Rock mass deformation monitoring method and computer program product

Abstract

The invention discloses a rock mass deformation monitoring method and a computer program product, and the method comprises the steps: carrying out the spherical target extraction of point cloud data in different periods, and obtaining the spherical center coordinates and the target radius of each spherical target in different periods; and determining the sphere center displacement offset of each spherical target in different periods according to the sphere center coordinates and the target radius, and calculating the displacement deviation spectrum of each point cloud section line in different periods by adopting a section line point cloud registration algorithm based on the sphere center coordinates of each spherical target in different periods. And three-orthogonal-plane point cloud data is constructed based on the sphere center target of the minimum sphere center displacement offset. And importing the three-orthogonal-plane point cloud data into the point cloud data of different periods, and introducing the selected displacement deviation spectrum as geometric constraint to carry out overall registration on the point cloud data of different periods. The displacement variation of the point cloud is calculated through the directed distances from the point cloud to the three orthogonal planes in different periods after overall registration, the overall deformation of the rock mass is reflected, and the overall deformation trend of the rock mass and the deformation trend of the key area are macroscopically reflected.

Description

Technical Field

[0001] This invention relates to the field of 3D modeling surface description technology for computer-aided drafting, and specifically to a rock mass deformation monitoring method and computer program product. Background Technology

[0002] In recent years, with the development of laser point cloud measurement technology and the maturity of point cloud registration algorithms, these technologies have been widely applied in dangerous rock engineering. Laser point cloud measurement technology can extract the DEM (Digital Image Model) of dangerous rocks from point cloud data obtained by lidar laser point cloud measurement, acquire 3D point cloud data of the dangerous rocks, generate 3D models, and effectively represent the characteristic information of the dangerous rocks. This eliminates the need for personnel to visit the dangerous rock site for surveying, greatly improving work efficiency, reducing manual labor, and increasing automation.

[0003] Point cloud registration algorithms can match 3D point cloud images of unstable rocks from different periods, thereby observing the deformation process of the rock surface at different times. Point cloud registration methods based on laser point cloud measurement technology can accurately reflect the actual situation of unstable rocks. Combined with point cloud registration, rock deformation can be quantitatively calculated, enabling more accurate observation of the deformation and stability of unstable rock structures. Comparative analysis with actual engineering monitoring systems can evaluate the measurement accuracy of laser point cloud measurement technology, meeting the needs of practical engineering. Laser point cloud measurement technology can analyze point deformation, line deformation, and overall deformation of unstable rock structure point cloud data from two periods, avoiding errors caused by single-point measurements, comprehensively evaluating the deformation of unstable rock structures, and improving the stability and sustainability of unstable rock engineering projects.

[0004] Current methods for monitoring rockfall deformation involve setting monitoring points on the stable rock structure and using a 3D laser scanner to perform detailed scanning of these points. The resulting displacement changes at these points are then calculated to analyze the overall trend of the unstable rock structure. This is primarily achieved by registering point cloud data from different periods of the unstable rock structure and then calculating the overall displacement change. Therefore, the most crucial stage in deformation monitoring is data processing, and point cloud registration is the most critical step in this process. The accuracy of point cloud registration directly impacts the accuracy of the rockfall deformation analysis.

[0005] Due to the influence of measurement distance and angle, using a planar target will result in uneven acquisition of target points, leading to a large error in the fitted target center point. Moreover, analysis based solely on target points can only obtain the deformation characteristics of a single point on the unstable rock mass, reflecting only the deformation of discrete points on the stable rock mass structure, and cannot macroscopically reflect the overall or local deformation trend of the stable rock mass structure. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention proposes a rock mass deformation monitoring method and computer program product, capable of macroscopically reflecting the precise overall or local deformation trend of unstable rock masses. The specific technical solution is as follows: In a first aspect, a method for monitoring rock mass deformation is provided, wherein in a first feasible mode of the first aspect, the method includes: Spherical targets were extracted from point cloud data at different times to obtain the center coordinates and radius of each spherical target at different times. Based on the center coordinates and radius of the spherical target at different times, the displacement deviation of the center of each spherical target is calculated, and the displacement deviation spectrum of each point cloud profile is calculated using the profile point cloud registration algorithm. Using the spherical target with the smallest center displacement as the reference target, the point cloud data of the three orthogonal surfaces is simulated and generated based on the center coordinates of the reference target. The point cloud data of the three orthogonal planes were imported into the point cloud data of different periods, and the displacement deviation spectrum of the selected key profile line was used as a constraint to perform overall registration of the point cloud data of different periods. Based on the registered point cloud data, the directed distances from each point cloud to the three orthogonal planes at different times are calculated respectively. Based on the directional distances from the point cloud to the three orthogonal planes at different times, the displacement changes of each point cloud are calculated to determine the overall deformation of the rock mass.

