detecting straight lines in the design cross-section contour map is as follows: detect straight lines in the design image contour using the cumulative probability Hough transform algorithm, and then output the coordinates of two nodes of the straight line segment to determine the range of the straight line segment. Construct a set of pixel points for each straight line segment and a set of curve segments using the pixel points within the straight line segment. Then, calculate the straight line formula based on the coordinates of the endpoints of the straight line, and construct the corresponding parallel straight line at the centroid of the design cross-section.

[0126] Optionally, the specific method for calculating the profile curve of the design cross-section is as follows:

[0127] S7.1. Let the coordinates of the centroid of the circular profile be the curve nodes;

[0128] S7.2. Further divide the set of curve segments according to the curve construction rules to obtain the pixel points of each curve segment;

[0129] S7.3. Using the coordinates of the first side endpoint (x1, y1), the other side endpoint (x2, y2), and the midpoint (x3, y3) of the curve, calculate the center coordinates (x0, y0) of the arc curve and the corresponding arc pixel radius r0; the specific formula is as follows:

[0130] In the formula, (x1, y1) are the column and row numbers of one endpoint of the curve, (x1, y1) are the column and row numbers of the other endpoint of the curve, (x3, y3) are the column and row numbers of the curve's center point, (x0, y0) are the column and row numbers of the center point of the arc corresponding to the curve, and r0 is the pixel radius of the arc corresponding to the curve (wherein, for details on the calculation of the arc parameters, please refer to...). Figure 17 As shown, the pixel row and column numbers for the design cross-sectional outline diagram are detailed in [reference needed]. Figure 14 As shown, the pixel row and column numbers of the excavation outline map are detailed in [reference needed]. Figure 15 (As shown).

[0131] Optionally, according to the defined curve construction rules, the curve nodes include:

[0132] ① If a node on one side of the curve and any other remaining node form an interval that contains the coordinates of one or more other nodes, then the two nodes do not form a curve; if the interval does not contain any other nodes, then a curve can be constructed.

[0133] ② Based on the maximum / minimum x and y coordinates in the curve segment set, calculate the tunnel span and tunnel centerline coordinates, thereby dividing it into left / right curves and top / bottom arch curves. Further divide the curve segment pixel set into multiple curve segment pixel point sets.

[0134] Step 8: Divide the contour curves of the tunnel excavation cross-section outline diagram.

[0135] Optionally, the specific process for dividing the contour curve of the tunnel excavation cross-section is as follows:

[0136] S8.1 Traverse the binary image of the tunnel excavation section outline, extract the black pixels with a pixel value of 255, obtain the row and column number of each black pixel, and then calculate the coordinate value of each point.

[0137] S8.2. Based on the design contour curve and the coordinates of the endpoints of the straight lines, the set of contour line segments of the excavation section is delineated according to the boundary division criteria. The division starts from the lower left endpoint of the tunnel contour, and the following rules are established:

[0138] ① First, divide the set of pixels of the excavation cross-section outline of the straight line segment, and then divide the set of pixels of the excavation cross-section outline of the curved line segment.

[0139] ② For the division of the range of a straight line segment, the intersection of the straight line perpendicular to the node and the outline of the excavation section is generally calculated (there are usually two intersection points, take the closest point). The two nodes will form two intersection points on both sides, and the range formed by the intersection points is the set of pixel points of the straight line segment.

[0140] ③ Considering that there are two intersecting straight segments in the tunnel outline, for over-excavation, refer to... Figure 4 As shown, the division is generally done according to rule ①: the remaining parts of two intersecting line segments are divided at their intersection nodes, and then each node is treated as an independent set. For under-excavation cases, refer to... Figure 3 As shown, the intersection of the two line segments and the intersection of the line with the slope of (length of the vertical line segment / length of the horizontal line) with the excavation section outline are calculated. This intersection point and the intersection point of the other node of the line segment in rule ② form the set of line segment pixel points.

[0141] ④ For the range of the curve segment, the intersection of the straight line between the curve segment node and the center of the circle and the outline of the excavation section is generally calculated. The two nodes of the curve segment form two intersection points on both sides. The range formed by the intersection points is the set of pixels of the curve segment.

