Mine water exploration and drainage drill rod automatic counting method and system based on static geometric calibration

Abstract

This invention discloses an automatic counting method and system for mine water exploration and drainage drill rods based on static geometric calibration, belonging to the field of computer vision technology. The method includes: S1: In the static calibration stage, extracting the position information of the drill rod and drilling rig from real-time images of the work site using a deep learning object detection algorithm; S2: Obtaining the drilling rig's operating direction vector based on the position information, obtaining the boundary direction vector of the work area based on the drilling rig's operating direction vector, and obtaining area determination rules based on the boundary direction vector of the work area; S3: Obtaining counting rules based on the work area dwell time mechanism, obtaining valid counting rules based on the counting rules and the area determination rules, and obtaining the number of drill rod operations based on the valid counting rules. This invention solves the technical problem that existing technologies struggle to improve counting accuracy, leading to low operational safety.

Description

Technical Field

[0001] This invention relates to the field of computer vision technology, specifically to an automatic counting method and system for mine water exploration and drainage drill rods based on static geometric calibration. Background Technology

[0002] In drilling operations such as water exploration and geological surveys in coal mines, deviations and errors in drill pipe counting are key factors leading to misjudgments of underground geological and hydrological conditions. Inaccurate counting not only causes misjudgments of borehole depth and geological structure but can also directly trigger safety accidents such as water inrush and dam failures. Therefore, accurately recording the number of drill pipes pushed in and pulled back is crucial for ensuring project safety, assessing geological conditions, and evaluating operational progress. Currently, drill pipe counting is primarily done manually. Workers must visually observe the raising and lowering of the drill pipe in complex environments with pervasive dust, dim lighting, and frequent fluctuations, while manually recording the count. This traditional method has significant technical limitations; it is prone to errors and omissions due to factors such as worker fatigue, lack of concentration, or obstructed vision, thus posing a safety hazard. Therefore, improving the accuracy of counting to enhance operational safety remains a technical challenge that current technology struggles to address. Summary of the Invention

[0003] To address the technical problem that existing technologies struggle to improve the accuracy of counting, leading to low operational safety, this invention provides an automatic counting method and system for mine water exploration and drainage drill rods based on static geometric calibration. By using deep learning target detection algorithms to identify and replace manual observation, geometric rules to replace subjective judgment, and spatiotemporal dual-judgment to filter out interference actions, this invention fundamentally overcomes the technical limitations of manual counting and solves the technical problem that existing technologies struggle to improve the accuracy of counting, resulting in low operational safety.

[0004] To solve the above-mentioned technical problems, the present invention provides an automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration, comprising the following steps: S1: In the static calibration phase, the position information of the drill rod and drilling rig is extracted from real-time images of the work site using a deep learning object detection algorithm; S2: Obtain the drilling rig operation direction vector based on the location information, obtain the operation area boundary direction vector based on the drilling rig operation direction vector, and obtain the area determination rules based on the operation area boundary direction vector. S3: Obtain counting rules based on the dwell time mechanism in the work area, obtain valid counting rules based on the counting rules and the area determination rules, and obtain the number of drill pipe operations based on the valid counting rules.

[0005] Preferably, in step S1: during the static calibration phase, the position information of the drill pipe and drilling rig is extracted from real-time images of the work site using a deep learning object detection algorithm, including: Bilateral filtering is used to filter out interference features in real-time images to obtain clean images, and adaptive histogram equalization is used to perform image enhancement processing on the clean images to obtain optimized images; An initial target detection model is constructed based on a deep learning target detection algorithm. An occlusion sample from the drill pipe and the drill rig is introduced to train the initial target detection model and obtain an optimized target detection model. During the static calibration phase, the position information of the drill rod and drilling rig is extracted from the optimized image by optimizing the target detection model.

