Main draw shaft wall analysis system based on laser point cloud and high-definition image

Abstract

The invention relates to the technical field of data processing, and particularly discloses a main draw shaft wall analysis system based on a laser point cloud and a high-definition image, and the system obtains point cloud data and images through a sensor platform composed of a laser radar and a high-definition camera. Three-dimensional well wall point cloud data obtained through multi-section scanning are spliced into a well wall point cloud model under the same coordinate system through an automatic registration algorithm, automatic recognition and evaluation of well wall defects are achieved by integrating high-definition images and the three-dimensional well wall point cloud data, and after a well wall section needing to be repaired is determined, the well wall section needing to be repaired is repaired. According to the intelligent well wall inspection system and method, defects can be intelligently recognized, the well wall health condition can be automatically evaluated, and the automation level, the safety guarantee capacity and the economic benefits of mine well wall inspection are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, to a main ore pass shaft wall analysis system based on laser point cloud and high-definition image. Background Art

[0002] An ore pass is an important ore transfer system project in underground metal mines. It realizes the continuous sliding of ore from the upper ore bin to the lower ore bin by utilizing the self-weight of the ore, providing an ore transfer passage for one or more mining stages. With the long-term operation of the mining ore pass, the high-speed impact of ore in the ore pass causes repeated spalling, depression, and rib spalling at the connection part between the branch ore pass and the main ore pass and at the vulnerable parts of the shaft wall, seriously threatening the safety of the wellhead ore unloading platform and underground operation. Existing literature (Yin Yue. Research on the movement law of ore and rock blocks and the failure mechanism of shaft wall during ore unloading in ore pass [D]. University of Science and Technology Liaoning, 2020. DOI: 10.26923 / d.cnki.gasgc.2020.000093.) has studied the movement law of ore and rock blocks, the failure characteristics, failure mechanism, and collision distribution range of the shaft wall during the ore unloading of the main ore pass, and given the schematic diagram of the failure area of an ore pass in a certain mine as shown in Figure 2 The traditional assessment of the current situation of shaft wall failure in ore passes mostly relies on manual inspection, two-dimensional photogrammetry, or single ultrasonic and sonar detection technologies, which have limitations such as many inspection blind spots, low efficiency, and difficulty in quantifying the geometric characteristics of diseases, and cannot provide accurate spatial distribution and damaged volume information for targeted support design. In addition, although wet dust suppression and ventilation control can, to a certain extent, inhibit the spillage of dust, they cannot evaluate the degree of damage to the shaft wall structure and are difficult to meet the health diagnosis requirements under the high-intensity and long-term operation of deep inclined ore passes.

[0003] In recent years, multi-source remote sensing technologies based on laser point cloud and high-definition image have shown advantages of high precision, visualization, and quantification in the health detection of underground engineering structures such as tunnels and culverts. LiDAR can quickly obtain the point cloud inside the shaft and construct a high-resolution three-dimensional model; high-definition cameras can record multi-view real scenes, providing intuitive evidence for subsequent defect identification; the fusion of the two can accurately extract parameters such as the spatial position, depth, width, and height of cracks, spalling, and depression, providing data support for disease grading assessment and optimized reinforcement design. However, existing research mostly focuses on shallow or open-cut scenarios, and there is still a lack of research on multi-segment continuous scanning and full-segment visualization diagnosis inside deep inclined ore passes. To solve the above problems, a technical solution is provided now. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a main chute wall analysis system based on laser point cloud and high-definition imaging, which realizes automatic identification and evaluation of shaft wall defects by integrating high-definition imaging and three-dimensional shaft wall point cloud data. It is used to solve the problem that existing research focuses on shallow or open excavation scenarios, and there is still a lack of research on multi-segment continuous scanning and full-segment visual diagnosis inside deep well inclined chutes. It can intelligently identify defects, automatically evaluate the health status of the shaft wall, and improve the automation level of mine shaft wall inspection to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: A main chute shaft wall analysis system based on laser point cloud and high-definition image, including a three-dimensional scanning modeling module, an interval segmentation analysis module and a linkage control mechanism module; the three-dimensional scanning modeling module is used to obtain data and images using a sensor platform composed of a laser radar and a high-definition camera, and use an automatic registration algorithm to splice the three-dimensional shaft wall point cloud data obtained by multiple-segment scanning into a shaft wall point cloud model in the same coordinate system; the interval segmentation analysis module is used to realize automatic recognition and evaluation of shaft wall defects by integrating high-definition images and three-dimensional shaft wall point cloud data; the interval segmentation analysis module includes a first shaft wall defect recognition unit; the first shaft wall defect recognition unit detects shaft wall linear defects on the shaft wall surface through image processing and pattern recognition algorithms, identifies linear defect features based on texture analysis and edge extraction of high-definition images, and filters misjudgments in combination with three-dimensional shaft wall point cloud data; the linkage control mechanism module is used to intelligently control local ventilators and air ducts in the shaft wall section after determining the shaft wall section to be repaired.