[0142] ⑤ Considering that there is a straight segment and a curved segment intersecting in the tunnel outline, refer to Figure 5 and Figure 6 As shown, according to rule ①, the range of the straight line segment is first divided, and the range of the curved line segment is the set of the intersection points of the straight line segment and the excavation section outline line and the intersection points of the curved line segment and the excavation section outline line calculated by rule ④.

[0143] ⑥ Considering the situation where two curved segments intersect in the tunnel outline, refer to Figure 7 and Figure 8As shown, the intersection points of the previous curve segment with the excavation section outline and the intersection points of the other side nodes of this curve segment with the excavation section outline calculated according to rule ④ are combined to form the pixel set of the excavation section outline curve segment of this curve segment.

[0144] Step 9: Calculate the over-excavation and under-excavation values ​​for the cross-section. The specific process is as follows:

[0145] S9.1. Based on the tunnel excavation cross-section contour pixels of the straight segments, curved segments, and nodes divided in step eight, calculate the over-excavation and under-excavation values ​​for different ranges. The specific calculation formula is as follows:

[0146] For a straight segment, the over-excavation / under-excavation value d = d2 - d1, where: d1 is the distance between the straight line and a parallel line passing through the centroid of the design section, and d2 is the perpendicular distance between a point and a parallel line passing through the centroid of the design section. For a detailed list of over-excavation / under-excavation values, please refer to [link to list]. Figure 18 As shown;

[0147] For the curved segment, the over- or under-excavation value is d = r - r0;

[0148] For a node, its over-excavation value d = [(x - x1)]. 2 +(y-y1) 2 ] 0.5 ;

[0149] S9.2 Determine whether there is over-excavation in the tunnel cross-section. If d is greater than 0, it is over-excavation; if d is less than 0, it is under-excavation. Wherein:

[0150]

[0151] In the formula, α is the actual pixel size in cm; l0 is the distance between the designed straight line segment and the parallel straight line passing through the centroid of the designed cross-section. i r0 is the perpendicular distance from a pixel on the excavation section outline to a parallel straight line passing through the centroid of the designed section; r0 is the radius of the arc corresponding to the curve segment. i x is the distance from a pixel on the excavation cross-section outline to the center of the arc corresponding to the curve segment; i ,y i Let m and n be the column and row numbers of a pixel in the excavation cross-section outline, where m and n are the column and row numbers of the node.

[0152] In addition, this embodiment also discloses the calculation of over-excavation and under-excavation areas; the specific process is as follows:

[0153] Based on the actual dimensions of the pixels, calculate the true coordinates of the excavation cross-section outline pixels and the design outline pixels, and plot them on the same graph according to their corresponding coordinates, such as... Figure 19As shown, the actual size of the image is set to 5*5mm. Considering the possibility of discontinuous contour pixels due to the original actual size being smaller than the actual size of the merged image, interpolation can be used to supplement the tunnel contour pixels, thus forming a continuous binary image of the tunnel design cross-section contour and the tunnel excavation cross-section contour. Then, a closed contour is found using a connected component algorithm, the number of pixels in the closed contour is counted, and the area of ​​the closed contour is calculated. Then, an over-excavation or under-excavation value is calculated for any point on the closed contour to determine whether the closed contour is over-excavated or under-excavated. The over-excavation area and under-excavation area of ​​the tunnel cross-section are statistically analyzed, such as... Figures 20-23 As shown.

[0154] As a further embodiment of the present invention, the present invention also provides an electronic device and a computer-readable medium.

[0155] The electronic devices included are:

[0156] One or more processors;

[0157] Storage device for storing one or more programs;

[0158] When the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method.

[0159] In practical use, users can interact with servers, which are also electronic devices, via a network to receive or send messages. Terminal devices are generally various electronic devices equipped with a display and used through a human-computer interface, including but not limited to smartphones, tablets, laptops, and desktop computers. Various specific application software can be installed on these terminal devices as needed, including but not limited to web browsers, instant messaging software, social media platforms, and shopping apps.

[0160] Optionally, the server is a network service provider that offers various services, such as a backend server that provides corresponding calculation services for the received excavation cross-section contour images transmitted from the terminal device. This enables automatic contour recognition of the received excavation cross-section contour images and returns the final recognition result to the terminal device.