[0007] In conjunction with the first possible implementation of the first aspect, in the second possible implementation of the first aspect, obtaining the coordinates of the center of the spherical target and the target radius includes: A spherical target extraction algorithm is used to extract spherical target point cloud data from the point cloud data; Based on the point cloud data of the spherical target, a fitting algorithm is used to calculate the center coordinates and radius of the spherical target.

[0008] In conjunction with the second possible implementation of the first aspect, in the third possible implementation of the first aspect, the extraction of the spherical target point cloud data includes: Based on the set geometric constraints, a random sampling consensus algorithm is used to extract point cloud data of spherical targets.

[0009] In conjunction with the second feasible method of the first aspect, in the fourth feasible method of the first aspect, the coordinates of the center of the spherical target and the target radius are calculated by fitting using the least squares method.

[0010] In conjunction with the first possible implementation of the first aspect, in the fifth possible implementation of the first aspect, calculating the displacement deviation spectrum of the point cloud profile line includes: Point cloud profiles for different periods are generated by fitting the corresponding sphere center coordinates, and the coordinates of multiple target base points corresponding to the point cloud profiles for different periods are determined by the sphere center coordinates. Using the coordinates of multiple target base points, the singular value decomposition registration algorithm is used to perform coarse registration of point cloud profiles at different times. The iterative nearest point algorithm is used to perform fine registration of the point cloud profiles at different times after coarse registration, and the displacement deviation spectrum of the point cloud profiles is obtained.

[0011] In conjunction with the fifth possible implementation of the first aspect, the sixth possible implementation of the first aspect involves fitting and generating the point cloud profile line, including: Determine the connecting line segment between the two spherical targets based on their respective center coordinates; A cutting cross section is set on the connecting line segment according to the set sliding value; Based on all the set cut cross sections, the point cloud profile lines are generated by fitting using a random sampling consensus algorithm.

[0012] In conjunction with the first feasible method of the first aspect, in the seventh feasible method of the first aspect, the displacement deviation spectrum of the selected key profile line is used as a constraint condition, including: Select the point cloud profiles that control the main structure of the unstable rock, pass through potential sliding surfaces or key joints from all the point cloud profiles as key profiles.

[0013] In conjunction with the first feasible method of the first aspect, the eighth feasible method of the first aspect involves overall registration of point cloud data from different periods, including: Feature points were extracted from point cloud data from different periods, and a singular value decomposition registration algorithm was used to perform coarse registration of point cloud data from different periods based on the extracted feature points. The iterative nearest-point algorithm is used to perform fine registration of point cloud data at different time points after coarse registration. In conjunction with the first possible implementation of the first aspect, in the ninth possible implementation of the first aspect, the extraction of spherical targets from the point cloud data includes: preprocessing the point cloud data using a filtering algorithm.

[0014] In a second aspect, a computer program product is provided, including a computer program / instructions that, when executed by a processor, implement the steps of the rock mass deformation monitoring method as described in any of the first to ninth implementable methods of the first aspect.

[0015] Beneficial effects: By using the rock mass deformation monitoring method and computer program product of the present invention, the scanning points of the spherical target can always be distributed on a hemisphere, which enables the accurate extraction of the target's center point coordinates and target radius from the point cloud data, providing a precise data basis for subsequent accurate overall or local deformation trend analysis of unstable rock structures.

[0016] By using the coordinates of the center of the spherical targets and their radii at different times, the displacement deviation of the center of each spherical target can be accurately calculated. Simultaneously, a point cloud registration algorithm can be used to calculate the displacement deviation spectrum of each point cloud profile, thus accurately reflecting the deformation characteristics of discrete points and continuous lines on the stable rock mass structure. Furthermore, based on the deformation characteristics of discrete points in the rock mass, and using the deformation characteristics of continuous lines as constraints, overall registration of the point cloud data allows for a macroscopic reflection of the precise overall and key deformation trends of the unstable rock mass structure. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention, the accompanying drawings used in the specific embodiments will be briefly described below. In all the drawings, the elements or parts are not necessarily drawn to scale.

[0018] Figure 1 A flowchart of a rock mass deformation monitoring method provided in an embodiment of the present invention; Figure 2 This is a point cloud profile of the rock mass at a certain time. Figure 3 Linear deformation analysis yields the displacement deviation spectrum of the profile line; Figure 4 This is a spectrum of overall rock mass deformation obtained using the rock mass deformation monitoring method provided in an embodiment of the present invention. Detailed Implementation

[0019] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. These embodiments are merely illustrative of the technical solution of the present invention and are therefore intended to limit the scope of protection of the present invention.

