Dual-mode-based half-pore rate calculation method, system, device, medium and product

Abstract

The present disclosure belongs to the technical field of image recognition, and provides a semi-hole rate calculation method, system, device, medium and product based on a dual mode, the method comprising: collecting original image data and original point cloud data of a blasting section, and obtaining a mapping relationship from point cloud to image through spatial registration; based on the original point cloud data and the mapping relationship obtained through spatial registration, a semi-hole index map is obtained; and based on the semi-hole index map, the blasting semi-hole rate is calculated. Through cross verification of three-dimensional geometry and two-dimensional texture, the probability of misjudging a natural pit or broken area of a rock mass as a semi-hole is reduced, and the reliability of the semi-hole rate evaluation result is improved.

Description

Technical Field

[0001] This disclosure belongs to the field of image recognition technology, and in particular relates to a method, system, device, medium and product for calculating half porosity based on dual-modality. Background Technology

[0002] Timely calculation of the half-hole ratio after blasting is crucial for optimizing blasting parameters and evaluating blasting quality, and to a certain extent reflects the quality of engineering excavation. Generally speaking, the higher the half-hole ratio, the less over-crushing, under-crushing, and irregularities there are, and the higher the excavation quality.

[0003] When calculating the blasting half-hole ratio, on-site manual measurement has the problem of low detection efficiency.

[0004] Image recognition technology based on RGB images of rock surfaces may suffer from problems such as the semi-hole marks being confused with the surrounding rock texture, lighting and shadows, and dust, resulting in unclear identification.

[0005] The method of intelligent calculation based on the three-dimensional point cloud of the rock surface may misjudge data such as natural unevenness, joint surface, and slight damage from the previous blasting as half-hole traces.

[0006] Therefore, it is necessary to develop an image recognition method that can accurately identify blasted half-holes. Summary of the Invention

[0007] To address the aforementioned issues, this disclosure provides a bimodal method for calculating half-porosity. By cross-validating three-dimensional geometry and two-dimensional texture, it reduces the probability of misclassifying natural pits or fractured areas in rock masses as half-porosity, thereby improving the reliability of the half-porosity assessment results.

[0008] In a first aspect, this disclosure provides a method for calculating semi-porosity based on a dual-mode approach, including: The original image data and original point cloud data of the blasting section are collected, and the mapping relationship from point cloud to image is obtained through spatial registration; based on the mapping relationship between the original point cloud data and spatial registration, the half-hole index map is obtained; and the blasting half-hole rate is calculated based on the half-hole index map.

[0009] Furthermore, The original image data and original point cloud data of the blasting section were collected, and the mapping relationship from point cloud to image was obtained through spatial registration, including: The lidar and camera are jointly calibrated; raw image data and raw point cloud data of the cross section are collected and preprocessed; for each point in the preprocessed raw point cloud data, rigid body transformation and perspective projection are performed in sequence, and the points are mapped to the image pixel coordinate system for spatial registration.

[0010] Furthermore, Based on the mapping relationship between the original point cloud data and spatial registration, a semi-aperture index map is obtained, including: Based on the original point cloud data, three-dimensional geometric feature analysis is performed to identify potential depressions; cluster analysis is performed on the set of potential depressions to generate candidate regions for identification; and the original point cloud data within the candidate regions is used to obtain a half-aperture index map on the image plane according to the spatial registration mapping relationship.

[0011] Furthermore, The original point cloud within the identified candidate region is used to obtain a half-aperture index map on the image plane based on the spatial registration mapping relationship, including: The original point cloud within the identified candidate region is projected onto the two-dimensional image pixel coordinate system according to the spatial registration mapping relationship; the half-aperture confidence index is obtained on the image plane; and a half-aperture index map is drawn based on the obtained half-aperture confidence index.

