MÉTODO E SISTEMA AUTOMATIZADOS PARA VARREDURA, DETECÇÃO E REVISÃO DE PERFURAÇÕES EM UMA FACE ROCHOSA

Abstract

The present invention relates to an automated method and system for detecting perforations in a rock face that detects perforations in a perforated face in order to position instruments relative to each hole. The method uses a record of light detection and ranging (LiDAR) sensor scans as information, as well as a time-of-flight (ToF) camera. The method requires complementing and comparing a set of perforation candidates obtained from scans with the information from a perforation plan executed on the rock face in order to define a set of perforations.

Description

1 / 18 Automated method and system for scanning, detecting, and reviewing boreholes in a rock face. FIELD OF THE INVENTION

[0001] The present invention relates to the mining and construction industry. In particular, the present invention relates to a system and method for scanning, which detects and examines blast holes and boreholes in a rock face. STATE OF THE ART

[0002] Currently in the mining and construction industry, new alternatives are being considered in order to automate the processes associated with the detection of holes drilled in a rock face, in order to facilitate the tasks associated with the autonomous or semi-autonomous loading of explosives and in the various state-of-the-art solutions that have been found, which partially solve the technical problem faced.

[0003] Within the state of the art is US patent 10,724,371 B2, which discloses a method consisting of obtaining, with a robot camera system, an image representation of a mining wall having drill holes. A drill hole map characterizing the mining wall is compared with the image representation to perform the identification of drill holes from the map not detected in the image representation. The operation is activated in a robot system based on an input from an operator in response to the identification of the drill hole from the map not detected in the image representation, where the image data from the image representation are designated as representative. Petition 870250092138, dated 08 / 10 / 2025, page 9 / 26 2 / 18 of the borehole. However, the cited document refers only to the distance detection of the boreholes and does not consider the examination of the detected boreholes, considering that these blast holes may actually collapse internally due to rock mechanics reasons or may have several deviations in their direction, which makes calculating the amount of explosive to be loaded more complex.

[0004] Another prior art document corresponds to publication WO2020028951 A1, which describes a blast hole measuring and recording apparatus comprising a housing configured to operationally house a solid-state LiDAR sensor array configured to transmit and direct light pulses to a blast hole by shifting a phase of said pulses through the array to compile volumetric data from the field of view of said sensor. Furthermore, a processor is comprised, which is configured to receive the volumetric data from the LiDAR sensor, and said volumetric data indicates an internal volume of said blast hole useful in calculating an explosive charge according to a blast plan, the processor being configured to store and / or transmit the volumetric data.However, the cited document involves the use of satellite data to identify boreholes, which is useful for open-pit mining but not for underground mining, as highlighted in this description. Additionally, volumetric data is obtained from the exterior using images, which can lead to errors given the reliefs that boreholes may have. Petition 870250092138, dated 08 / 10 / 2025, p. 10 / 26 3 / 18

[0005] Another prior art document corresponds to U.S. Patent US 11,055,546 B2, which describes an apparatus for signal detection comprising a point cloud analysis module, an image analysis module, a trunk comparison module, and a signal detector. The point cloud analysis module is configured to receive point cloud data associated with a geographic region and classify at least one point neighborhood in the point cloud data as planar and a signal position candidate. The image analysis module is configured to receive image data associated with the geographic region and calculate a sighting trunk from the image data, and said trunk comparison module is configured to perform a comparison of the sighting trunk to the signal position candidate, which has at least one point neighborhood classified as planar.The signal detector is configured to provide a location for signal detection in response to comparing the sighting trunk to the signal position candidate. However, the cited document is directed towards the detection of flat or 2D figures in a 3D environment, where the sensors move, calculating the geometry of the trunk that these 2D figures form in the light of the geographic location. BRIEF DESCRIPTION OF THE INVENTION

[0006] The present invention relates to an automated method for detecting blast holes, which allows the detection of blast holes in a drilled face by positioning instruments in front of each blast hole. The method uses as information a recording of scans from range sensors and Petition 870250092138, dated 08 / 10 / 2025, p. 11 / 26 4 / 18 Light detection (LiDAR) - preferably 2D LiDAR sensors, which measure distance using a laser mounted on a motor and additionally employing a time-of-flight (ToF) camera for phase detection. The method requires supplementing the information by means of a spatial description of positions (x, y, z) of blast holes or drill holes that involves a drilling plan executed on the rock face.