[0020] Example 1 like Figure 1 The flowchart shown illustrates a rock mass deformation monitoring method, which includes: Step 1: Extract spherical targets from point cloud data at different times to obtain the center coordinates and radius of each spherical target at different times; Step 2: Based on the center coordinates and radius of the spherical target at different times, calculate the displacement deviation of the center of each spherical target, and use the profile point cloud registration algorithm to calculate the displacement deviation spectrum of each point cloud profile. Step 3: Using the spherical target with the smallest center displacement as the reference target, simulate and generate point cloud data of three orthogonal surfaces based on the center coordinates of the reference target. Step 4: Import the three orthogonal plane point cloud data into the point cloud data of different periods, and use the displacement deviation spectrum of the selected key profile line as a constraint to perform overall registration of the point cloud data of different periods. Step 5: Based on the registered point cloud data, calculate the directed distance from each point cloud to the three orthogonal planes at different times; Step 6: Based on the directional distances from the point cloud to the three orthogonal planes at different times, calculate the displacement changes of each point cloud to determine the overall deformation of the rock mass.

[0021] Specifically, firstly, a laser scanning radar can be used periodically to scan a rock mass with multiple spherical targets, acquiring multiple frames of point cloud data containing the spherical targets. Each frame of point cloud data corresponds to a different period of the rock mass. By extracting the spherical targets from each frame of point cloud data, the center coordinates and radius of each spherical target at different periods can be obtained.

[0022] Then, by comparing the center coordinates and radius of each spherical target at two different times, the displacement offset of the center of each spherical target between the two times can be obtained, thus accurately reflecting the deformation characteristics of discrete points on the stable rock mass structure. Since single-point deformation characteristics can only reflect the deformation characteristics of discrete points on the stable rock mass structure and cannot macroscopically reflect the overall or local deformation trend of the stable slope structure, the displacement deviation spectrum of each point cloud profile line between the two times can also be calculated based on the center coordinates of each spherical target at two different times using a profile line point cloud registration algorithm.

[0023] The displacement deviation spectrum of each point cloud profile can macroscopically reflect the overall or local deformation trend of the stable rock mass structure. However, the lack of constraints on key structural deformation characteristics can lead to inaccurate overall analysis results or insensitivity to deformation in key risk areas. Therefore, a triorthogonal surface point cloud can be constructed based on the coordinates of the center of a target with the minimum center displacement offset. Then, the triorthogonal surface point cloud data is imported into point cloud data from two different periods, and the displacement deviation spectrum obtained from line deformation analysis is used as a key geometric constraint for overall registration of the point cloud data from the two periods.

[0024] Finally, by calculating the directional distances from the point clouds to the three orthogonal planes at two different times after overall registration, the displacement changes of the point clouds between the two different times can be calculated to reflect the overall deformation of the rock mass. This achieves a comprehensive analysis integrating points, lines, and surfaces, thereby macroscopically reflecting the precise overall and key deformation trends of the unstable rock mass structure.

[0025] In this embodiment, optionally, obtaining the coordinates of the center of the spherical target and the target radius includes: A spherical target extraction algorithm is used to extract spherical target point cloud data from the point cloud data; Based on the point cloud data of the spherical target, a fitting algorithm is used to calculate the center coordinates and radius of the spherical target.

[0026] Specifically, firstly, existing spherical target extraction algorithms, such as the random sampling consensus algorithm or the region-growing-based spherical extraction algorithm, can be used to extract the spherical target point cloud data from the point cloud data. Then, existing fitting algorithms, such as the Levenberg-Marquardt algorithm or the Hough transform method, can be used to calculate the coordinates of the spherical target's center and the target radius.

[0027] In this embodiment, optionally, extracting the spherical target point cloud data includes: Based on the set geometric constraints, a random sampling consensus algorithm is used to extract point cloud data of spherical targets.

[0028] Specifically, to reduce workload, point cloud data for spherical targets can be extracted from point cloud data using a random sampling consensus algorithm based on predefined geometric constraints, such as a threshold for the radius difference of the spherical target and a threshold for the ratio of spherical points. The radius threshold can be set according to the actual radius of the spherical target deployed on the rock mass. For example, if the actual radius of the spherical target is 14 cm, the radius difference threshold can be set to 1 cm. The ratio of spherical points can be set to 0.7 or 0.8 based on the actual point cloud situation of the spherical target. In the presence of noise interference, the incompleteness of the spherical target can be set to 0.5 or 0.6 to increase the accuracy of the fitting.

[0029] In this embodiment, optionally, the coordinates of the center of the spherical target and the target radius are calculated by fitting using the least squares method.