[0012] Furthermore, The calculation of the blasting half-hole ratio based on the half-hole index diagram includes: The SAM model is used to perform fine-grained hole wall segmentation on the candidate target region; based on the fine-grained hole wall segmentation, the effective half-holes are determined; based on the number of effective half-holes, the overall half-hole ratio of the cross-section is calculated.

[0013] Furthermore, Based on refined hole wall segmentation, the effective half-hole is determined, including: For each target region, it must simultaneously satisfy both 3D geometric verification and 2D texture continuity verification to be considered a valid half-hole.

[0014] Secondly, based on the same inventive concept, this disclosure also provides a dual-mode semi-porosity calculation system, including an acquisition and configuration module, a semi-porosity index diagram generation module, and a semi-porosity calculation module. The acquisition and configuration module is used to acquire the original image data and original point cloud data of the blasting section, and obtain the mapping relationship from point cloud to image through spatial registration. The half-aperture index map generation module is used to obtain the half-aperture index map based on the mapping relationship between the original point cloud data and spatial registration. The half-hole ratio calculation module is used to calculate the blasting half-hole ratio based on the half-hole index diagram.

[0015] Thirdly, based on the same inventive concept, this disclosure also provides an electronic device, including at least one processor and at least one memory electrically connected; The memory is electrically connected to the processor, wherein the memory stores instructions executable by at least one of the processors, the instructions being executed by at least one of the processors to enable at least one of the processors to perform the bimodal-based semi-porosity calculation method as described above.

[0016] Fourthly, based on the same inventive concept, this disclosure also provides a computer storage medium. The computer storage medium stores a computer program. When the computer program is executed by the processor, it implements the bimodal-based semi-porous ratio calculation method as described above.

[0017] Fifthly, based on the same inventive concept, this disclosure also provides a computer program product. The computer program product is stored in at least one storage medium; The computer program product includes several instructions for causing at least one electronic device to execute the bimodal-based half-porosity calculation method as described above.

[0018] Compared with existing technologies, this disclosure provides a method for calculating semi-porosity based on a dual-mode approach, which has the following advantages: In the embodiments of this disclosure, the cross-verification of three-dimensional geometry and two-dimensional texture reduces the probability of misjudging natural pits or fractured areas of rock mass as half-holes, thereby improving the reliability of the half-hole rate assessment results.

[0019] Other features and advantages of this disclosure will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the disclosure. The objects and other advantages of this disclosure may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart illustrating a method for calculating semi-porosity based on a dual-mode according to an embodiment of the present disclosure is shown. Figure 2 A schematic diagram illustrating the structural principle of an electronic device according to an embodiment of this disclosure is shown. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0023] In the embodiments of this disclosure, the design requirement for half-holes is that after smooth blasting or pre-splitting blasting, the visible borehole wall traces remaining on the excavation profile should be clear and complete. The axial length of the remaining wall trace of a single half-hole should generally not be less than a set proportion of the original design depth of the borehole. The half-hole ratio refers to the ratio of the sum of the lengths of the half-hole traces remaining on the rock wall on the acceptance section after smooth blasting or pre-splitting blasting (i.e., the cumulative axial length of the remaining wall traces of each borehole) to the sum of the original drilling depths of all boreholes on that section.

[0024] Figure 1 A flowchart illustrating a method for calculating semi-porosity based on a dual-mode approach according to an embodiment of this disclosure is shown.

[0025] like Figure 1 As shown in the figure, an embodiment of the present disclosure provides a method for calculating semi-porosity based on dual-mode operation, which includes the following steps: S1: Collect the original image data and original point cloud data of the blasting section, and obtain the mapping relationship from point cloud to image through spatial registration.

[0026] In this embodiment, LiDAR and camera are used to acquire blast cross-section data. Based on the pre-calibrated extrinsic parameters (rotation matrix R, translation vector L) between the LiDAR and camera and the camera intrinsic parameters (matrix K), coordinate transformation and perspective projection are performed on the synchronously acquired 3D point cloud of the blast cross-section, mapping it onto the corresponding distortion-corrected 2D image to obtain a spatially registered multimodal blast cross-section dataset.

