Intelligent detection system for construction quality of side slope anchor rod drilling platform based on machine vision

Abstract

The invention relates to the technical field of machine vision, and discloses a side slope anchor rod drilling platform construction quality intelligent detection system based on machine vision. The method comprises the steps that an acquisition module acquires a multi-angle original image with a space-time mark; the correction module performs preprocessing to obtain a standardized image; the extraction module constructs a drilling platform three-dimensional model; the monitoring module detects track deviation by using an improved adaptive Kalman filtering algorithm; the calculation module evaluates the drilling quality; and the analysis module comprehensively analyzes and generates quality grade and early warning information. According to the intelligent detection method for the construction quality of the side slope anchor rod drilling platform based on machine vision, real-time monitoring, dynamic analysis and quality early warning of the construction process are achieved, subjectivity and hysteresis of traditional manual detection are overcome, and the detection precision and efficiency of the anchor rod drilling construction quality are improved.

Description

Technical Field

[0001] The present application relates to the field of machine vision technology, and in particular to a machine vision-based intelligent detection system for the construction quality of a slope anchor drilling platform. Background Art

[0002] Slope anchor drilling technology is a commonly used reinforcement measure in slope engineering. By drilling anchor holes on the slope surface, installing anchors, and grouting to secure them, an anchoring system is formed to improve slope stability. Traditional slope anchor drilling platform construction quality inspections rely primarily on manual visual inspections and simple measurement tools, such as using a tape measure to measure drilling depth, using a spirit level to check platform levelness, and testing anchoring force through tests. With the development of computer vision and sensor technology, some projects have begun to introduce cameras to record the construction process and assess construction quality through post-analysis. In recent years, some studies have attempted to apply technologies such as laser scanning and 3D reconstruction to slope engineering monitoring, but these efforts have primarily focused on monitoring overall slope deformation. Research on intelligent inspection of drilling platform construction quality is still relatively lacking.

[0003] Existing technologies have significant shortcomings: First, manual inspection methods are highly subjective and have low measurement accuracy, making it impossible to obtain quality information about the interior of the borehole and the anchor installation process. Second, existing machine vision applications mostly focus on single-parameter detection or static image processing, lacking the ability to dynamically monitor and conduct comprehensive analysis of the entire construction process. Third, there is a gap between data collection and quality assessment, making it difficult to achieve real-time feedback and early warning. Fourth, correlation analysis between different quality parameters is lacking, making it impossible to identify the root causes of quality issues from a systematic perspective. Fifth, inspection results lack objective quantitative standards and a systematic evaluation system, making it difficult to provide a scientific basis for construction decision-making. These shortcomings have led to frequent quality problems in slope anchor projects, increasing project risks and maintenance costs. Summary of the Invention

[0004] The present application provides a machine vision-based intelligent detection system for the construction quality of a slope anchor drilling platform, which is used to construct a machine vision-based intelligent detection method for the construction quality of a slope anchor drilling platform, realize real-time monitoring, dynamic analysis and quality warning of the construction process, overcome the subjectivity and lag of traditional manual detection, and improve the detection accuracy and efficiency of the anchor drilling construction quality.

[0005] The present application provides a machine vision-based intelligent detection system for the construction quality of a slope anchor drilling platform. The machine vision-based intelligent detection system for the construction quality of a slope anchor drilling platform includes: The acquisition module is used to collect multi-angle images of the slope anchor drilling platform construction area using a high-resolution industrial camera to obtain a raw image dataset with time and space tags; A correction module, configured to perform image preprocessing and geometric correction on the original image data set to obtain a standardized drilling platform image; an extraction module, configured to extract geometric characteristic parameters of the drilling platform based on the standardized drilling platform image to obtain a three-dimensional spatial model of the drilling platform; A monitoring module, configured to apply an improved adaptive Kalman filter algorithm to the three-dimensional spatial model of the drilling platform to perform real-time monitoring of the drilling trajectory and obtain drilling trajectory deviation data; a calculation module, configured to calculate key drilling quality parameters based on the drilling trajectory deviation data to obtain a drilling quality evaluation index; The analysis module is used to comprehensively analyze the drilling quality evaluation index and the anchor installation process monitoring data to obtain the slope anchor drilling platform construction quality grade and early warning information.

