Roof bolter automatic hole finding and positioning system and method based on machine vision

Abstract

The invention provides an anchor rod drilling machine automatic hole finding and positioning system and method based on machine vision, and belongs to the technical field of anchor rod drilling machine automatic hole finding and positioning. In order to solve the technical problems of low manual operation precision, poor efficiency and poor automation scheme adaptability during anchor rod supporting, the adopted technical scheme is as follows: a binocular vision camera and a laser displacement sensor are arranged on an anchor rod drilling machine vehicle body; a drilling machine controller, a tilt angle sensor, an industrial switch, a drilling machine electric control box and a communication module are arranged in a jumbolter body, the drilling machine controller is connected with other modules or components through wires, a positioning target and a wireless AP base station are further arranged in a roadway, and the communication module is in wireless connection with the wireless AP base station through a wireless network. A positioning console is arranged in the control room, and a positioning control computer, a core switch and a drilling positioning server are further arranged in the positioning console; the method is applied to automatic hole finding and positioning of the jumbolter.

Description

Technical Field

[0001] The present invention provides a system and method for automatic hole finding and positioning of an anchor drill based on machine vision, belonging to the technical field of automatic hole finding and positioning of an anchor drill. Background Art

[0002] In mine tunnel support operations, anchor support is an important means to ensure the stability of the surrounding rock. Before anchor support is carried out, it is necessary to drill holes in the support area. The current manual drilling method has the following main defects:

[0003] Low positioning accuracy: Manually marking the drilling position is affected by the operator's experience, and the error usually exceeds ±10mm, which is difficult to meet the high-precision support requirements;

[0004] Low efficiency: Each hole positioning takes 3-5 minutes, which will seriously affect the construction progress in long-distance tunnels;

[0005] Poor environmental adaptability: The dusty, humid and low-light environment underground makes manual observation difficult, and problems such as drilling errors and missed drilling are prone to occur.

[0006] In response to the above-mentioned defects, corresponding automatic anchor hole positioning methods have been proposed, such as laser hole positioning, monocular vision positioning, and mechanical limit device drilling. However, the laser guidance solution relies on preset coordinates and cannot adapt to rock surface deformation. The error increases significantly in broken surrounding rock. Monocular vision positioning is limited by two-dimensional image information and cannot accurately solve the spatial position relationship between the drill bit and the hole position. The mechanical limit device has a limited adjustment range and cannot achieve dynamic compensation under complex rock surfaces. Summary of the Invention

[0007] In order to solve the technical problems existing in the background technology, the present invention adopts the following technical solutions: providing an automatic hole-finding and positioning system for an anchor drill rig based on machine vision, comprising an anchor drill rig body and a control room arranged in a tunnel, a binocular vision camera and a laser displacement sensor being arranged on the anchor drill rig body, and a drill controller, an inclination sensor, an industrial switch, an electric control box for the drill rig, and a communication module being arranged inside the anchor drill rig body;

[0008] Positioning targets and wireless AP base stations are also set up in the lanes;

[0009] The control room is provided with a positioning console, which is also provided with a positioning control computer, a core switch, and a drilling positioning server;

[0010] The drilling rig controller is connected to the binocular vision camera, the laser displacement sensor, the tilt sensor, the industrial switch, the drilling rig electric control box, and the communication module through wires;

[0011] The communication module is wirelessly connected to the wireless AP base station via a wireless network;

[0012] The wireless AP base station is connected to the core switch via a wire, and the core switch is connected to the positioning control computer and the drilling positioning server via wires respectively;

[0013] The wireless AP base station is wirelessly connected to the positioning target via a wireless network.

[0014] The anchor drilling rig body is also provided with an explosion-proof lighting lamp, and the control end of the explosion-proof lighting lamp is connected to the industrial switch through a wire.

[0015] The binocular vision camera includes a left camera and a right camera, and the left camera and the right camera are specifically arranged horizontally on the drilling rig crossbeam of the anchor drilling rig body through a shock-absorbing bracket.

[0016] The laser displacement sensor is specifically installed on a bracket on the side of the drill head of the anchor drilling rig body;

[0017] The measuring axis of the laser displacement sensor is arranged to form an angle of 30° with the drill rod.

[0018] The drilling rig controller is specifically connected to the drilling rig electric control box via a PROFINET bus, or via an EtherCAT network cable, or via a Modbus RTU bus.