[0006] Preferably, in S2, obtaining the drilling rig operating direction vector based on the location information includes: The average detection frame of the drill rod and the average detection frame of the drill rig are obtained based on the position information of the drill rod and the drill rig, respectively. The geometric center point of the drill rig is obtained through the average detection frame of the drill rig, and the Euclidean distance from the corner point of the average detection frame of the drill rod to the geometric center point is calculated. The corner point with the smallest Euclidean distance is taken as the drill pipe head point, and the corner point with the largest Euclidean distance is taken as the drill pipe tail point. The initial drilling rig operation direction vector is obtained based on the coordinates of the drill pipe head point and the drill pipe tail point, and the initial drilling rig operation direction vector is normalized to obtain the drilling rig operation direction vector.

[0007] Preferably, in S2, the step of obtaining the boundary direction vector of the working area based on the drilling rig's working direction vector, and obtaining the area determination rule based on the boundary direction vector of the working area, includes: The validity of the drilling rig's operating direction vector is checked. If the validity check fails, the static calibration stage is restarted after a preset time interval to obtain the area determination rules. Otherwise, the center point of the working area boundary is obtained based on the drill pipe tail point and the drilling rig's operating direction vector. The working area boundary center point is linearly extended in the direction perpendicular to the drilling rig's operating direction vector to obtain the working area boundary direction vector. The working area boundary line corresponding to the working area boundary direction vector is obtained, and the area determination rules are obtained based on the working area boundary line.

[0008] Preferably, the region determination rule includes: When the directed distance from the target feature point on the drill pipe to the push rod line in the straight line of the working area boundary is greater than the preset value, it indicates that the target feature point on the drill pipe is located in the near end area. When the directed distance from the target feature point on the drill pipe to the push rod line in the straight line of the working area boundary is less than the preset value, and the directed distance from the target feature point on the drill pipe to the retraction rod line in the straight line of the working area boundary is greater than the preset value, it indicates that the target feature point on the drill pipe is located in the counting area. When the directed distance from the target feature point on the drill pipe to the rod retraction line in the straight line of the working area boundary is less than the preset value, it indicates that the target feature point on the drill pipe is located in the far end area.

[0009] Preferably, in S3, the counting rule includes: The drill pipe is eligible for counting only when the target feature point on the drill pipe stays in the counting area for a period of time exceeding the preset minimum dwell time threshold.

[0010] Preferably, in S3, obtaining the effective counting rule based on the counting rule and the region determination rule includes: If the drill pipe meets the counting rules, and the target feature point on the drill pipe moves from the counting area to the near end area according to the area determination rules, it is determined as a valid push event. If the drill pipe meets the counting rules, and the target feature point on the drill pipe moves from the counting area to the far end area according to the area determination rules, it is determined as a valid pull event.

[0011] Preferably, in S3, obtaining the number of drill pipe operations based on the effective counting rule includes: After a valid push or pull event is determined by the valid counting rules, the drill pipe enters a valid counting cooldown period. During the valid counting cooldown period, no action of the drill pipe will trigger a valid count. After the valid counting cooldown period ends, the number of drill pipe operations will be obtained again by combining the number of valid push events and / or the number of valid pull events according to the valid counting rules.

[0012] By adopting the above technical solution, the present invention has the following advantages: By preprocessing real-time images through bilateral filtering and adaptive histogram equalization, the image quality problem caused by dust and uneven lighting in mining scenarios is solved, which greatly reduces the recognition difficulty of subsequent target detection models. By introducing occlusion samples to train the initial target detection model, the adaptability of the initial target detection model to complex working conditions is improved, avoiding target omissions and false detections caused by occlusion, and ensuring the stability of location information extraction. The drilling rig's operating direction vector is determined by the position information of the drill rod and the drilling rig. Then, the operating area boundary direction vector is delineated based on this vector, and finally, standardized area determination rules are formed. Through this process, the effective operating actions of the drill rod are transformed into quantifiable geometric spatial parameters, thereby effectively avoiding counting errors caused by ambiguity in manual action judgment. A counting rule is constructed by using a dwell time mechanism within the work area. Combined with area determination rules, this forms a valid counting rule. Only when the drill rod's dwell time within the work area exceeds a preset minimum dwell time threshold and the direction of movement meets the requirements will it be considered a valid operation and counted. Through dual determination logic of space and time, invalid movement interference can be accurately filtered out, ensuring that the counting result corresponds to the actual number of pushes and pulls. By using deep learning target detection algorithms to identify objects and replace manual observation, geometric rules to replace subjective judgment, and spatiotemporal dual-judgment to filter out interfering actions, the technical limitations of manual counting are fundamentally overcome. This solves the technical problem that existing technologies cannot improve the accuracy of counting, thus leading to low operational safety.