[0006] In the technical solution provided by the present application, multi-angle image acquisition is performed by a high-resolution industrial camera to obtain a raw image data set with time and space tags, which effectively solves the problem of incomplete and discontinuous traditional single-point observation data, and provides rich raw data resources for subsequent comprehensive quality analysis; image preprocessing and geometric correction are performed on the raw image data set to obtain a standardized drilling platform image, which overcomes interference factors such as unstable image quality and perspective distortion in complex on-site environments, and significantly improves the accuracy of subsequent analysis and processing; based on the standardized drilling platform image, the geometric feature parameters of the drilling platform are extracted to obtain a three-dimensional spatial model of the drilling platform, and the two-dimensional image information is converted into three-dimensional spatial geometric information, so that key parameters such as the drilling platform size, drilling position and angle are accurately quantified; an improved adaptive Kalman filtering algorithm is applied to the three-dimensional spatial model of the drilling platform to perform drilling trajectory simulation Real-time monitoring is carried out to obtain drilling trajectory deviation data. The algorithm is optimized for the geological change characteristics unique to slope engineering. Through real-time adjustment of the state noise covariance matrix, the robustness and accuracy of trajectory monitoring under complex geological conditions are significantly improved, so that small deviation changes in the drilling process can be captured in time; based on the drilling trajectory deviation data, the key parameters of drilling quality are calculated to obtain drilling quality evaluation indicators, which realizes the objective quantification of multi-dimensional quality characteristics such as drilling straightness, depth, diameter and wall roughness, avoiding the subjectivity and uncertainty of traditional manual measurement; the drilling quality evaluation indicators and the anchor installation process monitoring data are comprehensively analyzed to obtain the construction quality grade and early warning information of the slope anchor drilling platform. Through multi-source data fusion and comprehensive evaluation model, a systematic and standardized quality assessment system is established, realizing the leap from single indicator detection to comprehensive quality evaluation. It is particularly worth emphasizing that the improved adaptive Kalman filter algorithm introduced in the present invention for drilling trajectory monitoring fully considers the particularities of specific application areas in slope engineering. By sensing changes in geological conditions and automatically adjusting filter parameters, this algorithm effectively solves the problem of large trajectory prediction deviations in traditional Kalman filters in geologically complex areas. The weight allocation mechanism combining the entropy weight method with hierarchical analysis adopted in the comprehensive quality evaluation phase fully adapts to the multidimensional characteristics of slope anchor engineering quality evaluation, overcoming the arbitrariness of subjective weighting and the limitations of fixed weights that cannot reflect actual changes. The specific field optimization and combined application of these artificial intelligence algorithms give this solution significant advantages in the field of slope anchor drilling quality inspection. It not only improves the accuracy and comprehensiveness of inspection, but also realizes the intelligence of real-time monitoring and early warning, providing strong technical support for improving the quality and safety of slope engineering construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0008] Figure 1 This is a schematic diagram of an embodiment of an intelligent detection system for construction quality of a slope anchor drilling platform based on machine vision in an embodiment of the present application. DETAILED DESCRIPTION

[0009] An embodiment of the present application provides a machine vision-based intelligent detection system for the construction quality of a slope anchor drilling platform. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or devices.

[0010] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of the intelligent detection system for construction quality of a slope anchor drilling platform based on machine vision includes: The acquisition module 101 is used to acquire multi-angle images of the slope anchor drilling platform construction area using a high-resolution industrial camera to obtain a raw image dataset with time and space tags; The correction module 102 is used to perform image preprocessing and geometric correction on the original image data set to obtain a standardized drilling platform image; An extraction module 103 is configured to extract geometric characteristic parameters of the drilling platform based on the standardized drilling platform image to obtain a three-dimensional spatial model of the drilling platform; A monitoring module 104 is configured to apply an improved adaptive Kalman filter algorithm to the three-dimensional spatial model of the drilling platform to perform real-time monitoring of the drilling trajectory and obtain drilling trajectory deviation data; A calculation module 105 is configured to calculate key drilling quality parameters based on the drilling trajectory deviation data to obtain a drilling quality evaluation index; The analysis module 106 is used to comprehensively analyze the drilling quality evaluation index and the anchor installation process monitoring data to obtain the slope anchor drilling platform construction quality grade and early warning information.

[0011] Specifically, the acquisition module 101 implements multi-angle image acquisition of the slope anchor drilling platform construction area. This is accomplished by establishing a spatial coordinate system by setting at least four positioning markers around the construction area and deploying at least three 12-megapixel global shutter CCD sensor industrial cameras to form a multi-angle coverage network. During acquisition, the industrial cameras continuously capture images at a rate of 5 frames per second, simultaneously recording image timestamps and spatial location information. During critical construction stages, such as the start and completion of drilling, the acquisition rate is automatically increased to 15 frames per second to capture more detail. The captured images are stored in a lossless compression format, and each image is assigned a unique identification code, which is linked to the construction log for subsequent query and analysis. The correction module 102 first applies a Gaussian filter to the raw image dataset to reduce noise caused by factors such as dust and vibration in the construction environment. It then enhances image contrast through histogram equalization, enhancing the clarity of the drill hole edges and platform structure. A perspective transformation matrix is ​​calculated based on the positioning marker information to correct for lens distortion. Finally, the images captured at different angles are spatially registered and mapped to the same coordinate system to generate a standardized image of the drilling platform. The extraction module 103 applies the Canny edge detection algorithm to the standardized drilling platform image to extract contours. This algorithm uses a dual-threshold method, with the lower threshold set at 40% of the upper threshold to ensure detection continuity and accuracy. The extracted contours are then fed into the Hough transform algorithm to identify linear and circular structural elements and locate the drilling platform boundary and drill hole locations. Based on the matching feature points of images from different angles, structured light 3D reconstruction technology is used to construct point cloud data. Finally, through surface fitting and gridding, a 3D spatial model is established that includes the platform dimensions, drill hole positions, and angles. The monitoring module 104 first sets the ideal state vector of the drilling trajectory based on the 3D model of the drilling platform, including the drilling start point position, target depth, designed angle, and direction. The drill rod position and depth sensors collect real-time motion parameters of the drill rod to construct an observation vector. The observation vector is processed using an improved adaptive Kalman filter algorithm. This algorithm, based on the standard Kalman filter, adds the ability to adjust the state noise covariance matrix in real time based on geological parameters, adaptively correcting for measurement errors caused by changing geological conditions. Finally, the lateral and angular deviations between the filtered actual drilling trajectory and the ideal trajectory are calculated to generate drilling trajectory deviation data. The calculation module 105 applies the least squares method to the drilling trajectory deviation data to fit the drilling centerline. It calculates the maximum distance from the actual drilling trajectory point to the fitted line and determines the drilling straightness. The drilling depth is measured by analyzing the drill rod feed length and the change in drill bit position and compared with the designed depth. The drilling diameter is measured at different depths and a drilling diameter variation curve is constructed. The gray-level co-occurrence matrix method is used to analyze the texture features of the borehole wall image, calculate texture parameters such as energy, entropy, and contrast, and evaluate the borehole wall roughness to form a drilling quality evaluation index.The analysis module 106 first monitors the anchor installation process through a target tracking algorithm, records the anchor position, posture and grouting parameter changes, and obtains the anchor installation process monitoring data. The drilling quality evaluation index and the anchor installation process monitoring data are input into the hierarchical analysis model to establish a quality evaluation system including four first-level indicators: platform geometry parameters, drilling trajectory parameters, drilling quality parameters and anchor installation parameters. The entropy weight method is applied to each indicator in the hierarchical analysis model to assign weight coefficients. The entropy weight method automatically assigns weights according to the degree of indicator variation by calculating the information entropy of the indicator, avoiding the influence of subjective factors. Finally, a four-level warning threshold is set based on the comprehensive quality score, and the quality level is divided into four levels: excellent, qualified, critical and unqualified, to generate the slope anchor drilling platform construction quality level and warning information.