[0006] As a further solution of the present invention, the three-dimensional scanning modeling module includes a well section division unit, a sensor platform, a well wall data extraction unit and a well wall data fusion unit; the sensor platform includes a laser radar and a high-definition camera; the well section division unit is used to evenly divide the inclined well to be tested into a number of well sections to be tested; the well wall data extraction unit extracts the first well wall data according to the order of the well sections to be tested from top to bottom; the well wall data fusion unit is used to pre-calibrate the relative posture of the laser radar and the high-definition camera, align the collected first well wall data with the corresponding point cloud coordinate system, obtain three-dimensional well wall point cloud data with texture information, and use an automatic registration algorithm to splice the three-dimensional well wall point cloud data obtained by multi-segment scanning into a well wall point cloud model under a unified coordinate system.

[0007] As a further solution of the present invention, the sensor platform includes a laser radar and a high-definition camera; the sensor platform is driven by a track guide, and the high-definition camera is deployed around the laser radar at a constant speed with a side view to obtain the first wellbore wall data and high-definition images of the well section to be measured.

[0008] As a further solution of the present invention, the first wellbore defect recognition unit detects the linear defects on the wellbore surface through image processing and pattern recognition algorithms, and recognizes the linear defect features based on texture analysis and edge extraction of high-definition images. Specifically: Obtain the high-definition image of the wellbore, and preprocess the high-definition image, including image distortion correction, brightness and contrast enhancement, and denoising processing; Use the image texture feature extraction algorithm on the processed high-definition image to achieve preliminary crack recognition and obtain the crack candidate area; Perform morphological thinning processing on the crack candidate area, extract the skeleton curve of the crack candidate area, and use the connected component labeling method to label different crack areas as independent crack features as the crack recognition result; calculate the skeleton length of the crack feature and the average width of the crack area.

[0009] As a further solution of the present invention, filter misjudgments by combining three-dimensional wellbore point cloud data, and fuse and analyze the crack recognition result with the three-dimensional wellbore point cloud data. Specifically: Point cloud and image fusion: Establish the correspondence between the three-dimensional wellbore point cloud data and the high-definition image, and through image-point cloud projection mapping, realize the correspondence of crack features in three-dimensional space; Verify the authenticity of crack features by depth difference: Calculate the three-dimensional point cloud depth change of the crack candidate area where the crack feature is located through the three-dimensional wellbore point cloud data. If the three-dimensional point cloud depth change corresponding to the image crack area exceeds the set threshold, confirm it as a real crack; otherwise, judge it as a misjudgment and eliminate it; Final crack confirmation: According to the depth difference verification result, obtain the real three-dimensional coordinate information and image texture coordinate information of the crack.

[0010] As a further solution of the present invention, the interval segmentation analysis module further includes a second wellbore defect recognition unit and a defect grading and evaluation unit; the second wellbore defect recognition unit uses the geometric information of the three-dimensional wellbore point cloud data to detect the planar defects on the wellbore surface; the defect grading and evaluation unit is used to perform grading and evaluation according to the linear defects and planar defects on the wellbore surface to determine the wellbore section that needs to be repaired.

[0011] As a further solution of the present invention, the second wellbore defect recognition unit uses the geometric information of the three-dimensional wellbore point cloud data to detect the planar defects on the wellbore surface. Specifically: Perform point cloud preprocessing on the three-dimensional wellbore point cloud data to eliminate outlier noise, and divide the three-dimensional wellbore point cloud data into local regions based on the spatial clustering method; Calculate the change of the normal vector field of the three-dimensional wellbore point cloud data in each local region. The local region with the change of the normal vector field exceeding the preset change threshold is marked as the defect region; calculate the depth difference between the local region and the surrounding normal regions, and if it exceeds the depth difference threshold, it is marked as the defect region; Measure the two-dimensional area of the defect area using its convex hull, and calculate the maximum depth difference between the defect area and the adjacent normal area as the depth index of the planar defect.