[0019] A positioning method for an automatic hole-finding and positioning system of an anchor drilling rig based on machine vision includes the following positioning steps:

[0020] Step 1: Control the drilling rig controller to receive data collected by the laser displacement sensor and the inclination sensor respectively, and the drilling rig controller analyzes and processes the collected data and sends it to the communication module;

[0021] Step 2: Control the drilling rig controller to communicate data with the drilling rig electric control box, obtain the operating parameters of the drilling rig, and send the operating parameters of the drilling rig to the communication module;

[0022] The obtained operating parameters include the drill speed ω and the feed speed v, which are defined as satisfying ω=kv, where k is the material correlation coefficient;

[0023] Step 3: Control the drilling rig controller to receive the image data collected by the binocular vision camera and perform preprocessing, including image binarization, image enhancement, and stereo correction processing;

[0024] Step 4: The control communication module communicates data with the core switch set in the control room through the wireless AP base station;

[0025] Step 5: Control the positioning target set in the tunnel to send the position parameters to the wireless AP base station in real time through the wireless network;

[0026] Step 6: Estimate the acquisition distance through the structural parameters of the anchor drilling robot and the transformation relationship between the coordinate axes, and calculate the spatial distance between the camera position and the target when acquiring the image;

[0027] Step 7: Using the functional relationship between the acquisition distance and the anchor hole radius in the image, determine the radius of the anchor hole at any distance, and then determine the center coordinates and the maximum and minimum radius of the anchor hole;

[0028] Step 8: Based on the geometric constraints of the maximum and minimum radius of the anchor hole, the anchor holes on the steel strip are detected using Hough transform with adaptive radius parameter adjustment. The linear slope and geometric constraints are used to achieve stereo matching of the anchor hole contours.

[0029] Step 9: Determine the spatial coordinates of the anchor hole center in the camera coordinate system based on the binocular vision positioning principle;

[0030] Step 10: The anchor drilling rig body launches and installs the anchor into the corresponding anchor hole according to the determined spatial coordinates of the anchor hole center.

[0031] The specific method of step six is:

[0032] Based on the conversion relationship of each coordinate system, the conversion relationship between the camera coordinate system and the fuselage coordinate system is obtained, and the three-dimensional coordinates of the camera coordinate system in the world coordinate system are obtained in real time through the fuselage positioning method;

[0033] Construct the camera coordinate system, reference coordinate system, and world coordinate system respectively:

[0034] Among them, the spatial coordinates of point A on the camera plane in the camera coordinate system are defined as C A=(x1,y1,z1), and convert it to the world coordinate system W The coordinates in A=(x2,y2,z2) are converted as follows:

[0035] ;

[0036] in, Represents the conversion relationship between the camera coordinate system and the body coordinate system, and Indicates the conversion relationship between the fuselage coordinate system and the world coordinate system;

[0037] Define the coordinates of point B on the target in the world coordinate system as WB=(x3,y(3),z(3)), where x(3) and y(3) are unknowns;

[0038] The acquisition distance can be approximately expressed as the distance between point A on the camera and point B on the steel belt, expressed as:

[0039] ;

[0040] In order to reduce the impact of the difference between the X-axis and the Y-axis, the attenuation coefficient α is introduced, and the calculation formula of the spatial distance is:

[0041] .

[0042] The specific method of step seven is:

[0043] Based on step 6, the target contour is fitted. To compensate for the distance between the camera and the target that affects the center distance between the targets, the following compensation rules are used:

[0044] ;

[0045] Among them, x i and y i represents the coordinates of the i-th target in the image.

[0046] The specific method of step eight is:

[0047] Obtain the target center coordinates through parameter-adaptive Hough circle detection:

[0048] The circle function is defined as follows:

[0049] x=x0+r0cosα;

[0050] y=y0+r0sinα;

[0051] Where r0 is the target radius, x0 and y0 are the target center coordinates, x and y are the image coordinates, and α represents the angle, which ranges from 0° to 360°.

[0052] Since the contours on the image are oriented, the coordinates of the circle centers can be arranged from small to large in the horizontal direction and defined as:

[0053] O l1 (u l1 , v l1 ), O l2 (u l2 , v l2 ), O l3 (u l3 , v l3 ), ...,O lm (u lm , v lm ) is the coordinate of the center point of the anchor drill hole in the image captured by the left camera;

[0054] O r1 (u r1, v r1 ), O r2 (u r2 , v r2 ), O r3 (u r3 , v r3 ), ....,O rn (u rn , v rn ) is the coordinate of the center point of the anchor drill hole in the image captured by the right camera;

[0055] The above coordinates are arranged in ascending order of horizontal pixel coordinate values. Based on the fact that the pixel coordinate values ​​of the same point in the vertical direction of the left and right images are the same, stereo matching is performed on different anchor borehole centers to complete the three-dimensional matching of the target center and determine the spatial coordinates of the target center.