[0012] In a specific embodiment, the acquisition module 101 is configured to: At least four positioning marking points are set around the construction area of ​​the slope anchor drilling platform to construct a spatial coordinate system; Arrange at least three 12-megapixel global shutter CCD sensor industrial cameras in the construction area of ​​the slope anchor drilling platform to form a multi-angle coverage network; The high-resolution industrial camera continuously captures the construction process of the slope anchor drilling platform at a frequency of 5 frames per second, and records image timestamps and spatial position information; The collected image data is losslessly compressed and stored, and a unique identification code is assigned to each image. An associated index with the construction log is established to obtain the original image data set with time and space tags.

[0013] Specifically, at least four positioning markers are set up around the construction area of ​​the slope anchor drilling platform to construct a spatial coordinate system for precise three-dimensional positioning. These positioning markers are typically made of specially designed highly reflective material, ensuring clear identification at all angles and in all lighting conditions. These four markers are placed around the construction area to form a stereoscopic reference system. Each marker has precise three-dimensional coordinates, determined using a total station or RTK-GPS measurement. This arrangement ensures that at least three markers appear simultaneously in the image, regardless of the angle from which the image is taken, providing a reference for subsequent geometric correction of the image. At least three 12-megapixel industrial cameras with global shutter CCD sensors are deployed around the construction area of ​​the slope anchor drilling platform, forming a multi-angle coverage network to obtain comprehensive construction information. Global shutter CCD sensors simultaneously expose the entire sensor, preventing distortion of moving objects in the image, making them particularly suitable for capturing high-speed moving parts during the drilling process. These cameras are typically arranged in a 120° triangular array to monitor the drilling platform from multiple angles, eliminating occlusion issues caused by a single viewing angle. Each camera is equipped with a remote control unit to achieve synchronous parameter adjustment.

[0014] High-resolution industrial cameras continuously capture the slope anchor drilling platform's construction process at 5 frames per second. Recording image timestamps and spatial position information is key to achieving precise spatiotemporal positioning. Timestamps are recorded with millisecond accuracy and synchronized with a global time server to ensure time consistency across multiple cameras. Spatial position information, including the camera's six-degree-of-freedom attitude parameters (three position parameters and three angle parameters), is acquired in real time by an inertial measurement unit mounted externally to the camera. This high-frequency acquisition ensures that instantaneous changes during drilling are captured, particularly the critical moments of the drill bit entering and exiting the rock. The captured image data is stored losslessly, and each image is assigned a unique identification code. This is then indexed and linked to the construction log, resulting in a raw image dataset with spatiotemporal tags, which is the foundation for data management. Lossless compression is performed in PNG format to ensure image quality. The unique identification code, consisting of the camera ID, acquisition timestamp, and serial number, is formatted as "CAM01-20250519-153045-0001," ensuring that each image is uniquely identified. The association with construction logs is achieved through time matching, linking images with construction parameters recorded at the same time, such as drilling depth, drilling pressure, and drilling speed, to form a complete spatiotemporal dataset, facilitating subsequent data mining and analysis. This data organization method solves the problems of data fragmentation and poor correlation in traditional construction monitoring.

[0015] In one embodiment, the correction module 102 is configured to: Applying a Gaussian filter algorithm to the original image data set to perform noise reduction processing to eliminate image noise caused by environmental factors; Performing contrast enhancement on the original image data set after noise reduction by using a histogram equalization method to improve image detail performance; Calculating a perspective transformation matrix based on information of positioning marker points set around the construction area of ​​the slope anchor drilling platform to correct image distortion; The corrected images are spatially registered, and images collected at different angles are mapped into the same coordinate system to generate the standardized drilling platform image.