[0161] Similarly, the computer-readable medium of the present invention stores a computer program thereon, which, when executed by a processor, implements an automatic crack image recognition method according to an embodiment of the present invention.

[0162] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting over-excavation and under-excavation in tunnel excavation sections, characterized in that, Includes the following steps: Step 1: Acquire images of the tunnel excavation cross-section; Step 2: Find the tunnel design cross-section drawing that corresponds to the pile segment in the tunnel excavation cross-section image from the tunnel design drawing, and modify and mark the outline in the tunnel design cross-section drawing; Input both the tunnel design cross-section diagram and the tunnel excavation cross-section diagram into the over-excavation and under-excavation detection program; Step 3: Preprocess the tunnel design cross-section and tunnel excavation cross-section in the over-excavation and under-excavation detection program respectively: Bilateral filtering is used to blur and reduce noise in tunnel excavation cross-section diagrams. The pixel dimensions of the tunnel excavation cross-section image and the tunnel design cross-section image were calculated using pixel calculation method and size comparison method, respectively. Mark the target locations in the tunnel excavation cross-section diagram and the known engineering coordinate points in the tunnel design cross-section diagram; Input the engineering coordinates of the target point location corresponding to the tunnel excavation cross-section and the engineering coordinates of the tunnel design cross-section, and match the tunnel excavation cross-section and the tunnel design cross-section according to the engineering coordinates of the two. Step 4: Use the HSV color space image segmentation algorithm to process the tunnel excavation cross-section and the tunnel design cross-section to obtain the binarized image of the tunnel excavation cross-section outline, the binarized image of the tunnel design cross-section outline, and the binarized image of the circular outline of the connection point of the tunnel design cross-section outline curve. Step 5: Extraction of tunnel design cross-sectional profile and tunnel excavation cross-sectional profile: Based on the binarized images of the tunnel design cross-section profile and the circular profile, the tunnel design cross-section profile and the circular profile are extracted to obtain the design cross-section profile with a line width of one pixel, the design cross-section circular profile with a line width of one pixel, and the number of curve nodes. Extracting the profile of the excavation section based on the binarized image of the tunnel excavation section profile, and extracting the profile of the tunnel design section based on the binarized image of the tunnel design section profile and the binarized image of the circular profile; Step 6: Classify the contour shapes in the design cross-section contour drawing and determine the number of arcs that make up the tunnel contour based on the number of circular contours. Step 7: Perform straight line detection and contour curve calculation on the design cross-section outline of the tunnel. Step 8: Divide the contour curves of the tunnel excavation cross-section outline diagram; Step 9: Calculate the over-excavation and under-excavation values ​​for the cross-section; Step 10: Calculate the over-excavation and under-excavation areas. The specific method is as follows: Based on the actual size of the pixels, calculate the real coordinates of the excavation cross-section outline pixels and the design cross-section outline pixels, and draw them on the same image according to the corresponding coordinates. Set the real size of this image to 5*5mm. Considering that there may be discontinuous outline pixels due to the original real size being smaller than the real size of the merged image, interpolation is used to supplement the tunnel outline pixels, thereby forming a continuous binary image of the tunnel design cross-section outline and the tunnel excavation cross-section outline. Then, the closed outline is found according to the connected component algorithm, the number of closed outline pixels is counted, and the area of ​​the closed outline is calculated. Then, the over-excavation and under-excavation values ​​are calculated for any point on the closed outline to determine whether the closed outline is over-excavated or under-excavated. The over-excavation area and under-excavation area of ​​the tunnel cross-section are statistically analyzed.

2. The method for detecting over-excavation and under-excavation in tunnel excavation sections according to claim 1, characterized in that, The specific process of processing the tunnel excavation cross-section and tunnel design cross-section using a color segmentation algorithm in step four is as follows: S4.1 Convert the RGB values ​​of the image pixels of the tunnel excavation cross-section and the tunnel design cross-section to HSV values, and set the range of H, S, and V values ​​of the color of the outline. S4.2 Extract the pixels corresponding to the H, S, and V values ​​from the tunnel excavation cross-section and tunnel design cross-section, and assign the corresponding pixels to 0, which is black; assign pixels in other color ranges to 255, which is white. S4.3 Extract the color gamut of red and blue pixels from the tunnel design cross-section drawing to obtain a binarized image of the tunnel design cross-section outline and a binarized image of the circular outline connecting points of the tunnel design cross-section outline curves; and extract the color gamut of red pixels from the tunnel excavation cross-section drawing to obtain a binarized image of the tunnel excavation cross-section outline.