[0056] The stereo matching accuracy of the target center is then verified by the directionality of the contour:

[0057] Definition: If the i-th point on the left and the j-th point on the right are a set of matching points, then i±p and j±p on the image are a set of matching points;

[0058] In addition, the distance between two adjacent matching points is basically the same, with a small fluctuation range, and the expression is:

[0059] |O li O li+1 | = O rj O rj+1 .

[0060] The specific method of step nine is:

[0061] The coordinates of the target center are calculated using the ideal model of the binocular vision system:

[0062] Define X separately W 、Y W , Z W is the world coordinate, X C 、Y C , Z C is the camera coordinate, u and v are the pixel plane coordinates; O l and O r is the optical center of the two cameras, B is the distance between the two cameras, and f is the focal length; C l and C r They are perpendicular to the camera optical axis Z C The left image plane and the right image plane of ;

[0063] For stereo correction, point A (X, Y, Z) is a point in the world coordinate system, and its corresponding points in the image plane are a1 (u1, v1) and a2 (u2, v2). The conversion formula between the world coordinate system and the image coordinate system is obtained based on similar triangles:

[0064] ;

[0065] The above formula is used to calculate the coordinates of any point in space, where f ′ is one of the intrinsic parameters obtained through camera calibration;

[0066] By identifying and matching circles, the pixel coordinates in the drill hole image, including those in the left and right images, can be determined;

[0067] According to the calibration results, segmentation and stereo matching results, the spatial coordinates of a point are calculated using the following formula:

[0068] y=kx+b;

[0069] Where k is the slope of the line and b is the vertical coordinate of the origin.

[0070] The beneficial effects of the present invention compared with the prior art are as follows: the present invention proposes a high-precision, anti-interference automatic hole finding and positioning solution for anchor drilling rigs, which uses a monocular camera and a binocular camera to work together, and uses binocular stereo vision to capture the image of the steel belt on the tunnel roof, and performs a series of preprocessing operations such as image enhancement, denoising and stereo correction on the captured image. At the same time, a circle detection algorithm with parameter adaptive Hough transform is adopted, which can automatically adjust the maximum and minimum radii of the anchor hole according to the actual distance of the anchor hole, which helps to accurately and quickly identify and segment the anchor hole contour, and adopts a matching method based on the slope of the straight line where the anchor hole is located and geometric constraints to achieve fast image matching. Compared with manual positioning, the stability and robustness are improved; the present invention solves the problems of low precision and poor efficiency of traditional manual operation and insufficient adaptability of existing automation solutions, and is suitable for underground tunnel support operations such as mines and tunnels. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] The present invention will be further described below with reference to the accompanying drawings:

[0072] Figure 1 This is a schematic diagram of the circuit structure of the automatic hole-finding and positioning system for an anchor drill according to the present invention;

[0073] Figure 2 Schematic diagram of the binocular vision system model used in the present invention to calculate target coordinates;

[0074] The meanings of the serial numbers in the figure are: 1 is the anchor drilling rig body, 2 is the binocular vision camera, 3 is the laser displacement sensor, 4 is the drilling rig controller, 5 is the tilt sensor, 6 is the explosion-proof lighting, 7 is the industrial switch, 9 is the drilling rig electrical control box, 10 is the communication module, 11 is the positioning target, 12 is the positioning console, 13 is the wireless AP base station, 15 is the positioning control computer, 16 is the core switch, and 17 is the drilling positioning server. DETAILED DESCRIPTION

[0075] like Figure 1 and Figure 2 As shown, the present invention provides a machine vision-based automatic hole location solution for anchor drilling rigs. This solution primarily establishes a prediction model to determine the correlation between the target hole radius in the image and the shooting distance. The shooting distance range is determined based on the structural model of the anchor drilling robot and relevant sensor data. The solution then utilizes the inherent geometric constraints of adjacent anchor holes to accurately identify anchor holes using a Hough transformer with an adaptive parameter adjustment method. Furthermore, the solution uses linear slopes and geometric constraints to match anchor hole contours, and uses binocular vision positioning principles to determine the spatial coordinates of the anchor hole center in the camera coordinate system. This achieves high-precision, highly interference-resistant automatic hole location. Experimental verification demonstrates that this method achieves a positioning accuracy of 95.2% with an absolute error of approximately 1.52 mm, effectively improving drilling accuracy and auxiliary efficiency. Furthermore, the proposed method maintains a 95% recognition rate even under dust concentrations of 200 mg / m³, reducing single-hole location time to less than 30 seconds. The proposed method is widely applicable in scenarios such as coal mine roadway support, tunnel anchoring, and underground reinforcement.