[0016] Specifically, correction module 102 applies a Gaussian filter algorithm to the original image dataset to perform noise reduction, eliminating image noise caused by environmental factors. Gaussian filtering is a linear smoothing filter algorithm that achieves image smoothing by convolving with a normal (Gaussian) distribution function. At construction sites, factors such as dust, vibration, and lighting variations can introduce noise into images, affecting subsequent image analysis. The Gaussian filter algorithm uses a two-dimensional Gaussian function as the convolution kernel and performs a weighted average of each pixel in the image and its neighborhood. The weights are determined by the Gaussian function, with pixels closer to the center receiving a greater weight. In implementation, a convolution kernel size of 5×5 and a standard deviation of σ=1.5 were selected. This configuration effectively removes salt and pepper noise and Gaussian noise, common in construction environments, while preserving image edge detail. Histogram equalization is then used to enhance contrast on the denoised original image dataset to improve image detail. Histogram equalization is a nonlinear image enhancement technique that redistributes image brightness values ​​to make the image histogram more uniform across the entire brightness range. In slope construction environments, lighting conditions are often suboptimal, resulting in insufficient image contrast and difficulty distinguishing between borehole edges and platform structures. Histogram equalization calculates the image's cumulative distribution function and then performs grayscale mapping, redistributing the original image's grayscale values ​​to a range of 0-255. This brightens dark areas and appropriately compresses bright areas, thereby enhancing local contrast. To address the unique characteristics of images from slope anchor drilling platform construction, an adaptive histogram equalization method was employed. The image was divided into multiple sub-blocks, each of which was then equalized separately. The results were then combined through bilinear interpolation, effectively addressing the over-enhancement and noise amplification issues that can result from global equalization.

[0017] The perspective transformation matrix is ​​calculated based on the positioning markers set around the construction area of ​​the slope anchor drilling platform to correct image distortion. Camera lens distortion can cause straight lines to appear as curves in the image, affecting the accuracy of geometric measurements. Using the four pre-set positioning markers as a reference, a 3×3 perspective transformation matrix is ​​calculated based on the correspondence between the two-dimensional coordinates of the markers in the image and the actual three-dimensional coordinates. This process first uses a feature matching algorithm to automatically detect the positions of the markers in the image, then solves the parameters of the perspective transformation matrix using the least squares method. Finally, this matrix is ​​applied to transform the entire image to correct radial and tangential distortion. This step ensures that straight line elements in the image, such as the edge of the drilling platform, retain their straight line characteristics after correction, laying the foundation for subsequent dimensional measurement.

[0018] Spatial registration is performed on the rectified images, mapping images captured at different angles into the same coordinate system to generate a standardized image of the drilling platform. Spatial registration is a key step in multi-view image fusion, establishing the correspondence between images captured by different cameras. This process first extracts and matches feature points from images at different angles using a feature point matching algorithm (such as SIFT or ORB). False matches are then eliminated using the RANSAC algorithm to calculate the relative pose between the cameras. Based on this information and combined with the previously obtained perspective transformation matrix, all images are transformed to a unified "virtual camera" perspective, generating a complete, unobstructed, standardized image of the drilling platform. This standardized image eliminates geometric distortion caused by perspective differences, ensuring measurement consistency in subsequent geometric feature extraction and providing accurate image correspondence for 3D reconstruction.

[0019] In a specific embodiment, the extraction module 103 is configured to: Applying the Canny edge detection algorithm to the standardized drilling platform image to extract the contour, setting a double threshold parameter, with the low threshold being 40% of the high threshold; Inputting the extracted contour into a Hough transform algorithm to identify straight line and circular structural elements, and locate the boundary of the slope anchor drilling platform and the drilling position; Based on the matching feature points of the standardized drilling platform images at different angles, point cloud data is constructed using structured light 3D reconstruction technology; Surface fitting and gridding are performed on the point cloud data to establish a three-dimensional spatial model of the drilling platform including the size, drilling position, and drilling angle of the slope anchor drilling platform.

[0020] Specifically, extraction module 103 first applies the Canny edge detection algorithm to the standardized drilling platform image to extract the contours, setting dual threshold parameters, with the low threshold set at 40% of the high threshold. Canny edge detection is a multi-stage edge detection algorithm particularly suitable for extracting the contours of clearly structured engineering objects such as drilling platforms. The algorithm first applies Gaussian filtering to smooth the image, then calculates the image's gradient magnitude and direction. Non-maximum suppression is then performed to obtain edges with a single pixel width, and finally, edge connection is performed using a dual thresholding method. In the application scenario of slope anchor drilling platforms, the high threshold is typically set at the 70th percentile of the gradient magnitude, while the low threshold is set at 40% of the high threshold. This setting effectively suppresses noise interference while maintaining detection sensitivity. The algorithm outputs a binary edge image that clearly displays the platform outline and borehole boundary, providing basic data for subsequent geometric feature recognition. The extracted contours are then input into the Hough transform algorithm to identify linear and circular structural elements and locate the slope anchor drilling platform boundary and drillhole location. The Hough transform is a feature extraction technique that can detect parameterized shapes such as lines and circles in an image. To detect the drilling platform's boundaries, the Hough line transform is used to transform points in the binary edge image into the Hough parameter space (ρ, θ), where ρ represents the perpendicular distance from the origin to the line, and θ represents the angle between the perpendicular line and the x-axis. By searching for local maxima in the accumulator array in the parameter space, the main lines in the image are identified; these lines constitute the drilling platform's boundaries. To detect the drilling hole's location, the Hough circle transform is used, searching in the three-dimensional parameter space (a, b, r), where (a, b) are the coordinates of the circle's center and r is the radius. The algorithm then votes on the parameter space based on the edge points and finds the parameters corresponding to the local maxima in the accumulator array, thereby determining the location and diameter of the drill hole. This method is effective in detecting drill holes even when they are partially obscured or affected by lighting.