[0013] This invention also provides an automatic counting system for mine water exploration and drainage drill rods based on static geometric calibration, applicable to the aforementioned automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration, comprising: The location information acquisition module is used to extract the location information of the drill rod and drilling rig from real-time images of the work site using a deep learning target detection algorithm during the static calibration phase. The area determination rule acquisition module is used to obtain the drilling rig operation direction vector based on the location information, obtain the operation area boundary direction vector based on the drilling rig operation direction vector, and obtain the area determination rule based on the operation area boundary direction vector. The drill pipe operation count acquisition module is used to obtain counting rules based on the dwell time mechanism in the operation area, obtain valid counting rules based on the counting rules and the area determination rules, and obtain the number of drill pipe operations based on the valid counting rules.

[0014] By adopting the above technical solution, the present invention has the following advantages: By using deep learning target detection algorithms to identify objects and replace manual observation, geometric rules to replace subjective judgment, and spatiotemporal dual-judgment to filter out interfering actions, the technical limitations of manual counting are fundamentally overcome. This solves the technical problem that existing technologies cannot improve the accuracy of counting, thus leading to low operational safety.

[0015] The present invention also provides a computer device, including: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the computer device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration. Attached Figure Description

[0016] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings. The drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0017] Figure 1 This is a flowchart illustrating the automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only one preferred embodiment of this invention and are only used to explain this invention. They do not limit the scope of protection of this invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0019] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations (or steps) can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it may also have additional steps not included in the figures; the process may correspond to a method, function, procedure, subroutine, subroutine, etc.

[0020] Example 1: like Figure 1 As shown, the automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration includes: S1: In the static calibration phase, the position information of the drill rod and drilling rig is extracted from real-time images of the work site using a deep learning object detection algorithm.

[0021] In one embodiment, S1: During the static calibration phase, the position information of the drill pipe and drilling rig is extracted from real-time images of the work site using a deep learning object detection algorithm, including: Bilateral filtering is used to filter out interference features in real-time images to obtain clean images, and adaptive histogram equalization is used to perform image enhancement processing on the clean images to obtain optimized images; An initial target detection model is constructed based on a deep learning target detection algorithm. An occlusion sample from the drill pipe and the drill rig is introduced to train the initial target detection model and obtain an optimized target detection model. During the static calibration phase, the position information of the drill rod and drilling rig is extracted from the optimized image by optimizing the target detection model.