[0012] As a further solution of the present invention, the linkage control mechanism module is used to intelligently control the local ventilator and air duct of the shaft wall section after determining the shaft wall section to be repaired; if the existing air volume in the shaft wall section to be repaired is insufficient, remotely start or increase the wind speed of the local ventilator in the adjacent shaft wall section, and adjust the air duct valve or the outlet direction to lead the fresh air flow to the shaft wall section to be repaired; for the non-operating shaft wall section or the shaft wall section not to be repaired temporarily, reduce the air volume supply.

[0013] The technical effects and advantages of a main ore pass shaft wall analysis system based on lidar point cloud and high-definition image of the present invention: By dividing the shaft section, using a sensor platform composed of a lidar and a high-definition camera, high-precision continuous scanning and image acquisition are carried out under the guidance of the track, and the spatial geometric information and surface texture details of the shaft wall are accurately obtained, effectively capturing small defects; through the pre-calibrated pose relationship and automatic registration algorithm, multiple sections of scanned point clouds and high-definition texture images are efficiently fused into a complete and high-precision three-dimensional shaft wall point cloud model, ensuring the accuracy and efficiency of data fusion of different shaft sections; through the first shaft wall defect recognition unit, shaft wall cracks are accurately detected and sized, through the second shaft wall defect recognition unit, planar defects such as depressions and spallings are identified, and through the defect grading and evaluation unit, the severity of the defects is quantitatively graded to determine the specific shaft wall section to be repaired, integrating image and point cloud technologies to eliminate misjudgments and improve the accuracy of defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic structural diagram of a main ore pass shaft wall analysis system based on lidar point cloud and high-definition image provided by the present invention; Figure 2 It is a schematic diagram of the damaged area of a certain existing mine ore pass provided by the present invention; Figure 3 It is a graph showing the change in the quantity of real-time three-dimensional shaft wall point cloud data provided by the present invention; Figure 4 It is a pie chart showing the distribution of main ore pass shaft wall defect types provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] Next, the technical solutions in the present invention will be clearly and completely described in conjunction with the drawings in the present invention. Obviously, the described technical solutions are only a part of the present invention, rather than all of them. Based on the technical solutions in the present invention, all other technical solutions obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0016] Embodiment 1

[0017] As Figure 1 shown in the structural schematic diagram of a main ore pass shaft wall analysis system based on lidar point cloud and high-definition image. A main ore pass shaft wall analysis system based on lidar point cloud and high-definition image includes a three-dimensional scanning and modeling module, an interval segmentation analysis module, and a linkage control mechanism module; the three-dimensional scanning and modeling module is connected to the interval segmentation analysis module, and the interval segmentation analysis module is connected to the linkage control mechanism module.

[0018] The three-dimensional scanning and modeling module is used to obtain data and images by using a sensor platform composed of a lidar and a high-definition camera, and use an automatic registration algorithm to splice the three-dimensional shaft wall point cloud data obtained by multiple scans into a shaft wall point cloud model in the same coordinate system; The interval segmentation analysis module is used to realize the automatic identification and evaluation of shaft wall defects by integrating high-definition images and three-dimensional shaft wall point cloud data; The linkage control mechanism module is used to intelligently control the local ventilator and air duct of the shaft wall section after determining the shaft wall section to be repaired.

[0019] Specifically, the three-dimensional scanning and modeling module includes a well section division unit, a sensor platform, a shaft wall data extraction unit, and a shaft wall data fusion unit; the sensor platform includes a lidar and a high-definition camera; the well section division unit and the sensor platform are respectively connected to the shaft wall data extraction unit, and the shaft wall data extraction unit is connected to the shaft wall data fusion unit; the sensor platform includes a lidar and a high-definition camera; The well section division unit is used to evenly divide the inclined ore pass to be measured into several measured well sections; The sensor platform is driven by track guidance and uses a constant speed to arrange the high-definition camera around the lidar from a lateral perspective to obtain the first shaft wall data and high-definition images of the measured well section; The shaft wall data extraction unit extracts the first shaft wall data according to the order of the measured well section from top to bottom; The shaft wall data fusion unit is used to pre-calibrate the relative pose of the lidar and the high-definition camera, align the collected first shaft wall data with the corresponding point cloud coordinate system, obtain three-dimensional shaft wall point cloud data with texture information, and use an automatic registration algorithm to splice the three-dimensional shaft wall point cloud data obtained by multiple scans into a shaft wall point cloud model in a unified coordinate system.