[0021] Based on the matching feature points of standardized drilling platform images taken at different angles, point cloud data is constructed using structured light 3D reconstruction technology. Structured light 3D reconstruction is an active vision technology that restores 3D information by projecting a known pattern of light onto a scene and then analyzing the deformation of the light pattern. In this module, standardized images acquired at different angles are first used to establish pixel-level correspondences between image pairs using a feature matching algorithm (such as SIFT), identifying the projected positions of the same physical point in different images. Then, based on the camera's internal and external parameters and the established correspondences, the coordinates of the point in 3D space are calculated using the principle of triangulation. For complex objects such as slope anchor drilling platforms, a multi-view fusion strategy is adopted to align and fuse point cloud data from different viewpoints using an iterative closest point (ICP) algorithm, eliminating blind spots and obtaining dense point cloud data covering the entire platform. These point clouds accurately describe the platform's geometric shape and surface features.

[0022] Surface fitting and meshing are performed on the point cloud data to construct a 3D model of the drilling platform, including its dimensions, drill hole locations, and drill angles. The point cloud data inherently contains noise and outliers, requiring processing to construct a high-quality 3D model. First, the point cloud is denoised and downsampled. A statistical outlier filter algorithm is used to remove noise points. Then, a voxel grid filter is applied for uniform downsampling, reducing the data volume while preserving geometric detail. Surface reconstruction is then performed. The RANSAC plane fitting algorithm is applied to the platform's planar portion to extract key planes, and a cylinder fitting algorithm is applied to the drill hole portion to identify the drill hole axis and diameter. Finally, a complete triangulated mesh model is generated using the Poisson surface reconstruction algorithm. Mesh optimization, including mesh simplification and smoothing, is performed to produce an accurate and efficient 3D model. This model accurately records key geometric information, including the platform's length, width, and height, the spatial coordinates of the drill holes, and the angle between the drill hole axis and the horizontal plane, laying the foundation for subsequent drilling trajectory monitoring.

[0023] In one embodiment, the monitoring module 104 is configured to: Setting an ideal state vector of the drilling trajectory according to the three-dimensional spatial model of the drilling platform, including the drilling starting point position, target depth, design angle and direction; The drill rod position sensor and depth sensor are used to collect the real-time motion parameters of the drill rod and construct the observation vector; The observation vector is processed by the improved adaptive Kalman filter algorithm, wherein the state noise covariance matrix is ​​adjusted in real time according to geological parameters; The lateral deviation and angular deviation between the filtered actual drilling trajectory and the ideal trajectory are calculated to generate the drilling trajectory deviation data.

[0024] Specifically, the monitoring module 104 first sets the ideal state vector for the drilling trajectory based on the three-dimensional spatial model of the drilling platform. This vector includes the starting point position, target depth, design angle, and direction. The ideal state vector is a mathematical representation of the drilling design parameters and serves as a benchmark for subsequent actual trajectory comparisons. In slope anchor drilling projects, the ideal trajectory of each hole is a straight line from the starting point to the target point. The starting point is determined by the surface coordinates of the drilling platform extracted from the three-dimensional spatial model and is typically expressed in a global coordinate system. The design angle includes the angle between the hole and the horizontal plane and the angle between the hole's projection on the horizontal plane and the north direction. These two angles together determine the spatial direction of the hole. The target depth is the distance from the starting point along the hole axis. Based on these parameters, the ideal state vector is constructed as a six-dimensional vector, describing the ideal spatial trajectory that the hole should follow under the design requirements. The drill rod position sensor and depth sensor collect real-time drill rod motion parameters to construct an observation vector. The drill rod position sensor typically includes an inclination sensor and an azimuth sensor installed on the drilling rig, which measure the angle between the drill rod and the vertical and the direction of the drill rod's projection on the horizontal plane, respectively. Depth sensors monitor the feed length of the drill pipe, typically using encoders or displacement sensors. These sensors collect data at a high frequency, ensuring they can capture even subtle changes during drilling. The observation vector, composed of these real-time measurements, represents the current drill head position coordinates, the current inclination and azimuth of the drill pipe, and the current drilling depth. Each parameter in the observation vector contains measurement noise, which requires filtering algorithms to obtain accurate trajectory information.