[0022] Bilateral filtering is a non-linear image filtering technique whose core function is to smooth images, remove noise, and fully preserve edge information. In this embodiment, the filter kernel in bilateral filtering is larger than the pixel scale of the interfering features (such as weeds, pebbles, and dust spots in a mine) to ensure that the filter kernel can cover and smooth the interference area, and smaller than the minimum pixel size of the target features (such as the diameter of the drill rod or the outline size of key components of the drilling rig) to avoid blurring the edge details of the drill rod / rig due to an excessively large filter kernel. The spatial domain parameters in bilateral filtering directly determine the spatial neighborhood influence range during filtering. The larger the spatial domain parameter value, the more neighboring pixels participate in the weighted calculation, and the stronger the overall image smoothing effect, but at the same time, it increases the computational complexity and reduces the real-time processing speed; the smaller the spatial domain parameter value, the narrower the spatial neighborhood range, the higher the computational efficiency, and the better the preservation effect of local details. The value range parameters in bilateral filtering determine the tolerance for pixel grayscale differences during filtering. The smaller the value of the range parameter, the higher the sensitivity to grayscale differences. Only neighborhood points with grayscale values ​​close to the center pixel are given high weight, resulting in a more prominent edge preservation effect. The larger the value of the range parameter, the higher the tolerance for grayscale differences. Neighborhood points of different grayscale values ​​will participate in the weighting, resulting in a stronger smoothing effect, but it is easy to weaken the edge distinction between the target and the background. In order to finally select the optimal parameter combination that minimizes background interference and makes the target features clearest, candidate parameter combinations of spatial domain parameters and range parameters are first obtained. Several drill rod images and drill rig images containing typical interference (such as drill rod images with dust spots, weed obscuration, and drill rig images with cluttered backgrounds) are selected and input into a bilateral filter obtained based on the candidate parameter combination for subjective visual evaluation and objective index verification. The subjective visual evaluation results and objective index verification results are then weighted and fused to select the optimal parameter combination. The final bilateral filter is obtained through the optimal parameter combination, and the final bilateral filter is used to filter the interference features in the real-time image to obtain a clean image. The subjective visual evaluation includes observing whether the interference is effectively smoothed and whether the target edge is clear, while the objective indicator verification includes calculating the signal-to-noise ratio and edge preservation rate of the images before and after filtering.

[0023] Adaptive histogram equalization (HHEM) is an improved image enhancement technique. Its core function is to address the shortcomings of traditional HHEM in enhancing local contrast. By processing the image in regions, it optimizes the contrast of each local area while avoiding over-enhancement and distortion problems caused by overall HHEM. In drill rod and rig detection scenarios, after bilateral filtering, background interference is smoothly removed, and the core edge features of the drill rod and rig are fully preserved. Applying HHEM further amplifies the grayscale difference between the target (drill rod / rig) and the background. For example, it can accurately brighten previously blurred details in the dark areas of the rig to restore its outline; and optimize the contrast of reflective areas on the drill rod surface in strong light to mitigate overexposure. Through interference feature filtering and image enhancement, the accuracy of the target detection model in extracting drill rod and rig location information from the processed image is improved.

[0024] Specifically, real-time images of the work site are acquired using video capture equipment. During the static calibration phase, the drilling rig in the image needs to move to its outermost position and stop, with a complete drill rod inserted at the tail of the rig. The rig and drill rod remain stationary, and their position information is continuously detected and recorded within the static calibration time (e.g., 50 frames). The deep learning object detection algorithm used is YOLOv11x. YOLOv11x's core does not rely on the object's color, texture, or other easily affected appearance features, and it has strong resistance to changes in lighting, dust, and water mist. Using YOLOv11x to build the object detection model further improves the accuracy of the position information.

[0025] S2: Obtain the drilling rig operation direction vector based on the location information, obtain the operation area boundary direction vector based on the drilling rig operation direction vector, and obtain the area determination rules based on the operation area boundary direction vector.

[0026] In some embodiments, S2, obtaining the drilling rig operation direction vector based on the location information includes: The average detection frame of the drill rod and the average detection frame of the drill rig are obtained based on the position information of the drill rod and the drill rig, respectively. The geometric center point of the drill rig is obtained through the average detection frame of the drill rig, and the Euclidean distance from the corner point of the average detection frame of the drill rod to the geometric center point is calculated. The corner point with the smallest Euclidean distance is taken as the drill pipe head point, and the corner point with the largest Euclidean distance is taken as the drill pipe tail point. The initial drilling rig operation direction vector is obtained based on the coordinates of the drill pipe head point and the drill pipe tail point, and the initial drilling rig operation direction vector is normalized to obtain the drilling rig operation direction vector.

[0027] Drill pipe average detection box , This represents the coordinates of the top left corner of the drill pipe average detection frame. This indicates the coordinates of the lower right corner of the drill pipe average detection frame, and the drilling rig average detection frame. , This represents the coordinates of the top left corner of the drilling rig's average detection frame. This represents the coordinates of the lower right corner of the drilling rig's average detection frame. Geometric center point. Understandably, the corner points of the drill pipe average detection frame are specifically... , , as well as Drilling rig operating direction vector ,in, , This represents the initial drilling rig operating direction vector. Indicates the point at the head of the drill pipe. Indicates the point at the end of the drill pipe. express Euclidean length, This represents the x-component of the drilling rig's operating direction vector on the x-axis. This represents the component of the drilling rig's operating direction vector on the y-axis.