[0076] Furthermore, the present invention provides an automatic hole-finding and positioning system for an anchor drill based on machine vision, comprising an anchor drill body 1 disposed in a tunnel and a positioning console 12 disposed in a control room;

[0077] The outer side of the anchor drilling rig body 1 is provided with a binocular vision camera 2, a laser displacement sensor 3, a drilling rig controller 4, a tilt sensor 5, and an explosion-proof lighting lamp 6;

[0078] The interior of the anchor drilling rig body 1 is provided with an industrial switch 7, a drilling rig electric control box 9, and a communication module 10;

[0079] A circular positioning target 11 and a wireless AP base station 13 are also provided in the lane;

[0080] The positioning console 12 is also provided with a positioning control computer 15, a core switch 16, and a drilling positioning server 17;

[0081] The drilling rig controller 4 is connected to the binocular vision camera 2, the laser displacement sensor 3, the tilt sensor 5, the industrial switch 7, and the communication module 10 through wires, and the industrial switch 7 is connected to the explosion-proof lighting lamp 6 through wires;

[0082] The drilling rig controller 4 is specifically connected to the drilling rig electric control box 9 via a PROFINET bus, or via an EtherCAT network cable, or via a ModbusRTU bus;

[0083] The communication module 10 is specifically wirelessly connected to the wireless AP base station 13 via radio electromagnetic waves;

[0084] The wireless AP base station 13 is connected to the core switch 16 via optical fiber;

[0085] The wireless AP base station 13 is specifically wirelessly connected to the positioning target 11 via radio electromagnetic waves;

[0086] The core switch 16 is connected to the positioning control computer 15 and the drilling positioning server 17 through optical fibers.

[0087] Furthermore, the binocular vision camera 2 adopts a 2-megapixel global shutter camera with a frame rate of not less than 30fps and a built-in LED fill light;

[0088] The measurement accuracy of the laser displacement sensor 3 is ±0.1mm, and the measurement distance is 50-300mm;

[0089] The 5G communication module 10 supports the 5G NR standard, has an operating frequency band of 3.5 GHz, and a transmission delay of less than 10 ms;

[0090] The positioning target 11 is made of a high reflectivity material and has a specific geometric pattern mark on its surface;

[0091] The color temperature of the explosion-proof lighting lamp 6 is 5000K, and the adjustable illumination range is 100-1000 lux.

[0092] Based on the above positioning system, the present invention also provides a method for automatic hole finding and positioning of an anchor drilling rig based on machine vision, which specifically includes the following positioning control steps:

[0093] Step 1: A drilling rig controller 4 is set on the anchor drilling rig body 1 to receive data collected by the laser displacement sensor 3 and the inclination sensor 5 respectively, and send the collected data to the communication module 10; the laser displacement data is calculated using the formula d=ct / 2, where c is the speed of light and t is the laser round-trip time.

[0094] Step 2: Control the drill controller 4 to communicate data with the drill electronic control box 9 via the PROFINET bus or Modbus RTU bus, obtain the drill operating parameters, and send the operating parameters to the communication module 10; the operating parameters include the drill speed ω and the feed speed v, satisfying ω=kv, where k is the material correlation coefficient.

[0095] Step 3: Control the drilling rig controller 4 to receive the image data collected by the binocular vision camera 2, use the Otsu method to binarize the image, and use the median filter method to effectively eliminate salt and pepper noise, thereby completing comprehensive image preprocessing.

[0096] Step 4: The control communication module 10 performs data communication with the core switch 16 provided in the control room through the wireless AP base station 13. The data transmission delay Δt≤10ms meets the real-time requirement.

[0097] Step 5: Control the circular positioning target 11 set in the lane to send its position parameters to the wireless AP base station 13 in real time via radio electromagnetic waves; accurately locate the target through parameter-adaptive Hough circle detection, thereby obtaining its precise pixel coordinates. The circle function is defined as follows:

[0098] x=x0+r0cosα;

[0099] y=y0+r0sinα;

[0100] Where r0 is the target radius, x0 and y0 are the target center coordinates, x and y are the image coordinates, and α represents the angle, which ranges from 0° to 360°.

[0101] Step six: Calculate the spatial distance between the camera position and the target when acquiring the image. The joint variables are accurately measured in real time using the displacement sensors and angle sensors arranged on each joint of the manipulator, and the Denavitt-Hartenberg parameter table is constructed based on the manipulator parameters and structural parameters of the drilling and anchoring robot. Through the conversion relationship of each coordinate system, the conversion relationship between the camera coordinate system and the fuselage coordinate system can be obtained. Through the method of fuselage positioning, the three-dimensional coordinates of the camera coordinate system in the world coordinate system can be obtained in real time. The camera coordinate system, the reference coordinate system and the world coordinate system are constructed, where the spatial coordinates of point A on the camera plane in the camera coordinate system are C A=(x1,y1,z1), and convert it to the world coordinate system W The coordinates in A=(x2,y2,z2) are as follows:

[0102] ;

[0103] in Represents the conversion relationship between the camera coordinate system and the body coordinate system, and It represents the transformation relationship between the body coordinate system and the world coordinate system, both of which can be determined according to the robot kinematics principle.