[0007] The motor-mounted LiDAR sensor allows the scan point density to be adjusted by regulating the rotational speed of the motor that spins the LiDAR sensor. Thus, a higher rotational speed can be used to perform a coarse scan, while lower rotational speeds are preferred if higher resolution and higher point density are required. DESCRIPTION OF THE FIGURES

[0008] Figure 1 shows a schematic representation of image formation of a rock face with drill holes using a visible spectrum camera.

[0009] Figure 2 shows a schematic representation of the coarse scan of a rock face with boreholes using LiDAR sensors.

[00010] Figure 3 shows a schematic representation of the fine-grained scan of a rock face with drill holes using LiDAR sensors.

[00011] Figure 4 shows a schematic representation of the ultrafine scan of a borehole in a rock face using a ToF camera.

[00012] Figure 5 shows a schematic representation of the depth and geometry recording. Petition 870250092138, dated 08 / 10 / 2025, page 12 / 26 5 / 18 inside a borehole in a rock face, inserting a probe with a sensor. DETAILED DESCRIPTION OF THE INVENTION

[00013] The main objective of the borehole assessment and detection system is to find blast holes within the horizontal work face, associate them with the corresponding borehole according to the borehole diagram, and evaluate the deviation of the blast hole from the design requirements. To do this, the system relies on detections based on a three-dimensional reconstruction of the face using a 2D laser scanner and a sensor that enters the boreholes to determine their internal geometry. The following steps are taken to conduct the method.

[00014] Using a LiDAR sensor array, the rock face is scanned, performing a coarse scan and a fine scan, thus obtaining a point cloud of the rock face.

[00015] The point cloud information is analyzed and processed in order to define the set of candidates to be drilled, where a contiguity analysis of the points in the fine cloud of scanned points is performed, then the sets of points with the greatest separation from contiguous points are grouped, and a representative central point is defined for each grouping of sets of points with the greatest separation, in order to define each representative central point as a candidate to be drilled. The set of candidates to be drilled is compared with the drilling plan executed on the rock face in order to define the set of drill holes and their orientation. Petition 870250092138, dated 08 / 10 / 2025, page 13 / 26 6 / 18

[00016] Thus, by comparing the candidates to be drilled detected by fine scanning with the drilling diagram or drilling plan executed on the rock face, these candidates to be drilled are confirmed as drill holes on the rock face. The drilling plan executed on the rock face allows distinguishing the approximate relative position between the drill holes; however, before the scanning process is carried out, its spatial position on the rock face is unknown.

[00017] In order to compare the drill candidates detected by fine scanning with the drill pattern or drill plan executed on the rock face, a fitting process is performed within the executed drill plan. Each drill candidate is evaluated as a drill hole within the executed drill plan, and the number of fits of the other drill candidates is evaluated. Then, the rotation of the normal axis of the executed drill plan is iteratively adjusted to try to improve the fits at this translation point. This is evaluated for all drill candidates, and the translation that obtains the highest score in this evaluation is chosen, i.e., different possible positions and angles are evaluated, and the one that best fits the executed drill plan is chosen.

[00018] An alternative way to approach comparing candidates to be drilled detected by fine scanning with the drilling diagram or drilling plan executed on the rock face is to consider that all points in the point cloud are calculated for Petition 870250092138, dated 08 / 10 / 2025, page 14 / 26 7 / 18 having a certain number of neighbors within a certain neighborhood or spherical volume. Drill holes tend to have fewer neighbors since there is greater separation between points in the cloud.

[00019] It should be noted that rock face points that are farther from the LiDAR sensor have comparatively fewer neighbors within a given neighborhood than rock face points that are closer to the LiDAR sensor. Therefore, in order to perform the analysis, the point neighborhood assessment is adjusted in light of the point distance.

[00020] When there is a drill hole in the drilling plan executed on the rock face that has not been detected and is not included in the candidate set to be drilled, said drill hole is added to the drill hole set in the spatial position that it should occupy geometrically based on its relative distance from other drill holes in the drill hole set.