[0020] The lidar provides high-density three-dimensional spatial structure information, and the high-definition camera provides high-resolution texture images. The two complement each other's advantages. The camera is arranged laterally around the lidar to avoid imaging blind spots caused by narrow wellbores, uneven structures, or equipment occlusion. The sensor platform is driven by the track guiding system to move at a constant speed, ensuring the continuity and uniformity of data acquisition during scanning and avoiding problems such as frame skipping and missed scans. The well section division unit automatically performs equidistant segmentation according to the well depth to meet the customized scanning requirements for different well depths. Each section of data is independently extracted and processed, which is conducive to parallel modeling and segmented diagnostic analysis, reduces the overall processing delay, supports the stitching process between well sections, and ensures the integrated coherence of the three-dimensional model of the entire well. The automatic registration algorithm (such as ICP, feature matching, etc.) is used to seamlessly stitch the point clouds of each section and unify them to the global coordinate system to avoid manual alignment errors. The point cloud and the high-definition image are spatially mapped through pose calibration to generate a texture-enhanced three-dimensional model with real surface texture. The output model has geometric authenticity and visual realism and is suitable for high-precision tasks such as subsequent defect identification and crack measurement. The constructed three-dimensional point cloud model can be directly connected to the defect identification algorithm to achieve accurate analysis of wellbore diseases such as cracks, spalling, and depressions. The multi-section model supports interval analysis and dynamic update, providing data support for subsequent defect grading, repair planning, and ventilation linkage control.

[0021] Specifically, the interval segmentation analysis module includes a first wellbore defect identification unit, a second wellbore defect identification unit, and a defect grading and evaluation unit. The first wellbore defect identification unit and the second wellbore defect identification unit are respectively connected to the defect grading and evaluation unit. The first wellbore defect identification unit detects the linear defects on the wellbore surface through image processing and pattern recognition algorithms, and identifies the linear defect features based on texture analysis and edge extraction of the high-definition image. Combining the three-dimensional wellbore point cloud data to filter out false judgments (such as distinguishing shadows from real cracks) and calculate the size parameters and orientations of the wellbore cracks. The size parameters of the cracks include the length, width, and penetration depth of the cracks. The linear defect features include wellbore cracks.

[0022] The second wellbore defect identification unit detects the planar defects on the wellbore surface using the geometric information of the three-dimensional wellbore point cloud data.

[0023] The defect grading and evaluation unit is used to perform grading and evaluation based on the linear defects and planar defects on the wellbore to determine the wellbore sections that need to be repaired. For example, according to the crack length and width, they are divided into minor cracks and through cracks, and according to the spalling area / depth, they are judged as surface spalling or deep spalling grades.

[0024] Specifically, the first wellbore defect identification unit detects the linear defects on the wellbore surface through image processing and pattern recognition algorithms, and identifies the linear defect features based on texture analysis and edge extraction of the high-definition image. Specifically: Obtain high-definition images of the wellbore and preprocess the high-definition images, including image distortion correction, brightness and contrast enhancement, and denoising processing; Use an image texture feature extraction algorithm on the processed high-definition images to achieve preliminary crack identification and obtain crack candidate regions; Perform morphological thinning processing on the crack candidate regions, extract the skeleton curves of the crack candidate regions, and use the connected component labeling method to label different crack regions as independent crack features as the crack identification results; calculate the skeleton length of the crack features and the average width of the crack regions.

[0025] Specifically, filter misjudgments by combining three-dimensional wellbore point cloud data, and fuse and analyze the crack identification results with the three-dimensional wellbore point cloud data. Specifically: Point cloud and image fusion: Establish the correspondence between the three-dimensional wellbore point cloud data and the high-definition images, and through image-point cloud projection mapping, achieve the correspondence of crack features in three-dimensional space; Verify the authenticity of crack features through depth differences: Calculate the three-dimensional point cloud depth change of the crack candidate region where the crack features are located through the three-dimensional wellbore point cloud data. If the three-dimensional point cloud depth change corresponding to the image crack region exceeds the set threshold, confirm it as a real crack; otherwise, determine it as a misjudgment and eliminate it; Final crack confirmation: According to the depth difference verification results, obtain the real three-dimensional coordinate information and image texture coordinate information of the cracks.