[0027] Specifically, it includes: S11 performs joint calibration of the lidar and camera.

[0028] A checkerboard-like planar calibration board was used, and multiple images were acquired by the camera at different positions and angles to calibrate the camera intrinsic parameter matrix K and distortion coefficient vector D. Then, images of the calibration board and laser point cloud data were simultaneously acquired at multiple spatial positions and angles. The 3D corner coordinates on the calibration board plane were extracted from the point cloud data, and the 2D pixel coordinates of the same corner points were extracted from the corresponding image data. The rigid transformation parameters (rotation matrix R, translation vector L) from the LiDAR to the camera coordinate system were obtained using a Perspective-n-Point (PnP) problem-solving algorithm.

[0029] S12: Collect the original image data and original point cloud data of the cross section and perform preprocessing.

[0030] The camera and lidar are synchronously triggered by hardware to acquire grayscale or color raw image data and raw point cloud data of the blasting section at the same time, and then preprocess them.

[0031] Specifically: statistical or radius filtering is performed on the original point cloud data to remove outlier noise points, and then the point cloud of the target cross-section region is extracted based on distance or intensity threshold segmentation; the distortion correction of the original image is performed using the camera intrinsic parameter matrix K and distortion coefficient D to obtain a distortion-free cross-section image.

[0032] S13. For each point in the preprocessed original point cloud data, perform rigid body transformation and perspective projection in sequence, map it to the image pixel coordinate system, and perform spatial registration.

[0033] (1) Map the point cloud from the lidar coordinate system to the camera coordinate system through rigid body transformation: P m =( x m , y m , z m ) T , P c =( x c , y c , z c ) T , P c =R P m +L, Among them, P m Represents a point in the lidar coordinate system; x m , y m , z m Represents the coordinates of each point in the cross-sectional point cloud data in the lidar coordinate system; T represents matrix transpose; P c Represents a point in the camera coordinate system; x c , y c , z c L represents the coordinates of each point in the cross-sectional point cloud data in the camera coordinate system; L represents the translation vector.

[0034] (2) Perform perspective projection on the data after rigid body transformation and map it to the image pixel coordinate system: a. Convert the data after rigid body transformation into normalized coordinate data: , Among them, P n These are homogeneous coordinates on the normalized plane.

[0035] b. Normalized coordinate data is used to generate registration pixel coordinate data through the intrinsic parameter matrix K. : , in,( x n = , y n = ) represents the coordinates of each point in the cross-sectional point cloud data in the image pixel coordinate system.

[0036] S2, based on the mapping relationship between the original point cloud data and spatial registration, obtains the semi-aperture index map.

[0037] S21. Based on the original point cloud data, perform three-dimensional geometric feature analysis to determine potential depression points.

[0038] For each point in the cross-sectional point cloud data acquired by lidar, a local fitting plane is first constructed using its neighborhood point set. Then, the normal vector, curvature, and signed distance relative to the local fitting plane are calculated for that point. If this distance is significantly greater than the average level of the distance distribution within the neighborhood, the point is marked as a potential depression point.

[0039] S22, perform cluster analysis on the set of potential depression points to generate candidate regions for judgment.

[0040] A region growing algorithm based on normal vectors and curvature constraints is used to divide potential depressions into multiple point cloud clusters. Furthermore, a fitted cylindrical model is used for screening to identify regions that conform to the semi-hole morphological characteristics as candidate regions.

[0041] (1) Based on the potential concave point set, sort the points by curvature values ​​from smallest to largest and select the initial point set.

[0042] (2) For each point in the initial point set, search for its neighboring points. If the angle between the normal vector of a neighboring point and the normal vector of the initial point is less than the smoothing threshold, then include it in the point set of the current region. If the curvature of the current neighboring point is less than the curvature threshold, then add it to the initial point set to expand the set range. After completing the judgment, remove the current point from the initial point set, and repeat the above process until the initial point set is empty. Save the point set of the current region as a candidate region. Select a new initial point from the remaining unprocessed points and proceed to the next iteration.