[00021] Using robotic means, a ToF camera is positioned near a borehole in the borehole array, thus taking recordings of the individual boreholes or blast holes, and this information is used to determine a relative misalignment between the ToF camera and the borehole. The position of the ToF camera is then adjusted by robotic means in order to align the ToF camera with the borehole based on the information of the relative misalignment between the ToF camera and the borehole, thus allowing additional instrumentation to be positioned in front of each borehole. Petition 870250092138, dated 08 / 10 / 2025, page 15 / 26 8 / 18

[00022] A probe with a sensor is inserted into individual boreholes using a transfer medium, which may be, for example, a hose. However, other elements that are substantially flexible transversely but considerably more rigid longitudinally are also suitable for inserting the probe into individual boreholes, such as elongated cylindrical bodies, flexible rods, or retractable antennas. The probe has, for example, an inertial sensor inside, which provides changes in direction and acceleration experienced by the probe. This information is supplemented by the forward distance traveled by the probe to obtain, for example, the path inside the borehole, the internal geometry, the borehole direction, and / or the depth of each borehole or blast hole.

[00023] By integrating the laser scanner with a servomotor, an environment scanning function is performed, which allows for the three-dimensional reconstruction of the work face in order to establish, for example, the relative distance between the walls and the drilled rock face. Appropriately, it is possible to establish an operating volume so that the robotic equipment can move without colliding with the walls of the rock tunnel in which it is working or with other equipment that may be in the same work area.

[00024] Coarse scanning is performed using a laser scanner from the LiDAR sensor array, which is positioned at an initial distance of approximately 6 to 15 meters from the rock face, preferably at a distance of approximately 8 to 12 meters. Coarse scanning Petition 870250092138, dated 08 / 10 / 2025, page 16 / 26 9 / 18 allows for modeling the immediate environment, thus defining the surface comprising the perforated face, locating it spatially, and differentiating it from the walls, ceiling, and floor. This process makes it possible to determine the actual dimensions of the perforated face, avoiding the risk of assuming a predefined dimension and leaving perforation holes out of the analysis. Coarse scanning can be performed even if the LiDAR sensor array consists of a single LiDAR sensor.

[00025] Fine scanning is performed using a LiDAR sensor from the LiDAR sensor array positioned at a second distance of approximately 8 to 12 meters from the rock face, preferably at a distance of approximately 3 to 5 meters. The LiDAR sensor performs a vertical laser scan while a rotating means rotates it horizontally, allowing a three-dimensional image of the entire perforated wall to be obtained. Similarly, the laser scan can be performed horizontally with a rotating means that performs horizontal rotation. In another embodiment of the present technology, the front LiDAR sensor performs a horizontal laser scan while a rotating means rotates it vertically. The rotating means can be, for example, a servomotor that rotates the front LiDAR sensor.Depending on the size of the face and the technical capabilities of the LiDAR sensor itself, this process can be divided into two or more stages by scanning the rock face in subsections. Fine scanning allows for obtaining detailed information about the number of drill holes in the rock face – details that coarse scanning cannot necessarily obtain, and which allows for... Petition 870250092138, dated 08 / 10 / 2025, p. 17 / 26 10 / 18 Calculation of the relative position of each of the drill holes with respect to the laser scanner or the medium containing said laser scanner. Thus, it is possible to correlate the detected drill holes with the drilling diagram, which allows the elimination of drill hole debris or blast holes that failed from previous blasting processes that in principle appear to be a drill hole from the executed drilling plan, but are in reality drill hole debris too small to be a blast hole to be blasted.

[00026] Finally, once the drill holes in the rock face have been confirmed, an approach and centering process is conducted for each of the drill holes using a ToF camera. To perform this approach and centering process, the ToF camera is positioned in front of each of the possible drill holes and scans at a distance of approximately 5 cm and 50 cm from the rock face, more preferably at a distance of approximately 10 cm and 30 cm from the rock face, after the positioning equipment is positioned in front of the wall. The approach and centering process allows the position of a probe to be corrected and adjusted before it is inserted into a drill hole, so that the probe can enter the drill holes easily.