[0028] In some embodiments, S2, the step of obtaining the boundary direction vector of the working area based on the drilling rig's working direction vector and obtaining the area determination rule based on the boundary direction vector of the working area includes: The validity of the drilling rig's operating direction vector is checked. If the validity check fails, the static calibration stage is restarted after a preset time interval to obtain the area determination rules. Otherwise, the center point of the working area boundary is obtained based on the drill pipe tail point and the drilling rig's operating direction vector. The working area boundary center point is linearly extended in the direction perpendicular to the drilling rig's operating direction vector to obtain the working area boundary direction vector. The working area boundary line corresponding to the working area boundary direction vector is obtained, and the area determination rules are obtained based on the working area boundary line.

[0029] Understandably, when When the validity test fails, it indicates that the validity test has failed. Indicates the pixel length of the drill rod. , This is a preset minimum pixel length threshold (e.g., 200 pixels). The center points of the work area boundary specifically include the center point of the putter line and the center point of the return line. Center point of the withdrawal line , and These are the proportionality coefficients, respectively. Work area boundary direction vector .

[0030] Understandably, the general form of the equation of a straight line is: For putter lines and return lines, parameters A, B, and C can be obtained from the coordinates of their two endpoints. Taking a putter line as an example, if the coordinates of the two endpoints of the putter line are... Then the parameter For any point in space Its directed distance to the line .

[0031] Specifically, the region determination rules include: When the directed distance from the target feature point on the drill pipe to the push rod line in the straight line of the working area boundary is greater than the preset value, it indicates that the target feature point on the drill pipe is located in the near end area. When the directed distance from the target feature point on the drill pipe to the push rod line in the straight line of the working area boundary is less than the preset value, and the directed distance from the target feature point on the drill pipe to the retraction rod line in the straight line of the working area boundary is greater than the preset value, it indicates that the target feature point on the drill pipe is located in the counting area. When the directed distance from the target feature point on the drill pipe to the rod retraction line in the straight line of the working area boundary is less than the preset value, it indicates that the target feature point on the drill pipe is located in the far end area.

[0032] Understandably, the proximal zone is the area closest to the drilling rig, the counting zone is the area between the push rod line and the retraction line, and the distal zone is the area furthest from the drilling rig. Preset values ​​can be flexibly set according to user needs.

[0033] S3: Obtain counting rules based on the dwell time mechanism in the work area, obtain valid counting rules based on the counting rules and the area determination rules, and obtain the number of drill pipe operations based on the valid counting rules.

[0034] Specifically, in S3, the counting rules include: The drill pipe is eligible for counting only when the target feature point on the drill pipe stays in the counting area for a period of time exceeding the preset minimum dwell time threshold.

[0035] In this embodiment, after entering the stable counting mode, a target tracking algorithm is employed, specifically ByteTrack, to continuously track the drill rod target in the image and assign a unique tracking ID to each tracked drill rod. The system also continuously determines the target feature point of the tracked drill rod, i.e., the area where the center point of the drill rod's tail is currently located. To prevent misjudgment caused by the rapid movement of the target feature point at the area boundary, a working area dwell time mechanism is introduced. When a tracked drill rod enters the counting area, a timer is started. Only when the drill rod continuously stays in the counting area for a period exceeding a preset minimum dwell time threshold (e.g., 8 seconds) is the drill rod eligible for counting. If it leaves midway, the timer is reset.

[0036] Specifically, in S3, obtaining the effective counting rule based on the counting rule and the region determination rule includes: If the drill pipe meets the counting rules, and the target feature point on the drill pipe moves from the counting area to the near end area according to the area determination rules, it is determined as a valid push event. If the drill pipe meets the counting rules, and the target feature point on the drill pipe moves from the counting area to the far end area according to the area determination rules, it is determined as a valid pull event.