[0026] Specifically, the second wellbore defect identification unit uses the geometric information of the three-dimensional wellbore point cloud data to detect planar defects on the wellbore. Specifically: Perform point cloud preprocessing on the three-dimensional wellbore point cloud data to eliminate outlier noise, and divide the three-dimensional wellbore point cloud data into local regions based on a spatial clustering method; Calculate the change in the normal vector field of the three-dimensional wellbore point cloud data in each local region. The local regions with a change in the normal vector field exceeding the preset change threshold are marked as defect regions; calculate the depth difference between the local region and the surrounding normal regions, and mark it as a defect region if it exceeds the depth difference threshold; Measure the two-dimensional area of the defect region using the convex hull of the defect region, and calculate the maximum depth difference between the defect region and the adjacent normal region as the planar defect depth index.

[0027] Specifically, the defect grading and evaluation unit is used to perform grading and evaluation based on the linear defects and planar defects on the wellbore to determine the wellbore sections that need to be repaired. For example, classify them into minor cracks and through cracks according to the crack length and width, and determine the surface spalling or deep spalling grade according to the spalling area / depth. The defect grading criteria can be set in advance according to mine safety specifications, and the evaluation results will provide a basis for subsequent maintenance and reinforcement.

[0028] Combining visual texture features with three-dimensional structure information effectively avoids the common problems of false detection and missed detection in single recognition methods; differentiates real cracks from shadows and stains, greatly improving the detection accuracy and robustness; the multi-dimensional quantification (length, width, depth, area) of cracks and spalls supports refined diagnosis. Automatically executes complex steps such as image preprocessing, region marking, skeleton extraction, and normal vector judgment; transforms the wellbore detection from "relying on manual subjective judgment" to "objective analysis based on parameter drive"; reduces the pressure of manual inspection and improves the data processing ability in complex well sections (such as deep wells below 70 meters). Maps the identified defects to specific well sections to achieve repair marking at the well section level; can generate a repair work list (including defect type, grade, coordinate range); provides precise guidance for subsequent construction tasks such as shotcrete reinforcement, grouting sealing, and bolt layout, constructing the core analysis link of the intelligent identification and maintenance assistance system for mine wellbore diseases. It not only has the technical characteristics of high precision, high automation, and high scalability, but also can be deeply integrated with safety specifications and construction decisions, becoming a key technical unit in the digital upgrade and intelligent operation and maintenance system of mines.

[0029] Specifically, preprocess the high-definition image: Image distortion correction: Use the camera internal parameter calibration result to perform radial distortion correction on the image to eliminate the lens distortion effect and ensure the dimensional accuracy of crack recognition; Brightness and contrast enhancement: Aiming at the problem of uneven illumination in the underground wellbore environment, adopt the adaptive histogram equalization method to improve the overall brightness and local contrast of the image and highlight the wellbore crack features; Denoising processing: Use median filtering or non-local means filtering methods to eliminate the image noise caused by dust or water mist and reduce misjudgment.

[0030] Use the image texture feature extraction algorithm to achieve preliminary crack recognition for the processed high-definition image, including: Grayscale conversion and edge detection: Use the Sobel operator and Canny operator to extract the image edges and obtain the edge image of the candidate area of the wellbore surface crack; Crack enhancement processing: Use the Gabor filter or Hessian matrix method to enhance the linear features and highlight the crack morphology; Preliminary selection of crack candidate areas: Set a threshold to screen the areas that meet the obvious crack texture features as crack candidate areas.

[0031] Specifically, the linkage control mechanism module is used to intelligently control the local ventilator and air duct of the wellbore section after determining the wellbore section to be repaired; if the existing air volume in the wellbore section to be repaired is insufficient, remotely start or increase the wind speed of the local ventilator (local fan) in the adjacent wellbore section, and adjust the air duct valve or the air outlet direction to direct the fresh air flow to the wellbore section to be repaired; for non-operating wellbore sections or wellbore sections that are not to be repaired temporarily, reduce the air volume supply and concentrate the limited ventilation capacity to serve the key working face.

[0032] Through the above linkage control, on the one hand, it ensures that maintenance personnel can obtain sufficient fresh air supply in deep well environments such as -30m to -70m, diluting the concentration of harmful gases. On the other hand, it enhances the discharge of local dust, reducing the risk of dust retention caused by repair operations such as rock drilling and shotcreting. The ventilation control module can be connected to the mine ventilation monitoring system interface to realize real-time monitoring of the wind speed and air volume in the operation area, and automatically adjust the fan in a closed loop according to the preset safety threshold, increasing the ventilation and air change intensity in the fault section area to ensure that the entire repair operation process is carried out in a safe and low-dust environment.