[00027] This technology additionally incorporates an internal exploration system that inserts a probe with sensors into each of the drill holes, for example, allowing the recording of the internal geometry, drilling direction, and / or depth of each drill hole. Petition 870250092138, dated 08 / 10 / 2025, page 18 / 26 11 / 18 An internal map of the drill holes is then created by combining information obtained from coarse and fine scans with information obtained from the exploration of the internal borehole. The internal exploration of the drill holes provides data of significant importance for calculating the amount of explosives, such as the length, path, and volume of each drill hole. The internal volume is estimated by considering the forward distance traveled by the drill within each drill hole and the drill hole diameter used for the executed drilling plan.

[00028] When the probe is inserted into the borehole using a hose, said hose is inserted using a puller that also records the hose's progress. When the hose tip reaches the end of the borehole, an increase in the force exerted by the puller is detected.

[00029] The system and method for scanning, detecting, and examining blast holes and boreholes in a rock face using present technology identifies, compares, and correlates information that comes from a drilling plan executed in the rock face and a set of borehole candidates scanned in a rock face after executing the drilling plan, where the depth of the blast holes is determined by the probe that enters and measures them. Thus, the system and method for scanning, detecting, and examining blast holes and boreholes in a rock face using present technology do not require geolocation and can operate even in underground environments without geolocation. Petition 870250092138, dated 08 / 10 / 2025, p. 19 / 26 12 / 18

[00030] The current technology also allows the detection of potential blast holes using a visible spectrum camera, taking color or black and white images of the rock face, which are then binarized in order to define a set of 2D candidates to be drilled into the rock face in light of the binarized images. The set of 2D candidates to be drilled is compared with the executed drilling plan, eliminating from the set of 2D candidates to be drilled those candidates that – due to their relative distance from other candidates to be drilled – do not correlate with any drill hole within the executed drilling plan.

[00031] The definition of the set of 2D candidates to be drilled into the rock face in light of the binarized images is determined by the following steps: binarization by means of a boundary based on the midpoint of the image histogram; performing morphological closure operations to eliminate remaining scattered points in the image; applying a circularity filter to generate sets of pixels to be analyzed; applying a center contour test to find centers that satisfy the condition of being able to drill holes.

[00032] On the other hand, it has been observed that state-of-the-art solutions significantly reduce their effectiveness in detecting blast holes that are radially further away from cameras and light sources, due to the fact that as blast holes move further radially away, they appear distorted and oval-shaped. To overcome this difficulty, the present method for the Petition 870250092138, dated 08 / 10 / 2025, page 20 / 26 13 / 18 borehole detection incorporates a correction in the evaluation of candidates based on their relative position with respect to the cameras.

[00033] Figure 1 shows a schematic representation of the image capture of a rock face with boreholes 100 using a visible spectrum camera 101. The images are recorded in color or black and white and then binarized in order to detect possible blast holes by identifying darker tones within the color spectrum or grayscale, as appropriate.

[00034] Figure 2 shows a schematic representation of a coarse scan of a rock face with 100 drill holes using a set of 200 LiDAR sensors that scan the rock face at a first distance 401 from the rock face.

[00035] Figure 3 shows a schematic representation of the fine scan of a rock face with 100 drill holes using a 201 front LiDAR sensor positioned at a second distance 402 from the rock face.

[00036] Figure 4 shows a schematic representation of the extra-fine scan of a borehole 100 in a rock face using a ToF camera 102, where the ToF camera 102 is positioned at a third distance 403 from the rock face.

[00037] Figure 5 shows a schematic representation of the recording of the depth and internal geometry of a borehole 100 in a rock face by inserting the sensor-equipped probe 300, where the probe is inserted through a hose 301. Petition 870250092138, dated 08 / 10 / 2025, page 21 / 26 14 / 18

[00038] The exemplary embodiments are described in detail below to illustrate the principles of the invention. The embodiments are provided to illustrate aspects of the invention, but the invention is not limited to any one embodiment. The scope of the invention encompasses various alternatives, modifications, and equivalents, limited only by the embodiments of the claims.