[0037] Understandably, when a drill pipe qualifies for counting, it means that the drill pipe meets the counting rules.

[0038] In one embodiment, step S3, obtaining the number of drill pipe operations based on the effective counting rule, includes: After a valid push or pull event is determined by the valid counting rules, the drill pipe enters a valid counting cooldown period. During the valid counting cooldown period, no action of the drill pipe will trigger a valid count. After the valid counting cooldown period ends, the number of drill pipe operations will be obtained again by combining the number of valid push events and / or the number of valid pull events according to the valid counting rules.

[0039] To prevent the same target from being counted repeatedly due to lingering at the area boundary, a valid count cooldown period is introduced. After any drill pipe completes a count, i.e., after being determined as a valid push or retraction event, it enters a short valid count cooldown period (e.g., 8 seconds). During the valid count cooldown period, any area change behavior of that drill pipe will not trigger a new count. By introducing the valid count cooldown period, combined with the work area dwell time mechanism and area division, noise signals such as shaking, obstruction, and rapid movement are effectively filtered out, enabling accurate differentiation of valid work behaviors and further solving the problems of misjudgment and duplicate counting.

[0040] Example 2: This embodiment also provides an automatic counting system for mine water exploration and drainage drill rods based on static geometric calibration, applicable to the aforementioned automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration, including: The location information acquisition module is used to extract the location information of the drill rod and drilling rig from real-time images of the work site using a deep learning target detection algorithm during the static calibration phase. The area determination rule acquisition module is used to obtain the drilling rig operation direction vector based on the location information, obtain the operation area boundary direction vector based on the drilling rig operation direction vector, and obtain the area determination rule based on the operation area boundary direction vector. The drill pipe operation count acquisition module is used to obtain counting rules based on the dwell time mechanism in the operation area, obtain valid counting rules based on the counting rules and the area determination rules, and obtain the number of drill pipe operations based on the valid counting rules.

[0041] The system also overlays and renders real-time information such as the number of pushes, the number of pulls, and the system status onto the video screen for visualization output, and can report the counting results to the host computer or scheduling system through the data interface.

[0042] Example 3: This embodiment also provides a computer device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration.

[0043] The specific embodiments described above are preferred embodiments of the automatic counting method and system for mine water exploration and drainage drill rods based on static geometric calibration of the present invention. They are not intended to limit the specific scope of the present invention. The scope of the present invention includes, but is not limited to, these specific embodiments. All equivalent changes made in accordance with the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. An automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration, characterized in that, Includes the following steps: S1: In the static calibration phase, the position information of the drill rod and drilling rig is extracted from real-time images of the work site using a deep learning object detection algorithm; S2: Obtain the drilling rig operation direction vector based on the location information, obtain the operation area boundary direction vector based on the drilling rig operation direction vector, and obtain the area determination rules based on the operation area boundary direction vector. S3: Obtain counting rules based on the dwell time mechanism in the work area, obtain valid counting rules based on the counting rules and the area determination rules, and obtain the number of drill pipe operations based on the valid counting rules.

2. The automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to claim 1, characterized in that, S1: In the static calibration phase, the position information of the drill pipe and drilling rig is extracted from real-time images of the work site using a deep learning object detection algorithm, including: Bilateral filtering is used to filter out interference features in real-time images to obtain clean images, and adaptive histogram equalization is used to perform image enhancement processing on the clean images to obtain optimized images; An initial target detection model is constructed based on a deep learning target detection algorithm. An occlusion sample from the drill pipe and the drill rig is introduced to train the initial target detection model and obtain an optimized target detection model. During the static calibration phase, the position information of the drill rod and drilling rig is extracted from the optimized image by optimizing the target detection model.