3. The method for detecting over-excavation and under-excavation in tunnel excavation sections according to claim 1, characterized in that, The specific process for extracting the tunnel design cross-sectional profile and circular profile in step five, based on the binary image of the tunnel design cross-section profile and the binary image of the circular profile, is as follows: S5.1 First, a 3×3 sliding pane is constructed to traverse the binarized image. The eight-neighborhood algorithm is used to detect the surrounding neighborhood of pixels with a pixel value of 0, determine the contour boundary and remove single pixels. Then, the TWO-PASS algorithm is used to mark the connected components to determine the connectivity between contours, obtain the design section contour boundary map and the circular contour boundary map, and determine the number of curve nodes N according to the number of marked circular contours. S5.

2. The design cross-sectional contour boundary map and circular contour boundary map are iteratively processed using the erosion algorithm and opening operation until the pixel width of the contour boundary map and the circular contour boundary map is a single pixel. The zero-order moment and first-order moment of the circular contour are calculated by statistically analyzing the number of pixels and pixel coordinates within the contour, thereby obtaining the centroid coordinates of the design cross-sectional contour and the circular contour.

4. The method for detecting over-excavation and under-excavation in tunnel excavation sections according to claim 1, characterized in that, The specific process of classifying the contour shapes in the design cross-sectional contour drawing in step six is ​​as follows: S6.1 Construct a database of cross-sectional profiles for four typical tunnel types and calculate the Hu invariant moment for different cross-sectional types. Specifically, the different cross-sectional types are divided into four categories: circular or arched, rectangular, horseshoe-shaped, and elliptical. Among them: the profile of a circular or arched tunnel consists of a single circle or multiple arcs, with no straight segments; the profile of a rectangular tunnel consists only of straight segments; the profile of a horseshoe-shaped tunnel consists of straight segments and arc segments, with arc segments only in the upper or middle part of the profile; the profile of an elliptical tunnel consists of straight segments and arc segments, with straight segments only in the middle part of the profile. S6.2 Calculate the Hu invariant moment of the design cross-section profile, find the tunnel cross-section type with the smallest difference in Hu invariant moment value in the profile cross-section database, and the design cross-section type is this cross-section type.

5. The method for detecting over-excavation and under-excavation in tunnel excavation sections according to claim 1, characterized in that, The specific method for detecting straight lines in the design cross-section contour map in step seven is as follows: the cumulative probability Hough transform algorithm is used to detect straight lines in the design image contour and output the coordinates of two nodes of the straight line segment to determine the range of the straight line segment. The pixel set of each straight line segment and the set of curve segments are constructed based on the pixel points within the straight line segment. Then, the straight line formula is calculated based on the coordinates of the endpoints of the straight line, and the corresponding parallel straight line is constructed at the centroid of the design cross-section.

6. The method for detecting over-excavation and under-excavation in tunnel excavation sections according to claim 1, characterized in that, The specific method for calculating the contour curve of the design cross-section in step seven is as follows: S7.

1. Let the coordinates of the centroid of the circular profile be the curve nodes; S7.

2. Further divide the set of curve segments according to the curve construction rules to obtain the pixel points of each curve segment; S7.

3. Using the coordinates of the first side endpoint (x1, y1), the other side endpoint (x2, y2), and the midpoint (x3, y3) of the curve, calculate the center coordinates (x0, y0) of the arc curve and the corresponding arc pixel radius r0; the specific formula is as follows: In the formula, (x1,y1) are the column and row numbers of the endpoints on one side of the curve, (x1,y1) are the column and row numbers of the endpoints on the other side of the curve, (x3,y3) are the column and row numbers of the center of the curve, (x0,y0) are the column and row numbers of the center of the arc corresponding to the curve, and r0 is the pixel radius of the arc corresponding to the curve.