[0025] The observation vectors are processed using an improved adaptive Kalman filter algorithm, in which the state noise covariance matrix is ​​adjusted in real time based on geological parameters. The standard Kalman filter algorithm assumes constant system and measurement noise. However, during actual drilling, the noise characteristics also change due to changing geological conditions. The improved adaptive Kalman filter algorithm incorporates a dynamic adjustment mechanism for the covariance matrix, enabling adaptive updates based on geological parameters and real-time feedback. The algorithm first establishes a state transition equation and an observation equation. The former describes the dynamic changes in the drill bit's position and attitude, while the latter describes the relationship between sensor measurements and the actual state. The algorithm then proceeds through two main steps: prediction and update. The prediction step predicts the current state based on the previous state and the state transition equation; the update step corrects the prediction based on observation data. The improvement lies in the introduction of geological parameter influencing factors. When the drill bit encounters hard rock formations, the state noise covariance is increased to accommodate potential trajectory deviations; when drilling into soft rock formations, the covariance is reduced to improve filtering accuracy. This adaptive mechanism significantly improves the accuracy of drilling trajectory monitoring in complex geological conditions.

[0026] The lateral and angular deviations between the filtered actual drilling trajectory and the ideal trajectory are calculated to generate drilling trajectory deviation data. Lateral deviation is the shortest distance from the actual drilling trajectory point to the ideal trajectory line, calculated using the point-to-line distance formula. During drilling, the lateral deviation value is recorded every 10 centimeters of drilling depth to construct a complete deviation curve. Angular deviation is the angle between the actual drilling direction and the designed direction. The current drilling direction vector is obtained by fitting three consecutive drilling points, and the angle is then calculated with the designed direction vector. Drilling trajectory deviation data contains three key pieces of information: depth, lateral deviation, and angular deviation, forming a complete drilling quality monitoring dataset. This dataset not only reflects the current drilling quality status but also predicts the final hole quality through trend analysis, providing a decision-making basis for timely adjustment of drilling parameters. This real-time monitoring and analysis effectively solves the problem of delayed detection of quality issues in traditional drilling construction, significantly improving the construction quality and efficiency of slope anchor drilling.

[0027] In a specific embodiment, the calculation module 105 is configured to: Applying the least square method to the drilling trajectory deviation data to fit the drilling center line, calculating the maximum distance value from the actual drilling trajectory point to the fitting line, and determining the drilling straightness; The drilling depth is measured by analyzing the drill rod feed length and the drill bit position change, and compared with the designed depth; Based on the edge detection results, the borehole diameter is measured at different depths and a borehole diameter change curve is constructed; The gray level co-occurrence matrix method is used to analyze the texture characteristics of the borehole wall image, calculate texture parameters including energy, entropy, and contrast, evaluate the roughness of the borehole wall, and form the drilling quality evaluation index.

[0028] Specifically, the calculation module 105 first applies the least squares method to the drilling trajectory deviation data to fit the drilling centerline. The maximum distance between the actual drilling trajectory point and the fitted line is calculated to determine the drilling straightness. The least squares method is a mathematical optimization technique that seeks the optimal function matching the data by minimizing the sum of squared errors. During the drilling centerline fitting process, the set of spatial points in the drilling trajectory deviation data is used as input. An objective function is constructed as the sum of the squared distances from all points to the line to be fitted. The optimal drilling centerline is obtained by solving the line parameters that minimize this objective function. After fitting is complete, the distance from each point on the actual drilling trajectory to the fitted centerline is calculated, and the maximum value is taken as the drilling straightness indicator. Drilling straightness directly reflects the degree of curvature of the drilling trajectory and is an important parameter for evaluating drilling quality. The drilling depth is measured by analyzing the drill rod feed length and the change in drill bit position and compared with the designed depth. The drill rod feed length is directly measured by a displacement sensor installed on the drilling rig, recording the distance the drill rod moves from its initial position. The drill bit position change is detected by the position sensor in the aforementioned monitoring module, recording the real-time coordinates of the drill bit in three-dimensional space. The drilling depth calculation takes these two data sources into account. By comparing the drill rod feed length with the projected distance of the drill bit's three-dimensional position change, it eliminates errors caused by drill rod bending or expansion. The calculated actual drilling depth is compared with the target depth specified in the design documents to assess the depth error and serve as the basis for drilling quality evaluation.

[0029] Based on the edge detection results, the borehole diameter is measured at different depths and a curve of the change in borehole diameter is constructed. This step uses the edge detection image obtained in the extraction module to process sections at different depths. For each depth section image, the Hough circle transform algorithm is applied to identify the borehole boundary, calculate the center position and radius, and thus obtain the borehole diameter at that depth. To ensure measurement accuracy, multi-angle images are collected at each depth position, and the multi-view measurement results are integrated to eliminate occlusions and errors that may be caused by a single view. By measuring the diameter values ​​at fixed intervals (usually 20 cm) throughout the entire length of the borehole, a curve reflecting the change in borehole diameter with depth is constructed. This curve can clearly show whether there are quality problems such as taper, necking or reaming in the borehole.