[0104] The size of the coal mine tunnel is fixed. The coordinates of point B on the target in the world coordinate system are WB=(x3,y(3),z(3)), where x(3) and y(3) are unknown. The acquisition distance can be approximately expressed as the distance between point A on the camera and point B on the steel belt:

[0105] ;

[0106] To ensure the steel strip is within the camera's field of view, the X- and Y-axis variations at points A and B are minimal. Experimental results show that the measured distance is primarily determined by the Z-axis variation. To mitigate the effects of the X- and Y-axis variations, an attenuation coefficient, denoted by α, is introduced. Experimental results indicate that α should be 1.1, calculated as:

[0107] ;

[0108] Therefore, the above method is used to calculate the acquisition distance and then determine the maximum and minimum radius.

[0109] Step 7: Get the center coordinates and radius information of the anchor hole. Based on step 6, the target contour fitting can be completed, but considering that the distance between the camera and the target will affect the center distance between the targets, the following rule formula needs to be used:

[0110] ;

[0111] Among them, x i and y i represents the coordinates of the i-th target in the image.

[0112] Based on the above method, the radius parameters can be obtained through the target radius prediction model on the image and adaptively adjusted. The anchor hole contour on the image is fitted using the Hough detection method with adaptive parameter adjustment to obtain the target center coordinates and radius information.

[0113] Step 8: Perform stereo matching. Based on the image segmentation in the above steps, the pixel coordinates of the target center can be determined. Since the contours on the image are directional, the coordinates of the circle center can be arranged from small to large in the horizontal direction and defined as:

[0114] O l1 (u l1 , v l1 ), O l2 (u l2 , v l2 ), O l3 (ul3 , v l3 ), ...,O lm (u lm , v lm ) is the coordinate of the center point of the anchor drill hole in the image captured by the left camera;

[0115] O r1 (u r1 , v r1 ), O r2 (u r2 , v r2 ), O r3 (u r3 , v r3 ), ....,O rn (u rn , v rn ) is the coordinate of the center point of the anchor hole in the image captured by the right camera;

[0116] They are arranged in ascending order of horizontal pixel coordinate values. Theoretically, the pixel coordinate values ​​for the same point in the vertical direction of the left and right images are identical. This property can be used to perform stereo matching of different anchor borehole centers. Since noise can cause certain errors in the corrected image, the corrected image will also contain certain errors. Multiple experiments have shown that the vertical coordinate error of the same point in the left and right images is less than 8 pixels. Therefore, we set the threshold to 10, and any point in the left image can be stereo matched. This method can quickly complete the three-dimensional matching of the target center and determine the spatial coordinates of the target center.

[0117] The accuracy of stereo matching of the target center can be verified based on the directionality of the contour. If the i-th point on the left and the j-th point on the right are a set of matching points, then i±p and j±p on the image are also a set of matching points. In addition, the distance between two adjacent matching points is basically the same, with a small fluctuation range:

[0118] |O li O li+1 | = O rj O rj+1 ;

[0119] Step 9: Calculate the coordinates of the target center:

[0120] The ideal model based on the binocular vision system is as follows Figure 2 As shown, where X is defined W 、Y W , Z W is the world coordinate, X C 、Y C , Z C is the camera coordinate, u and v are the pixel plane coordinates; Ol and O r is the optical center of the two cameras, B is the distance between the two cameras, and f is the focal length; C l and C r Perpendicular to Z C The left and right image planes are aligned with the camera's optical axis. A parallel optical system between the two cameras is ideal, but in reality, most systems do not adhere to this principle. Therefore, stereoscopic correction is an essential step before determining the position of a point in space. Point A (X, Y, Z) is a point in the world coordinate system, and its corresponding points in the image plane are a1(u1, v1) and a2(u2, v2), respectively. Due to the similarity of the triangles, we can express the conversion formula between the world coordinate system and the image coordinate system as follows:

[0121] ;

[0122] The above formula can be used to calculate the coordinates of a point; f′ is one of the intrinsic parameters, obtained through camera calibration. By identifying and matching circles, we can determine the coordinates of pixels in the drilling image, including those in the left and right images. Based on the calibration results, segmentation, and stereo matching, the spatial coordinates of a point are calculated using the formula y=kx+b (k is the slope of the line, b is the ordinate of the origin). Using coordinate transformation, the target's spatial coordinates are converted to the coordinate system of the anchor drilling robot, facilitating precise and rapid movement of the drill toward the target.