[0043] (3) After dividing the potential depression point set into multiple candidate regions, extract key features that can describe its shape from each candidate region, such as point cloud coordinates, normal vector, curvature, color / reflectivity.

[0044] (4) Cylindrical Model Fitting and Verification: A random sampling consensus algorithm is used to fit cylindrical models in each candidate region, and their conformity to the physical characteristics of a semi-aperture is evaluated. A semi-aperture typically appears as a concave surface in the shape of a semi-cylindrical cone, with a regular shape, extending inward from the rock wall. The judgment criteria are as follows: (1) The error between the fitted cylinder radius and the designed borehole diameter is less than the set threshold (e.g., 15%). (2) The deviation between the cylinder axis direction and the designed drilling direction is less than the set threshold (e.g., 30°). (3) Cylinder depth > set threshold three (e.g., 1 / 2 of the aperture); (4) The point cloud in the region has high continuity and no obvious breaks.

[0045] S23, the original point cloud in the determined candidate region is used to obtain the half-aperture index map on the image plane according to the spatial registration mapping relationship.

[0046] (1) Project the original point cloud in the candidate region into the two-dimensional image pixel coordinate system according to the spatial registration mapping relationship.

[0047] (2) Obtain the half-hole confidence index on the image plane.

[0048] Calculate the candidate region for each decision. k goodness of fit ρ k (i.e., the proportion of interior points fitted to the cylindrical model) and average depth (Average effective depth along the cylinder axis). Define the candidate region for judgment. k Half-pore confidence index α k The weighted combination of the two: , in, ; d min ,d max Candidate regions for determination k The boundary value of the depth statistical range.

[0049] (3) Draw a half-hole index diagram based on the half-hole confidence index.

[0050] Half-hole confidence index α k Normalizing to the [0,1] interval yields , where H and W are the height and width of the candidate region to be determined, respectively.

[0051] Draw a semi-pore index diagram based on M.

[0052] The half-aperture index image is a grayscale image. Each pixel value in the image represents the confidence level that the location belongs to the half-aperture region. Highlighted areas indicate the presence of high-confidence half-apertures, and the color intensity intuitively reflects the overall score.

[0053] S3, calculate the blasting half-hole rate based on the half-hole index diagram.

[0054] S31, the target region in the cross-sectional image is extracted using the half-aperture index diagram.

[0055] Set the confidence threshold for the half-hole index plot. The half-aperture index map is thresholded to obtain a binary mask, and then each independent candidate target region is extracted through connected component analysis.

[0056] S32 uses the SAM model to perform fine-grained hole wall segmentation on the candidate target region.

[0057] For each candidate target region, its minimum axis-aligned bounding box or its centroid coordinates in the image are used as spatial cues input into the SAM (Segment Anything Model). The minimum axis-aligned bounding box is the smallest rectangle that can completely enclose the candidate target region and whose edges are parallel to the horizontal / vertical coordinate axes of the image.

[0058] At the same time, set text prompts (such as "continuous, smooth cylindrical borehole retains hole wall").

[0059] The SAM model receives spatial and textual cues, performs refined foreground and background segmentation in the neighborhood of candidate target regions, and outputs a binary mask image corresponding to each target region. White pixels (with a value of 1) represent areas where the borehole wall is retained, and black pixels (with a value of 0) represent non-borehole wall areas, such as broken rock surfaces, fissures, or noise.

[0060] The purpose of a binary mask is to describe the visible outline of a borehole on the preserved rock surface, effectively distinguishing between a continuous, smooth borehole wall and irregular fractured areas such as broken rock surfaces and fissures.

[0061] S33, based on refined hole wall segmentation, determines the effective half-hole.