7. The method for detecting over-excavation and under-excavation in tunnel excavation sections according to claim 1, characterized in that, The specific process of dividing the contour curve of the tunnel excavation cross-section in step eight is as follows: S8.1 Traverse the binary image of the tunnel excavation section outline, extract the black pixels with a pixel value of 255, obtain the row and column number of each black pixel, and then calculate the coordinate value of each point. S8.

2. Based on the design cross-sectional profile curve and the coordinates of the endpoints of the straight lines, the set of excavation cross-sectional profile segments is delineated according to the range division criteria. The division starts from the endpoint at the lower left of the tunnel profile, and the following rules are established: ① First, divide the set of pixels of the excavation cross-section outline of the straight line segment, and then divide the set of pixels of the excavation cross-section outline of the curved line segment. ② For the division of the range of a straight line segment, the intersection of the straight line perpendicular to the node and the outline of the excavation section is generally calculated. Two nodes will form two intersection points on both sides, and the range formed by the intersection points is the set of pixel points of the straight line segment. ③ Considering that there are two intersecting straight line segments in the tunnel outline, for over-excavation, the division is generally done according to rule ①. The remaining part of the two straight line segments is divided at the intersection node and then independently regarded as the set of that node. For under-excavation, the intersection point of the two straight line segments, the intersection point of the straight line with the slope of the vertical line segment length / horizontal line length and the excavation section outline is calculated. The intersection point and the intersection point of the other node of the straight line segment according to rule ② form the set of pixel points of the straight line segment. ④ For the range of the curve segment, the intersection of the straight line between the curve segment node and the center of the circle and the outline of the excavation section is generally calculated. The two nodes of the curve segment form two intersection points on both sides. The range formed by the intersection points is the set of pixels of the curve segment. ⑤ Considering that there is a straight segment and a curved segment in the tunnel outline, according to rule ①, the range of the straight segment is first divided, and the range of the curved segment is the set of the intersection points of the straight segment and the excavation section outline and the intersection points of the curved segment and the excavation section outline calculated by rule ④. ⑥ Considering that there are two intersecting curve segments in the tunnel outline, the intersection point of the previous curve segment with the excavation section outline and the intersection point of the other node of this curve segment with the excavation section outline calculated according to rule ④ are combined to form the pixel set of the excavation section outline curve segment of this curve segment.

8. The method for detecting over-excavation and under-excavation in tunnel excavation sections according to claim 7, characterized in that, The specific process for calculating the over-excavation and under-excavation values ​​of the cross-section in step nine is as follows: S9.

1. Based on the tunnel excavation cross-section contour pixels of the straight segments, curved segments, and nodes divided in step eight, calculate the over-excavation and under-excavation values ​​for different ranges. The specific calculation formula is as follows: For a straight segment, the over-excavation / under-excavation value d = d2 - d1, where: d1 is the distance between the straight line and the parallel line passing through the centroid of the design section, and d2 is the perpendicular distance between a point and the parallel line passing through the centroid of the design section. For a curve segment, the over- or under-excavation value d = r - r0; for a node, the over-excavation value d = [(x - x1)]. 2 +(y-y1) 2 ] 0.5 ; S9.2 Determine whether there is over-excavation in the tunnel cross-section. If d is greater than 0, it is over-excavation; if d is less than 0, it is under-excavation. Wherein: In the formula, α is the actual pixel size in cm; l0 is the distance between the designed straight line segment and the parallel straight line passing through the centroid of the designed cross-section. i r0 is the perpendicular distance from a pixel on the excavation section outline to a parallel straight line passing through the centroid of the designed section; r0 is the radius of the arc corresponding to the curve segment. i x is the distance from a pixel on the excavation cross-section outline to the center of the arc corresponding to the curve segment; i ,y i Let m and n be the column and row numbers of a pixel in the excavation cross-section outline, where m and n are the column and row numbers of the node.

9. An electronic device, characterized in that, When applied to the tunnel excavation section over-excavation and under-excavation detection method as described in any one of claims 1-8, the electronic device includes one or more processors; Storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned method.

10. A storage medium, applied to the electronic device of claim 9, characterized in that, It contains a computer program that implements all the steps of the over-excavation and under-excavation calculation method.

Images