[0123] The hole location method proposed in this paper first preprocesses the captured image, including image binarization, image enhancement, and stereo correction. It then uses the Canny edge detection method to process the image and divides the steel strip into regions of interest (ROIs) using upper and lower edge lines as features. Based on this, the acquisition distance is estimated using the transformation relationship between the structural parameters of the anchor drilling robot and the coordinate axes. The radius at any distance is determined using the functional relationship between the acquisition distance and the anchor hole radius in the image. Finally, based on the geometric constraints of maximum and minimum radii, the Hough transform with adaptive radius parameter adjustment is used to detect anchor holes in the steel strip. Experimental results demonstrate that the proposed anchor hole detection method based on the adaptive parameter Hough transform achieves high accuracy and effectively eliminates invalid results.

[0124] This system uses high-precision industrial binocular cameras as its visual acquisition module. The left and right cameras are arranged horizontally, with a precisely calibrated spacing of 250mm. The workflow is as follows: the two cameras are synchronously triggered to capture images. These images are then pre-processed, including denoising and enhancement. Finally, the processed images are transmitted to an industrial switch via a gigabit Ethernet port.

[0125] The laser distance measurement module used in this invention is mounted on a side bracket of the drill bit, with its measuring axis at a 30° angle to the drill rod. The module's operating process involves first emitting a modulated laser beam and receiving the reflected signal, then calculating the real-time distance value based on the phase difference, and finally outputting the measured data via an RS485 interface.

[0126] The central control unit used in the present invention adopts a sandwich structure design. The bottom layer is FPGA, which is responsible for the synchronous acquisition of sensor data; the middle layer is an ARM processor, which runs the control algorithm; and the upper layer is a communication interface module, which is responsible for communicating with external devices.

[0127] The present invention adopts a dual-channel redundant design to ensure the reliability of data transmission. The data transmission path is as follows: sensor data is first transmitted to the industrial switch, then transmitted to the core switch through the 5G module or wireless AP base station, and finally reaches the control computer.

[0128] The positioning system of this invention is mechanically connected: the binocular camera is fixed to the drilling rig's crossbeam via a shock-absorbing bracket to reduce the impact of vibration on image acquisition. The laser sensor is installed using a quick-release interface for easy maintenance and replacement. All cables are protected by drag chains to prevent wear and tear.

[0129] In terms of electrical connections: the sensor uses 24V DC power supply to ensure safety and reliability. The actuators are connected via CAN bus to achieve fast data transmission and timely response to control instructions. The optical fiber backbone network connects to the control room to ensure stable transmission of large amounts of data.

[0130] In terms of data flow: Image data is transmitted from the camera to the switch, where it is initially processed by the FPGA before being transferred to the ARM processor for further processing. Control instructions are transmitted from the ARM processor to the electrical control box via the CAN bus, controlling the movements of the actuators. Status feedback signals are transmitted from the sensor to the PLC, and then to the control computer, achieving closed-loop control of the system.

[0131] The positioning system provided by the present invention mainly performs the following steps when working:

[0132] Step 1: System initialization:

[0133] During the system initialization phase, the binocular camera performs autofocus and exposure calibration, the laser sensor completes temperature compensation, and each module establishes communication connections and performs self-tests.

[0134] Step 2: Environmental Scan:

[0135] During the environmental scanning phase, the drill remains stationary. A binocular camera captures three sets of stereo images, each separated by 10°. Simultaneously, a laser sensor scans the rock surface contours, acquiring three-dimensional information about the environment.

[0136] Step 3: Feature Recognition:

[0137] During the feature recognition phase, the vision system detects natural feature points in the rock mass and matches them to the corresponding features in the left and right images. The three-dimensional coordinates of the feature points are calculated using the ranging data acquired by the laser sensor.

[0138] Step 4: Hole Positioning:

[0139] During the hole positioning stage, the target hole position is matched according to the design drawing, and the deviation between the current drill bit and the target hole position is calculated, including the horizontal deviation ΔX, vertical deviation ΔY and angular deviation Δθ.

[0140] Step 5: Pose adjustment:

[0141] During the posture adjustment stage, the actions of the drill rig's actuators are controlled, including: adjusting the horizontal position of the drill bit through the hydraulic cylinder, adjusting the angle of the drill rod through the pitch motor, and correcting the orientation of the drill bit through the rotating platform, thereby achieving precise positioning of the drill bit.