[0062] For each candidate target region, dual verification is performed using both 3D geometry and 2D texture: (1) Three-dimensional geometric conditions: Completed in step S2, determining whether the candidate region presents a cylindrical depression that meets the design specifications, including four indicators: radius error, axial direction deviation, effective depth and point cloud continuity.

[0063] (2) Two-dimensional texture continuity condition: Analyze the binary mask image output by SAM. Wherein: a. Edge Closure and Smoothness Analysis: Extract the edge contours from the binary mask image and calculate the convex hull integrity and standard deviation of curvature variation. Convex hull integrity is defined as the ratio of the contour perimeter to the convex hull perimeter. A value close to 1 indicates a convex and complete edge contour with no significant depressions or breaks; a value significantly less than 1 indicates missing or concave defects in the edge contour, corresponding to a fractured borehole wall. A smaller standard deviation of curvature variation indicates a small curvature change, suggesting a smooth contour without jagged edges or severe local bending; conversely, a larger standard deviation reflects irregular edges caused by rock wall fracture.

[0064] b. Texture Consistency Analysis: The gray-level variance and local binary pattern histogram entropy are calculated within the binary mask image region. The gray-level variance is defined as the second central moment of the pixel gray-level values ​​within the region, reflecting the dispersion of pixel value distribution. A smaller value indicates that the gray-level changes within the region are smooth and uniform, consistent with the texture characteristics of a complete semi-hole inner wall. The local binary pattern histogram entropy is defined as the Shannon entropy of the local binary pattern histogram, reflecting the complexity and randomness of the texture pattern. A smaller value indicates that the texture pattern is simple and regular, corresponding to a uniform structure of the real semi-hole inner wall with minimal blasting disturbance.

[0065] The system sets a lower limit for convex hull integrity, an upper limit for the standard deviation of curvature change, an upper limit for grayscale variance, and an upper limit for the entropy of the local binary pattern histogram. When the calculated convex hull integrity is greater than the lower limit, the standard deviation of curvature change is less than the upper limit, the grayscale variance is less than the upper limit, and the entropy of the local binary pattern histogram is less than the upper limit, the candidate region is considered to meet the two-dimensional texture conditions and is marked as a valid half-hole. Otherwise, even if the three-dimensional geometric conditions are met, the two-dimensional texture conditions are not met, and the candidate region is not marked as a valid half-hole.

[0066] S34, calculate the overall porosity P of the cross-section based on the effective number of half-holes.

[0067] The half-porosity is used to evaluate the quality of cross-sectional profile control. The larger the P value, the better the quality of cross-sectional profile blasting control.

[0068] , in, L valid The effective depth of the cylinder for each effective half-hole The accumulated value; L total It is the sum of the original borehole depths of all designed blast holes in the cross section.

[0069] Based on the same inventive concept as the method disclosed above, this disclosure also provides a dual-mode semi-porosity calculation system, including an acquisition and configuration module, a semi-porosity index diagram generation module, and a semi-porosity calculation module. The acquisition and configuration module is used to acquire the original image data and original point cloud data of the blasting section, and obtain the mapping relationship from point cloud to image through spatial registration. The half-aperture index map generation module is used to obtain the half-aperture index map based on the mapping relationship between the original point cloud data and spatial registration. The half-hole ratio calculation module is used to calculate the blasting half-hole ratio based on the half-hole index diagram.

[0070] Based on the same inventive concept as the above-disclosed content, this disclosure also provides an electronic device. For example... Figure 2 As shown, the electronic device of this disclosure includes at least one processor and at least one memory (storage medium) electrically connected to each other. The memory is electrically connected to the processor, wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the bimodal-based semi-porosity calculation method as described above.

[0071] It should be noted that the electrical connection between the above-mentioned units does not necessarily mean the connection between lines. The indirect connection method can be applied to the embodiments of this disclosure as long as it achieves the purpose of this disclosure.