[0030] The gray-level co-occurrence matrix method is used to analyze the texture characteristics of borehole wall images. Texture parameters including energy, entropy, and contrast are calculated to assess the roughness of the borehole wall and form a drilling quality evaluation index. The gray-level co-occurrence matrix is ​​a statistical method for describing image texture characteristics. It constructs a matrix reflecting texture information by calculating the co-occurrence frequency of grayscale values ​​of pixel pairs with specific positional relationships in the image. For the borehole wall image, the distance and direction relationship between pixel pairs is first defined. Typically, four directions (upward, downward, leftward, rightward, and diagonal) are selected, and the corresponding gray-level co-occurrence matrix is ​​calculated within a distance range of 1-3 pixels. Based on the gray-level co-occurrence matrix, second-order statistical features such as energy, entropy, and contrast are calculated. Energy reflects the uniformity of the texture; a higher value indicates a more uniform texture; entropy reflects the complexity of the texture; a higher value indicates a more complex texture; and contrast reflects the clarity of the texture; a higher value indicates a more distinct texture boundary. These parameters comprehensively assess the roughness of the borehole wall. Excessive roughness may affect the adhesion of the anchor bolt to the hole wall, while low roughness may lead to poor grouting results.

[0031] In one embodiment, the analysis module 106 is configured to: The anchor bolt installation process is monitored by a target tracking algorithm, and the anchor bolt position, posture and grouting parameter changes are recorded to obtain the anchor bolt installation process monitoring data; Input the drilling quality evaluation index and the anchor installation process monitoring data into a hierarchical analysis model to establish a quality evaluation system; Applying the entropy weight method to assign weight coefficients to each indicator in the hierarchical analysis model, and calculating the comprehensive quality score of each anchor drilling hole; Based on the comprehensive quality score, four-level warning thresholds are set, and the quality level is divided into four levels: excellent, qualified, critical and unqualified, to generate the construction quality level and warning information of the slope anchor drilling platform.

[0032] Specifically, the analysis module 106 first monitors the anchor installation process using a target tracking algorithm, recording the anchor position, posture, and changes in grouting parameters to obtain anchor installation process monitoring data. The target tracking algorithm is a computer vision technique used to track moving targets. A feature point-based tracking method is used to monitor the anchor installation process. This method first identifies distinctive marker points on the anchor, such as the junction between the anchor head and tail and the boundary of the anti-corrosion layer. The motion trajectory of these feature points is then tracked in a continuous image sequence. The anchor's movement speed, insertion depth, and posture angle are calculated based on the spatial position changes of the feature points. Simultaneously, image recognition technology is used to read the pressure gauge and flow meter values ​​on the grouting equipment to obtain parameters such as grouting pressure, grouting volume, and grouting speed. This data is recorded in a time series format, forming a complete anchor installation process monitoring dataset that comprehensively reflects the quality status of the anchor installation. The drilling quality evaluation indicators and anchor installation process monitoring data are input into a hierarchical analysis model to establish a quality evaluation system. The hierarchical analysis model is a multi-criteria decision-making method that simplifies the decision-making process by decomposing complex problems into a hierarchical structure. In evaluating the construction quality of slope anchor drilling platforms, the model constructs a three-tiered structure: the top tier represents the overall goal, or a comprehensive evaluation of construction quality; the middle tier comprises primary indicators, encompassing four categories: platform geometry, drilling trajectory, drilling quality, and anchor installation parameters; and the bottom tier comprises secondary indicators, detailing the specific evaluation criteria for each parameter. Platform geometry includes platform levelness and stability; drilling trajectory parameters include straightness and angular deviation; drilling quality parameters include depth, diameter, and wall roughness; and anchor installation parameters include anchor position accuracy, grouting fullness, and anchoring force. Each indicator has clear quantitative standards and evaluation criteria, forming a comprehensive quality evaluation system.

[0033] The entropy weight method was applied to assign weight coefficients to each indicator in the hierarchical analysis model (AHP) and calculate the comprehensive quality score for each anchor borehole. The entropy weight method is an objective weighting method based on information entropy theory, which determines weights by calculating the information entropy of indicators. A greater information entropy indicates greater uncertainty in the indicator, less impact on decision-making, and a correspondingly smaller weight. The specific process involves: first, standardizing each indicator value to eliminate dimensionality; then calculating the information entropy of each indicator, which takes into account the distribution of the indicator across all evaluation objects; then, calculating the entropy weight based on the information entropy, which is inversely proportional to the information entropy; finally, combining the AHP structure and, in a bottom-up order, aggregating the entropy weights of the secondary indicators to form the weights of the primary indicators. The entropy weights of the primary indicators are then aggregated to form the weights for the overall evaluation. Based on the determined weight coefficients and the standardized indicator values, a weighted summation is used to calculate the comprehensive quality score for each anchor borehole, which comprehensively reflects the overall quality of the drilling and anchor installation.

[0034] Based on the comprehensive quality score, four warning thresholds are set, categorizing the quality level into excellent, acceptable, marginal, and unacceptable. This generates quality grades and warning information for the slope anchor drilling platform. The warning thresholds are based on extensive engineering practice data and quality standards. Typically, scores above 90 are considered excellent, 75-89 are acceptable, 60-74 are marginal, and scores below 60 are unacceptable. The system generates corresponding warning messages for each level: excellent requires no special action; acceptable provides general acceptance recommendations; marginal issues issue a yellow warning, indicating the need for significant attention and necessary reinforcement measures; and unacceptable issues issue a red warning, requiring subsequent work to be halted and rework or reinforcement measures performed. The warning messages also include specific problem descriptions and recommended solutions, such as recommending back-drilling for insufficient straightness and additional grouting for incomplete grouting. The system also generates a quality distribution map, visually displaying the quality status of each borehole in the project, helping construction personnel quickly identify problem areas and conduct targeted quality control.