[0142] Step 6: Accuracy Verification:

[0143] During the accuracy verification phase, data at the current position is collected again and the residual error is calculated. If the residual error does not meet the requirements, steps 4 and 5 are repeated until the accuracy requirements are met.

[0144] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An automatic hole-finding and positioning system for an anchor drilling rig based on machine vision, comprising an anchor drilling rig body (1) and a control room arranged in a tunnel, characterized in that: A binocular vision camera (2) and a laser displacement sensor (3) are provided on the anchor drilling rig body (1); a drilling rig controller (4), an inclination sensor (5), an industrial switch (7), a drilling rig electric control box (9), and a communication module (10) are provided inside the anchor drilling rig body (1); A positioning target (11) and a wireless AP base station (13) are also provided in the lane; A positioning console (12) is provided in the control room, and a positioning control computer (15), a core switch (16), and a drilling positioning server (17) are also provided in the positioning console (12); The drilling rig controller (4) is connected to the binocular vision camera (2), the laser displacement sensor (3), the tilt sensor (5), the industrial switch (7), the drilling rig electric control box (9), and the communication module (10) through wires; The communication module (10) is wirelessly connected to the wireless AP base station (13) via a wireless network; The wireless AP base station (13) is connected to the core switch (16) via a wire, and the core switch (16) is connected to the positioning control computer (15) and the drilling positioning server (17) via wires; The wireless AP base station (13) is wirelessly connected to the positioning target (11) via a wireless network.

2. The automatic hole-finding and positioning system for anchor drilling rigs based on machine vision according to claim 1, characterized in that: An explosion-proof lighting lamp (6) is also provided on the anchor drilling rig body (1), and a control end of the explosion-proof lighting lamp (6) is connected to an industrial switch (7) via a wire.

3. The automatic hole-finding and positioning system for anchor drilling rigs based on machine vision according to claim 1, characterized in that: The binocular vision camera (2) comprises a left camera and a right camera, and the left camera and the right camera are specifically arranged horizontally on the drilling rig crossbeam of the anchor drilling rig body (1) via a shock-absorbing bracket.

4. The automatic hole-finding and positioning system for anchor drilling rigs based on machine vision according to claim 1, characterized in that: The laser displacement sensor (3) is specifically mounted on a bracket on the side of the drill bit of the anchor drilling rig body (1); The measuring axis of the laser displacement sensor (3) is arranged to form an angle of 30° with the drill rod.

5. The automatic hole-finding and positioning system for anchor drilling rigs based on machine vision according to claim 1, characterized in that: The drilling rig controller (4) is specifically connected to the drilling rig electric control box (9) via a PROFINET bus, an EtherCAT network cable, or a Modbus RTU bus.

6. The method for positioning an anchor drilling rig automatic hole-finding and positioning system based on machine vision according to claim 1, characterized in that: The positioning steps include the following: Step 1: Controlling the drilling rig controller (4) to receive data collected by the laser displacement sensor (3) and the tilt sensor (5), respectively; the drilling rig controller (4) analyzes and processes the collected data and then sends the data to the communication module (10); Step 2: Control the drilling rig controller (4) to communicate with the drilling rig electric control box (9) to obtain the operating parameters of the drilling rig, and send the operating parameters of the drilling rig to the communication module (10); The obtained operating parameters include the drill speed ω and the feed speed v, which are defined as satisfying ω=kv, where k is the material correlation coefficient; Step 3: Control the drilling rig controller (4) to receive the image data collected by the binocular vision camera (2) and perform pre-processing, including image binarization, image enhancement, and stereo correction processing; Step 4: The control communication module (10) performs data communication with the core switch (16) provided in the control room through the wireless AP base station (13); Step 5: Control the positioning target (11) set in the tunnel to send the position parameters to the wireless AP base station (13) in real time via the wireless network; Step 6: Estimate the acquisition distance through the structural parameters of the anchor drilling robot and the transformation relationship between the coordinate axes, and calculate the spatial distance between the camera position and the target when acquiring the image; Step 7: Using the functional relationship between the acquisition distance and the anchor hole radius in the image, determine the radius of the anchor hole at any distance, and then determine the center coordinates and the maximum and minimum radius of the anchor hole; Step 8: Based on the geometric constraints of the maximum and minimum radius of the anchor hole, the anchor holes on the steel strip are detected using Hough transform with adaptive radius parameter adjustment. The linear slope and geometric constraints are used to achieve stereo matching of the anchor hole contours. Step 9: Determine the spatial coordinates of the anchor hole center in the camera coordinate system based on the binocular vision positioning principle; Step 10: The anchor drilling rig body launches and installs the anchor into the corresponding anchor hole according to the determined spatial coordinates of the anchor hole center.