[0072] Based on the same inventive concept, this disclosure also provides a computer storage medium storing a computer program, which, when executed by a processor, implements the bimodal-based semi-porosity calculation method as described above.

[0073] Based on the same inventive concept, this disclosure also provides a computer program product stored in at least one storage medium; the computer program product includes several instructions to cause at least one computer device to execute the bimodal-based semi-porosity calculation method as described above.

[0074] Although the present disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for calculating semi-porosity based on dual-mode operation, characterized in that, The method includes: The original image data and original point cloud data of the blasting section were collected, and the mapping relationship from point cloud to image was obtained through spatial registration. Based on the mapping relationship between the original point cloud data and spatial registration, a semi-aperture index map is obtained; The blasting half-hole ratio is calculated based on the half-hole index diagram.

2. The method according to claim 1, characterized in that, The original image data and original point cloud data of the blasting section were collected, and the mapping relationship from point cloud to image was obtained through spatial registration, including: Joint calibration of lidar and camera; Collect raw image data and raw point cloud data of the cross section and perform preprocessing; For each point in the preprocessed raw point cloud data, rigid body transformation and perspective projection are performed sequentially to map it to the image pixel coordinate system for spatial registration.

3. The method according to claim 1, characterized in that, Based on the mapping relationship between the original point cloud data and spatial registration, a semi-aperture index map is obtained, including: Based on the original point cloud data, perform three-dimensional geometric feature analysis to identify potential depressions; Cluster analysis is performed on the set of potential depressions to generate candidate regions for determination; The original point cloud within the identified candidate region is used to obtain a semi-aperture index map on the image plane based on the spatial registration mapping relationship.

4. The method according to claim 3, characterized in that, The original point cloud within the identified candidate region is used to obtain a half-aperture index map on the image plane based on the spatial registration mapping relationship, including: The original point cloud within the identified candidate region is projected onto the two-dimensional image pixel coordinate system according to the spatial registration mapping relationship. On the image plane, obtain the half-aperture confidence index; A half-hole index diagram is plotted based on the obtained half-hole confidence index.

5. The method according to any one of claims 1-4, characterized in that, The calculation of the blasting half-hole ratio based on the half-hole index diagram includes: The SAM model is used to perform fine-grained hole wall segmentation on the candidate target region; Based on refined hole wall segmentation, the effective half-hole is determined; The overall porosity of the cross-section is calculated based on the number of effective half-holes.

6. The method according to claim 5, characterized in that, Based on refined hole wall segmentation, the effective half-hole is determined, including: For each target region, it must simultaneously satisfy both 3D geometric verification and 2D texture continuity verification to be considered a valid half-hole.

7. A semi-porosity calculation system based on dual-mode operation, characterized in that, The system includes a data acquisition and configuration module, a half-pore index diagram generation module, and a half-pore ratio calculation module. The acquisition and configuration module is used to acquire the original image data and original point cloud data of the blasting section, and obtain the mapping relationship from point cloud to image through spatial registration. The half-aperture index map generation module is used to obtain the half-aperture index map based on the mapping relationship between the original point cloud data and spatial registration. The half-hole ratio calculation module is used to calculate the blasting half-hole ratio based on the half-hole index diagram.

8. An electronic device, characterized in that, Includes at least one processor and at least one memory electrically connected; The memory is electrically connected to the processor, wherein the memory stores instructions executable by at least one of the processors, the instructions being executed by at least one of the processors to enable at least one of the processors to perform the bimodal-based semi-porosity calculation method as described in any one of claims 1-6.

9. A computer storage medium, characterized in that, The computer storage medium stores a computer program. When the computer program is executed by the processor, it implements the bimodal semi-porosity calculation method according to any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product is stored in at least one storage medium; The computer program product includes several instructions for causing at least one electronic device to execute the bimodal-based semi-porosity calculation method according to any one of claims 1-6.

Images