[00039] In a first embodiment, the present technology relates to an automated method for detecting drill holes in a rock face comprising the steps of: scanning the rock face at a first distance with a first scan density using a LiDAR sensor array with at least one LiDAR sensor to obtain a coarse cloud of scan points; defining the dimensions and relative position of the rock face from the information in the coarse cloud of scan points; scanning the rock face at a second scan density using the LiDAR sensor array to obtain a fine cloud of scan points; wherein the second scan density is greater than or equal to the first scan density; defining a set of candidates to be drilled in the rock face based on the information from the fine cloud of scan points;Compare the set of candidate holes to be drilled with a drilling plan executed on the rock face to define a set of drill holes.

[00040] The method further comprises the steps of: moving a ToF camera by robotic means to a borehole in the borehole array; scanning the rock face around the borehole with Petition 870250092138, dated 08 / 10 / 2025, page 22 / 26 15 / 18 the ToF camera to determine a relative misalignment between the ToF camera and the drill hole; adjust the position of the ToF camera by robotic means to align the ToF camera with the drill hole based on the information of the relative misalignment between the ToF camera and the drill hole; and insert into each drill hole a probe with sensors aligned with the ToF camera to record the internal geometry, drill hole direction, and / or depth of each drill hole. An additional technical advantage of inserting a probe is that it allows the determination of the internal condition of the drill hole in real time. When the inserted probe carries an attached IMU and advanced information is obtained from the coil, the information from inside the drill hole can be processed immediately.

[00041] In another embodiment of the present technology, the step of defining the set of candidates to be drilled comprises performing a contiguity analysis of the points in the fine cloud of scanned points, grouping the sets of points with the greatest separation from contiguous points, defining a representative central point for each grouping of sets of points with the greatest separation, and defining each representative central point as a candidate to be drilled.

[00042] In another embodiment of the present technology, the method further comprises removing from the set of candidates to be drilled those candidates to be drilled that - due to their relative distance from other candidates to be drilled, they do not correlate with Petition 870250092138, dated 08 / 10 / 2025, pp. 23 / 26 16 / 18 any drill hole within the executed drilling plan.

[00043] In another embodiment of the present technology, the method further comprises constructing a three-dimensional map of the rock face based on the coarse cloud of scan points, the fine cloud of scan points, and the set of drill holes.

[00044] In another embodiment of the present technology, the front LiDAR sensor performs a vertical laser scan while a rotating means rotates it horizontally, or the front LiDAR sensor performs a horizontal laser scan while a rotating means rotates it vertically.

[00045] In another embodiment, the present technology relates to an automated system for detecting drill holes in a rock face, comprising: a LiDAR sensor array with at least one LiDAR sensor, wherein the LiDAR sensor array scans the rock face at a first distance with a first scan density to obtain a coarse cloud of scan points and wherein the LiDAR sensor array scans the rock face with a second scan density greater than or equal to the first scan density to obtain a fine cloud of scan points; at least one processor configured to define a set of candidates to be drilled in the rock face based on the information from the fine cloud of scan points; wherein the processor compares the set of candidates to be drilled with a drilling plan executed in the rock face to define a set of drill holes. Petition 870250092138, dated 08 / 10 / 2025, pp. 24 / 26 17 / 18

[00046] In another embodiment of the present technology, the system further comprises a ToF camera mounted in the robotic medium and operatively connected to the processor, wherein the ToF camera scans the rock face around a borehole and transmits the information in such a way that the processor determines a relative misalignment between the ToF camera and the borehole.

[00047] In another embodiment of the present technology, the system further comprises a sensor probe aligned with a ToF camera that is inserted into each borehole of the actual borehole assembly to record the internal geometry, borehole direction, and / or depth of each borehole.

[00048] In another embodiment of the present technology, the system further comprises a visible spectrum camera for imaging the rock face, wherein the processor is further configured to binarize the rock face images and define a set of 2D candidates to be drilled into the rock face in the light of the binarized images. REFERENCE NUMBERS

[00049] In order to clarify the present description, a list of the components of the invention and their respective reference numbers in the figures are provided below. 100 Drilling hole 101 Visible Spectrum Camera 102 ToF camera 200 LiDAR sensor set 201 Front LiDAR Sensor Petition 870250092138, dated 08 / 10 / 2025, pp. 25 / 26 18 / 18 300 Probe 301 Hose 401 First distance 402 Second distance 403 Third distance