[0033] Figure 3 This is the graph showing the change in the quantity of real-time three-dimensional shaft wall point cloud data provided by the present invention; the horizontal axis is time (00:00~12:00), and the vertical axis is the quantity of three-dimensional shaft wall point cloud data (unit: number of points). The initial point cloud data quantity is approximately 900 points; the point cloud quantity starts to increase significantly from 06:00, reaching a peak of approximately 1450 points at 08:00; the data quantity then tends to be stable, indicating that the scanning is gradually completed and the system is operating stably; it reflects that the scanning platform obtains a large amount of shaft wall point cloud data during the morning period (possibly the high-intensity scanning stage), and the system has the ability to gradually construct a high-density three-dimensional model. There may be a correlation between the point cloud quantity and the complexity of the well section or the concentrated fracture section.

[0034] Figure 4 This is the pie chart showing the distribution of the types of defects in the main ore pass shaft wall provided by the present invention; it shows the distribution of various structural defects in the main ore pass shaft wall in the form of a pie chart, divided into four categories according to the types of defects: cracks (blue): the largest number, reaching 1048, which is the main defect type; spalling (green): the second largest, reaching 735, indicating that there is more material shedding on the shaft wall surface; water seepage (yellow): reaching 580, referring to the phenomenon of local shaft wall leakage; deformation (red): reaching 484, indicating that there is structural deformation or bulging in the shaft wall. It shows that the main problems of the shaft wall structure are cracks and spalling, reflecting obvious stability and durability risks in the shaft wall, and it is necessary to give priority to reinforcement and maintenance.

[0035] Example 2

[0036] The specific situation of the ore pass repair based on the above system: The main ore pass repair project adopts a two-stage construction method. First, the ore pass is filled with ore. Before the repair, the cleaning work of the ore pass is carried out. By fabricating and installing temporary derricks, self-made hanging platforms, winding engines and other hoisting facilities, personnel enter the shaft through the 60m horizontal hanging bucket for construction. It adopts 3-shift cyclic operation and is divided into 3 teams. In the first stage, one team is responsible for cleaning loose rocks on the shaft wall and pre-support work, one team is engaged in the processing of anchor bolts, steel bars and manganese steel plates, and one team is responsible for pouring and curing work.

[0037] In the first stage, work such as shaft cleaning, brushing and expanding, supporting, and stockpiling materials is carried out. The ore is filled into the entire shaft, and then the ore is gradually discharged, with the discharge height of 2m each time, serving as the preliminary supporting working platform. Personnel are lowered into the well through a kibble. The rubble on the shaft wall and the remaining rubble at the damaged parts are cleaned and pried clean to prevent loose stones from falling and injuring construction workers during the next step of work. When local fractures are found during the process of cleaning and inspecting the shaft, bolt support is carried out. In the second stage, based on the completion of cleaning on both sides, final casting support is carried out on the damaged area from top to bottom.

[0038] Repair situation from 60m to 30m: The damage degree of the shaft wall from 60m to 30m is relatively light. The "patching" repair method is adopted, and the support method for the shaft wall is double-layer steel bars + light rails (parallel to the shaft wall).

[0039] Repair situation from 30m to 0m: The damaged manganese steel plates are replaced for the repair of the 30m branch shaft. A buffer pit is reserved on the opposite side of the 30m branch shaft. Manganese steel plates are arranged at the bottom plate of the buffer pit and the collision part of the ore. The bottom plate of the buffer pit is supported with double-layer steel bars + light rails, and the shaft wall is supported in the same form.

[0040] Repair situation from 0m to -30m: Local support repair and replacement of manganese steel plates are adopted for the 0m branch shaft, and the bottom plate is reinforced and cast; a buffer pit is reserved on the opposite side of the 0m branch shaft. Manganese steel plates are arranged at the bottom plate of the buffer pit and the anti-smashing part of the ore. Double-layer steel bars are arranged on both sides and rails are installed. Anchor bolts are arranged at the bottom plate of the buffer pit and connected to the double-layer reinforcement; to increase the anti-smashing ability and bearing strength, the entire shaft wall below 0m is supported with reinforced concrete + light rails (parallel to the shaft wall).