7. The method for positioning an anchor drilling rig automatic hole-finding and positioning system based on machine vision according to claim 6, characterized in that: The specific method of step six is: Based on the conversion relationship of each coordinate system, the conversion relationship between the camera coordinate system and the fuselage coordinate system is obtained, and the three-dimensional coordinates of the camera coordinate system in the world coordinate system are obtained in real time through the fuselage positioning method; Construct the camera coordinate system, reference coordinate system, and world coordinate system respectively: Among them, the spatial coordinates of point A on the camera plane in the camera coordinate system are defined as C A=(x1,y1,z1), and convert it to the world coordinate system W The coordinates in A=(x2,y2,z2) are converted as follows: ； in, Represents the conversion relationship between the camera coordinate system and the body coordinate system, and Indicates the conversion relationship between the fuselage coordinate system and the world coordinate system; Define the coordinates of point B on the target in the world coordinate system as WB=(x3,y(3),z(3)), where x(3) and y(3) are unknowns; The acquisition distance can be approximately expressed as the distance between point A on the camera and point B on the steel belt, expressed as: ； In order to reduce the impact of the difference between the X-axis and the Y-axis, the attenuation coefficient α is introduced, and the calculation formula of the spatial distance is: 。 8. The method for positioning an anchor drilling rig automatic hole-finding and positioning system based on machine vision according to claim 7, characterized in that: The specific method of step seven is: Based on step 6, the target contour is fitted. To compensate for the distance between the camera and the target that affects the center distance between the targets, the following compensation rules are used: ； Among them, x i and y i represents the coordinates of the i-th target in the image.

9. The method for positioning an anchor drilling rig automatic hole-finding and positioning system based on machine vision according to claim 8, characterized in that: The specific method of step eight is: Obtain the target center coordinates through parameter-adaptive Hough circle detection: The circle function is defined as follows: x=x0+r0cosα; y=y0+r0sinα; Where r0 is the target radius, x0 and y0 are the target center coordinates, x and y are the image coordinates, and α represents the angle, which ranges from 0° to 360°. Since the contours on the image are oriented, the coordinates of the circle centers can be arranged from small to large in the horizontal direction and defined as: O l1 (u l1 , v l1 ), O l2 (u l2 , v l2 ), O l3 (u l3 , v l3 ), ...,O lm (u lm , v lm ) is the coordinate of the center point of the anchor drill hole in the image captured by the left camera; O r1 (u r1 , v r1 ), O r2 (u r2 , v r2 ), O r3 (u r3 , v r3 ), ....,O rn (u rn , v rn ) is the coordinate of the center point of the anchor drill hole in the image captured by the right camera; The above coordinates are arranged in ascending order of horizontal pixel coordinate values. Based on the fact that the pixel coordinate values ​​of the same point in the vertical direction of the left and right images are the same, stereo matching is performed on different anchor borehole centers to complete the three-dimensional matching of the target center and determine the spatial coordinates of the target center. The stereo matching accuracy of the target center is then verified by the directionality of the contour: Definition: If the i-th point on the left and the j-th point on the right are a set of matching points, then i±p and j±p on the image are a set of matching points; In addition, the distance between two adjacent matching points is basically the same, with a small fluctuation range, and the expression is: |O li The li+1 | = O rj The rj+1 。 10. The method for positioning an anchor drilling rig automatic hole-finding and positioning system based on machine vision according to claim 9, characterized in that: The specific method of step nine is: The coordinates of the target center are calculated using the ideal model of the binocular vision system: Define X separately W 、Y W , Z W is the world coordinate, X C 、Y C , Z C is the camera coordinate, u and v are the pixel plane coordinates; O l and O r is the optical center of the two cameras, B is the distance between the two cameras, and f is the focal length; C l and C r They are perpendicular to the camera optical axis Z C The left image plane and the right image plane of ; For stereo correction, point A (X, Y, Z) is a point in the world coordinate system, and its corresponding points in the image plane are a1 (u1, v1) and a2 (u2, v2). The conversion formula between the world coordinate system and the image coordinate system is obtained based on similar triangles: ； The above formula is used to calculate the coordinates of any point in space, where f ′ is one of the intrinsic parameters obtained through camera calibration; By identifying and matching circles, the pixel coordinates in the drill hole image, including those in the left and right images, can be determined; According to the calibration results, segmentation and stereo matching results, the spatial coordinates of a point are calculated using the following formula: y=kx+b; Where k is the slope of the line and b is the vertical coordinate of the origin.

Images