[0035] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A machine vision-based intelligent detection system for the construction quality of a slope anchor drilling platform, characterized in that the system include: The acquisition module is used to collect multi-angle images of the slope anchor drilling platform construction area using a high-resolution industrial camera to obtain a raw image dataset with time and space tags; A correction module, configured to perform image preprocessing and geometric correction on the original image data set to obtain a standardized drilling platform image; an extraction module, configured to extract geometric characteristic parameters of the drilling platform based on the standardized drilling platform image to obtain a three-dimensional spatial model of the drilling platform; A monitoring module, configured to apply an improved adaptive Kalman filter algorithm to the three-dimensional spatial model of the drilling platform to perform real-time monitoring of the drilling trajectory and obtain drilling trajectory deviation data; a calculation module, configured to calculate key drilling quality parameters based on the drilling trajectory deviation data to obtain a drilling quality evaluation index; The analysis module is used to comprehensively analyze the drilling quality evaluation index and the anchor installation process monitoring data to obtain the slope anchor drilling platform construction quality grade and early warning information.

2. The machine vision-based intelligent detection system for slope anchor drilling platform construction quality according to claim 1 is characterized in that: The acquisition module is used to: At least four positioning marking points are set around the construction area of ​​the slope anchor drilling platform to construct a spatial coordinate system; Arrange at least three 12-megapixel global shutter CCD sensor industrial cameras in the construction area of ​​the slope anchor drilling platform to form a multi-angle coverage network; The high-resolution industrial camera continuously captures the construction process of the slope anchor drilling platform at a frequency of 5 frames per second, and records image timestamps and spatial position information; The collected image data is losslessly compressed and stored, and a unique identification code is assigned to each image. An associated index with the construction log is established to obtain the original image data set with time and space tags.

3. The machine vision-based intelligent detection system for slope anchor drilling platform construction quality according to claim 1 is characterized in that: The correction module is used to: Applying a Gaussian filter algorithm to the original image data set to perform noise reduction processing to eliminate image noise caused by environmental factors; Performing contrast enhancement on the original image data set after noise reduction by using a histogram equalization method to improve image detail performance; Calculating a perspective transformation matrix based on information of positioning marker points set around the construction area of ​​the slope anchor drilling platform to correct image distortion; The corrected images are spatially registered, and images collected at different angles are mapped into the same coordinate system to generate the standardized drilling platform image.

4. The machine vision-based intelligent detection system for slope anchor drilling platform construction quality according to claim 1 is characterized in that: The extraction module is used to: Applying the Canny edge detection algorithm to the standardized drilling platform image to extract the contour, setting a double threshold parameter, with the low threshold being 40% of the high threshold; Inputting the extracted contour into a Hough transform algorithm to identify straight line and circular structural elements, and locate the boundary of the slope anchor drilling platform and the drilling position; Based on the matching feature points of the standardized drilling platform images at different angles, point cloud data is constructed using structured light 3D reconstruction technology; Surface fitting and gridding are performed on the point cloud data to establish a three-dimensional spatial model of the drilling platform including the size, drilling position, and drilling angle of the slope anchor drilling platform.

5. The machine vision-based intelligent detection system for slope anchor drilling platform construction quality according to claim 1 is characterized in that: The monitoring module is used to: Setting an ideal state vector of the drilling trajectory according to the three-dimensional spatial model of the drilling platform, including the drilling starting point position, target depth, design angle and direction; The drill rod position sensor and depth sensor are used to collect the real-time motion parameters of the drill rod and construct the observation vector; The observation vector is processed by the improved adaptive Kalman filter algorithm, wherein the state noise covariance matrix is ​​adjusted in real time according to geological parameters; The lateral deviation and angular deviation between the filtered actual drilling trajectory and the ideal trajectory are calculated to generate the drilling trajectory deviation data.

6. The machine vision-based intelligent detection system for slope anchor drilling platform construction quality according to claim 1 is characterized in that: The computing module is configured to: Applying the least square method to the drilling trajectory deviation data to fit the drilling center line, calculating the maximum distance value from the actual drilling trajectory point to the fitting line, and determining the drilling straightness; The drilling depth is measured by analyzing the drill rod feed length and the drill bit position change, and compared with the designed depth; Based on the edge detection results, the borehole diameter is measured at different depths and a borehole diameter change curve is constructed; The gray level co-occurrence matrix method is used to analyze the texture characteristics of the borehole wall image, calculate texture parameters including energy, entropy, and contrast, evaluate the roughness of the borehole wall, and form the drilling quality evaluation index.

7. The machine vision-based intelligent detection system for slope anchor drilling platform construction quality according to claim 1 is characterized in that: The analysis module is used to: The anchor bolt installation process is monitored by a target tracking algorithm, and the anchor bolt position, posture and grouting parameter changes are recorded to obtain the anchor bolt installation process monitoring data; Input the drilling quality evaluation index and the anchor installation process monitoring data into a hierarchical analysis model to establish a quality evaluation system; Applying the entropy weight method to assign weight coefficients to each indicator in the hierarchical analysis model, and calculating the comprehensive quality score of each anchor drilling hole; Based on the comprehensive quality score, four-level warning thresholds are set, and the quality level is divided into four levels: excellent, qualified, critical and unqualified, to generate the construction quality level and warning information of the slope anchor drilling platform.

Images