[0041] Repair situation from -30m to -70m: This section of the shaft wall is the most severely damaged. According to the actual situation on site, the shaft diameter at the severely damaged part is changed from 2m to 5m (the shape at some locations depends on the damage shape, achieving the goal of reducing the amount of support without reducing the strength to shorten the construction period. The concrete pouring volume is reduced by about 350m³ through the diameter change). A transition section is added at the diameter change, and the support strength of the transition section is increased. The shaft wall support adopts the support method of double-layer steel bars + steel wires + anchor bolts + light rails; the -30m branch shaft and the buffer pit are severely damaged sections. This section adopts the support method of double-layer steel bars + steel wires + anchor bolts + light rails to ensure the integrity of the support. And the rails are vertically arranged in a "plum blossom dot pattern" at the bottom plates of the buffer pit and the branch shaft to increase the anti-smashing ability.

[0042] Repair situation from -70m to -110m: This section of the shaft wall is severely damaged. Since the diameter cannot be changed in this section, it is repaired by restoring the original 2m-diameter shaft wall. The support method is reinforced concrete + light rail, and steel ropes are added at locations with larger damage. For the damaged floor of the -70m branch chute, the support method is reinforced concrete + light rail + "light rail in plum blossom shape vertically arranged". Since the damaged area on the opposite side caused by ore impact is not sufficient to reserve a buffer pit, no buffer pit is left on the opposite side of the -70m branch chute, and the support method of double-layer steel bars + steel ropes + anchor bolts + light rail + "light rail in plum blossom shape vertically arranged" is adopted to ensure the anti-smashing strength.

[0043] Repair situation from -110m to -132m: This section enters the ore bin part. Only the manganese steel plates at the damaged parts of the ore bin are replaced, and the rubble in the ore bin and the -110m branch chute is cleaned up.

[0044] In the embodiment of the present invention, by dividing the shaft sections, using a sensor platform composed of a lidar and a high-definition camera, high-precision continuous scanning and image acquisition are carried out under the guidance of the track, the spatial geometric information and surface texture details of the shaft wall are accurately obtained, and tiny defects are effectively captured; through the pre-calibrated pose relationship and the automatic registration algorithm, multiple sections of scanned point clouds and high-definition texture images are efficiently fused into a complete and high-precision three-dimensional shaft wall point cloud model, ensuring the accuracy and efficiency of data fusion of different shaft sections; through the first shaft wall defect recognition unit, shaft wall cracks are accurately detected and sized, through the second shaft wall defect recognition unit, planar defects such as depressions and spallings are identified, and through the defect grading and evaluation unit, the severity of the defects is quantitatively graded to determine the specific shaft wall sections to be repaired, integrating image and point cloud technologies to eliminate misjudgments and improve the accuracy of defect detection.

[0045] As mentioned above, the above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights.

[0046] Finally: The above is only the preferred solution of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A main ore pass shaft wall analysis system based on laser point cloud and high-definition image, comprising a three-dimensional scanning and modeling module, an interval segmentation analysis module, and a linkage control mechanism module; characterized in that, The 3D scanning modeling module is used to acquire data and images using a sensor platform composed of a laser radar and a high-definition camera, and to stitch the 3D wellbore point cloud data obtained from multiple scans into a wellbore point cloud model in the same coordinate system using an automatic registration algorithm; the interval segmentation analysis module is used to realize the automatic identification and evaluation of wellbore defects by integrating high-definition images and 3D wellbore point cloud data; The interval segmentation analysis module includes a first well wall defect recognition unit; the first well wall defect recognition unit detects well wall linear defects on the well wall surface through image processing and pattern recognition algorithms, identifies linear defect features based on texture analysis and edge extraction of high-definition images, and filters misjudgments in combination with three-dimensional well wall point cloud data; The linkage control mechanism module is used to intelligently control the local ventilators and air ducts in the shaft wall section after determining the shaft wall section that needs to be repaired.

2. The main chute shaft wall analysis system based on laser point cloud and high-definition image according to claim 1, wherein The three-dimensional scanning modeling module includes a well section division unit, a sensor platform, a well wall data extraction unit and a well wall data fusion unit; the sensor platform includes a laser radar and a high-definition camera; the well section division unit is used to evenly divide the inclined well to be tested into several well sections to be tested; the well wall data extraction unit extracts the first well wall data according to the order of the well sections to be tested from top to bottom; the well wall data fusion unit is used to pre-calibrate the relative posture of the laser radar and the high-definition camera, align the collected first well wall data with the corresponding point cloud coordinate system, obtain three-dimensional well wall point cloud data with texture information, and use an automatic registration algorithm to splice the three-dimensional well wall point cloud data obtained from multiple scans into a well wall point cloud model under a unified coordinate system.