3. The automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to claim 1, characterized in that, In S2, obtaining the drilling rig's operating direction vector based on the location information includes: The average detection frame of the drill rod and the average detection frame of the drill rig are obtained based on the position information of the drill rod and the drill rig, respectively. The geometric center point of the drill rig is obtained through the average detection frame of the drill rig, and the Euclidean distance from the corner point of the average detection frame of the drill rod to the geometric center point is calculated. The corner point with the smallest Euclidean distance is taken as the drill pipe head point, and the corner point with the largest Euclidean distance is taken as the drill pipe tail point. The initial drilling rig operation direction vector is obtained based on the drill pipe head point and the drill pipe tail point, and the initial drilling rig operation direction vector is normalized to obtain the drilling rig operation direction vector.

4. The automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to claim 3, characterized in that, In S2, the step of obtaining the boundary direction vector of the working area based on the drilling rig's working direction vector, and obtaining the area determination rule based on the boundary direction vector of the working area, includes: The validity of the drilling rig's operating direction vector is checked. If the validity check fails, the static calibration stage is restarted after a preset time interval to obtain the area determination rules. Otherwise, the center point of the working area boundary is obtained based on the drill pipe tail point and the drilling rig's operating direction vector. The working area boundary center point is linearly extended in the direction perpendicular to the drilling rig's operating direction vector to obtain the working area boundary direction vector. The working area boundary line corresponding to the working area boundary direction vector is obtained, and the area determination rules are obtained based on the working area boundary line.

5. The automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to claim 4, characterized in that, The region determination rules include: When the directed distance from the target feature point on the drill pipe to the push rod line in the straight line of the working area boundary is greater than the preset value, it indicates that the target feature point on the drill pipe is located in the near end area. When the directed distance from the target feature point on the drill pipe to the push rod line in the straight line of the working area boundary is less than the preset value, and the directed distance from the target feature point on the drill pipe to the retraction rod line in the straight line of the working area boundary is greater than the preset value, it indicates that the target feature point on the drill pipe is located in the counting area. When the directed distance from the target feature point on the drill pipe to the rod retraction line in the straight line of the working area boundary is less than the preset value, it indicates that the target feature point on the drill pipe is located in the far end area.

6. The automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to claim 5, characterized in that, In S3, the counting rules include: The drill pipe is eligible for counting only when the target feature point on the drill pipe stays in the counting area for a period of time exceeding the preset minimum dwell time threshold.

7. The automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to claim 5, characterized in that, In S3, the step of obtaining valid counting rules based on counting rules and region determination rules includes: If the drill pipe meets the counting rules, and the target feature point on the drill pipe moves from the counting area to the near end area according to the area determination rules, it is determined as a valid push event. If the drill pipe meets the counting rules, and the target feature point on the drill pipe moves from the counting area to the far end area according to the area determination rules, it is determined as a valid pull event.

8. The automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration according to claim 7, characterized in that, In S3, obtaining the number of drill pipe operations based on the effective counting rules includes: After a valid push or pull event is determined by the valid counting rules, the drill pipe enters a valid counting cooldown period. During the valid counting cooldown period, no action of the drill pipe will trigger a valid count. After the valid counting cooldown period ends, the number of drill pipe operations will be obtained again by combining the number of valid push events and / or the number of valid pull events according to the valid counting rules.

9. An automatic counting system for mine water exploration and drainage drill rods based on static geometric calibration, applicable to the automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration as described in any one of claims 1-8, characterized in that, include: The location information acquisition module is used to extract the location information of the drill rod and drilling rig from real-time images of the work site using a deep learning target detection algorithm during the static calibration phase. The area determination rule acquisition module is used to obtain the drilling rig operation direction vector based on the location information, obtain the operation area boundary direction vector based on the drilling rig operation direction vector, and obtain the area determination rule based on the operation area boundary direction vector. The drill pipe operation count acquisition module is used to obtain counting rules based on the dwell time mechanism in the operation area, obtain valid counting rules based on the counting rules and the area determination rules, and obtain the number of drill pipe operations based on the valid counting rules.

10. A computer device, comprising: The computer device includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the automatic counting method for mine water exploration and drainage drill rods based on static geometric calibration as described in any one of claims 1-8.

Images