[00050] Finally, it should be noted that several particular parameters of the invention, such as dimensions, choice of materials, and specific aspects of the preferred configurations described above, may vary or may be modified depending on operational requirements. Consequently, the specific configurations described above are not intended to be limiting, and such variations and / or modifications are within the spirit and scope of the invention. Petition 870250092138, dated 08 / 10 / 2025, page 26 / 26

Claims

1 / 4 CLAIMS 1.An automated method for detecting boreholes in a rock face, characterized in that it comprises the steps of: scanning the rock face at a first distance with a first scan density using a LIDAR sensor array with at least one LIDAR sensor to obtain a coarse scan point cloud; determining the dimensions and relative position of the rock face from the information in the coarse scan point cloud; scanning the rock face with a second scan density using the LIDAR sensor array to obtain a fine scan point cloud, wherein the second scan density is greater than or equal to the first scan density; defining a set of candidate boreholes in the rock face based on the information in the fine scan point cloud; comparing the set of candidate boreholes with a drilling plan executed in the rock face to define a borehole set.

2. Method according to claim 1, characterized in that it further comprises the steps of: robotically approaching a ToF camera to a borehole in the borehole array; scanning the rock face around the borehole with the ToF camera to determine a relative misalignment between the ToF camera and the borehole; adjusting the position of the ToF camera robotically to align the ToF camera with the borehole based on information about the relative misalignment between the ToF camera and the borehole.

3. Method, according to claim 2, characterized in that it further comprises the insertion of a probe with sensors aligned with the ToF camera in each borehole to record the internal geometry, drilling direction and / or depth of each borehole.

4. Method, according to claim 1, characterized in that the step of defining the set of candidate boreholes comprises performing a contiguity analysis of the points in the fine scan point cloud, grouping the sets of points with the greatest separation from adjacent points, defining a representative central point for each grouping of sets of points with the greatest separation, and defining each representative central point as a candidate borehole.

5. Method, according to claim 1, characterized in that it further comprises removing from the set of candidate boreholes those candidate boreholes which, due to their relative distance to other candidate boreholes, do not correlate with any borehole within the executed drilling plan.

6. Method, according to claim 1, characterized in that it further comprises the construction of a three-dimensional map of the rock face based on the coarse scan point cloud, the fine scan point cloud and the set of boreholes. Petition 870250084791, dated 09 / 19 / 2025, p. 25 / 31 3 / 4 7. A method according to claim 1, characterized in that the LIDAR sensor performs a vertical laser scan while a rotating means rotates it horizontally, or the LIDAR sensor performs a horizontal laser scan while a rotating means rotates it vertically.

8. Method, according to claim 1, characterized in that it further comprises the steps of obtaining images of the rock face using a visible spectrum camera; binarization of the rock face images; and definition of a set of 2D borehole candidates on the rock face based on the binarized images.

9. Automated system for detecting boreholes in a rock face, characterized in that it comprises: a LIDAR sensor array with at least one LIDAR sensor, wherein the LIDAR sensor array scans the rock face at a first distance with a first scan density to obtain a coarse scan point cloud, and wherein the LIDAR sensor array scans the rock face with a second scan density greater than or equal to the first scan density to obtain a fine scan point cloud; at least one processor configured to define a set of borehole candidates on the rock face based on the fine scan point cloud information; wherein the processor compares the set of borehole candidates with a drilling plan executed on the rock face to define a set of boreholes.

10. System, according to claim 9, characterized in that it further comprises a ToF camera mounted on robotic means and operatively connected to the processor, wherein the ToF camera scans the rock face around a borehole and transmits the information in such a way that the processor determines a relative misalignment between the ToF camera and the borehole.

11. System according to claim 10, characterized in that it further comprises a probe with sensors aligned with the ToF camera that is inserted into each borehole of the set of actual boreholes to record the internal geometry, drilling direction and / or depth of each borehole.

12. System, according to claim 9, characterized in that it further comprises a visible spectrum camera for obtaining images of the rock face, wherein the processor is further configured to binarize the images of the rock face and define a set of 2D borehole candidates on the rock face based on the binarized images. Petition 870250084791, dated 09 / 19 / 2025, pp. 27 / 31

Images