3. The main ore pass shaft wall analysis system based on laser point cloud and high-definition image according to claim 2, characterized in that, The sensor platform includes a laser radar and a high-definition camera; the sensor platform is driven by a track guide, and the high-definition camera is deployed around the laser radar at a constant speed with a lateral perspective to obtain the first wellbore wall data and high-definition images of the well section to be measured.

4. The main chute shaft wall analysis system based on laser point cloud and high-definition image according to claim 1, characterized in that, The first wellbore defect recognition unit detects wellbore linear defects on the wellbore surface through image processing and pattern recognition algorithms, and recognizes linear defect features based on texture analysis and edge extraction of high-definition images, specifically: Obtain high-definition images of the well wall and pre-process the high-definition images, including image distortion correction, brightness and contrast enhancement, and denoising; The processed high-definition images are used to extract image texture features to achieve preliminary crack identification and obtain crack candidate areas; The crack candidate areas are morphologically refined and the skeleton curves of the crack candidate areas are extracted. The connected domain labeling method is used to label different crack areas as independent crack features as the crack recognition results. The skeleton length of the crack features and the average width of the crack area are calculated.

5. The main ore pass shaft wall analysis system based on laser point cloud and high-definition image according to claim 4, characterized in that, Combined with the 3D wellbore point cloud data to filter misjudgment, the fracture identification results are fused and analyzed with the 3D wellbore point cloud data, specifically: Point cloud and image fusion: Establish the correspondence between 3D wellbore point cloud data and high-definition images, and achieve the correspondence of fracture features in 3D space through image-point cloud projection mapping; Depth difference verifies the authenticity of fracture features: Calculate the depth change of the 3D point cloud of the fracture candidate area where the fracture features are located through the 3D well wall point cloud data. If the depth change of the 3D point cloud corresponding to the image fracture area exceeds the set threshold, it is confirmed to be a real fracture. Otherwise, it is determined as a misjudgment and eliminated; Final crack confirmation: According to the depth difference verification results, the true three-dimensional coordinate information and image texture coordinate information of the cracks are obtained.

6. The main ore pass shaft wall analysis system based on laser point cloud and high-definition image according to claim 1, characterized in that The interval segmentation analysis module further includes a second wellbore defect identification unit and a defect grading and evaluation unit; the second wellbore defect identification unit detects the planar defects of the wellbore using the geometric information of the three-dimensional wellbore point cloud data; the defect grading and evaluation unit is used to perform grading and evaluation based on the linear defects and planar defects of the wellbore to determine the wellbore section that needs to be repaired.

7. The main ore pass shaft wall analysis system based on laser point cloud and high-definition image according to claim 6, characterized in that, The second wellbore defect identification unit detects the planar defects of the wellbore using the geometric information of the three-dimensional wellbore point cloud data, specifically: Perform point cloud preprocessing on the three-dimensional wellbore point cloud data to eliminate outlier noise, and divide the three-dimensional wellbore point cloud data into local regions based on the spatial clustering method; Calculate the change in the normal vector field of the three-dimensional wellbore point cloud data in each local region, and mark the local region where the change in the normal vector field exceeds the preset change threshold as the defect region; Calculate the depth difference between the local region and the surrounding normal region, and mark it as the defect region if it exceeds the depth difference threshold; Measure the two-dimensional area of the defect region using the convex hull of the defect region, and calculate the maximum depth difference between the defect region and the adjacent normal region as the depth index of the planar defect.

8. The main ore pass shaft wall analysis system based on laser point cloud and high-definition image according to claim 1, characterized in that The linkage control mechanism module is used to intelligently control the local ventilator and air duct of the wellbore section after determining the wellbore section that needs to be repaired.

9. The main chute shaft wall analysis system based on laser point cloud and high-definition image according to claim 8, wherein If the existing air volume in the wellbore section that needs to be repaired is insufficient, remotely start or increase the wind speed of the local ventilator in the adjacent wellbore section, and adjust the air duct valve or the outlet direction to direct the fresh air flow to the wellbore section that needs to be repaired.

10. The main ore pass shaft wall analysis system based on laser point cloud and high-definition image according to claim 9, characterized in that, For non-operating wellbore sections or wellbore sections that are not to be repaired temporarily, reduce the air volume supply.

Images