[0030] Specifically, the sphere fitting problem can be constructed as a nonlinear least squares optimization problem, with the objective function being to minimize the sum of the squares of the distances from all points to the fitted sphere. An iterative optimization algorithm is then used to solve for the optimal coordinates of the sphere's center and its radius. This method offers high fitting accuracy and allows for the addition of constraints. Therefore, the least squares method is preferred for calculating the coordinates of the spherical target's center and its radius. Specifically, let the coordinates of the sphere's center be... The target radius is Any point cloud in the spherical target point cloud data is Therefore, the observation model for the spherical target can be set as follows: ; A multiple linear regression model is constructed using the observation model for parameter estimation, specifically as follows: ; in, , , ; The following can be obtained by calculating the multiple linear regression model using the least squares method: ; Given the coordinates of the sphere's center and the radius of the target, It is the identity matrix. This represents the number of points in the point cloud data.

[0031] In this embodiment, optionally, calculating the displacement deviation spectrum of the point cloud profile line includes: Point cloud profiles for different periods are generated by fitting the corresponding sphere center coordinates, and the coordinates of multiple target base points corresponding to the point cloud profiles for different periods are determined by the sphere center coordinates. Using the coordinates of multiple target base points, the singular value decomposition registration algorithm is used to perform coarse registration of point cloud profiles at different times. The iterative nearest point algorithm is used to perform fine registration of the point cloud profiles at different times after coarse registration, and the displacement deviation spectrum of the point cloud profiles is obtained.

[0032] Specifically, the deformation characteristics of a single point cloud can only reflect the deformation characteristics of discrete points on a stable rock mass structure, not the deformation characteristics of continuous lines. Therefore, it is essential to perform linear deformation analysis on the deviations of point clouds in stable rock mass structures. For existing point cloud registration methods, registration in unstructured scenes mainly relies on non-target registration, which is computationally intensive and lacks accuracy. Therefore, a combination of coarse and fine registration can be used to calculate the displacement deviation spectrum of the point cloud profile lines.

[0033] Specifically, firstly, the centers of each spherical target in the point cloud data from different periods are connected pairwise to form corresponding line segments. Point cloud profiles are then generated based on these line segments, thus obtaining the point cloud profiles for different periods. Next, the midpoint coordinates of the connecting line segment are determined based on the centers of the two spherical targets. Then, using the centers of the two spherical targets and the midpoint of the connecting line segment as target reference points, a singular value decomposition registration algorithm is used to perform coarse point cloud registration between the corresponding point cloud profiles from the two periods, based on the coordinates of these three target reference points. The coordinates of the three target reference points corresponding to the point cloud profile at two different times are respectively expressed as follows: and Based on the target reference point corresponding to the point cloud profile line, the centroid coordinates for different periods are calculated. The specific calculation formula is as follows: ; ; The point cloud data corresponding to two different periods are processed in a decentralized manner to obtain decentralized point sets for each period. The specific calculation formula is as follows: ; ; The rotation matrix is ​​calculated based on the decentralized point sets corresponding to two different periods, as shown in the following formula: ; ; in, Let covariance matrix be the variance matrix. For rotation matrix, It is a left singular vector matrix. It is a singular value matrix. It is a right singular value vector matrix.

[0034] Then, based on the calculated rotation matrix and the centroid coordinates, the corresponding translation vector between the two different periods is calculated. The specific calculation formula is as follows: .

[0035] Finally, the point cloud data from two different periods are coarsely registered using the calculated rotation matrix and translation vector to align the coordinates of the point cloud data from the two different periods.

[0036] Based on coarse registration, the iterative nearest-point algorithm can be used to perform fine registration of point cloud profiles from two different periods, obtaining the displacement deviation spectrum of the point cloud profiles. The iterative nearest-point algorithm can handle local deformation and non-rigid changes between point clouds, specifically including: First, using the rotation matrix and translation vector obtained from coarse registration as initial values, the registered point cloud data is determined based on the translation vector. The specific calculation formula is as follows: ; Then, the average Euclidean distance between the point cloud in the registered point cloud data and the same point cloud in the point cloud data from another period is calculated. The specific calculation formula is as follows: ; in, For the registered point cloud coordinates Point clouds for another period The coordinates.

[0037] Next, it is determined whether the calculated average Euclidean distance is less than a set threshold or whether the number of iterations has reached a set standard. If so, the iteration ends.

[0038] If this is not achieved, the singular value decomposition (SVD) registration algorithm is used to search for the rotation matrix and translation vector that minimizes the registration error function of the corresponding point cloud profiles between the two periods. Based on the searched rotation matrix and translation vector, the mean Euclidean distance between the two registered point cloud datasets is calculated until the mean Euclidean distance is less than a set threshold or the number of iterations reaches a set standard. The registration error function is: ; in, This is the number of reliable point pairs used to calculate the current optimal transformation after outlier removal.

[0039] As the number of iterations increases, the distance between points in two frames of point cloud data from different periods gradually decreases, and the registration accuracy continuously improves. Through coarse and fine registration, the deformation and displacement deviation spectrum of the point cloud profile lines from the two periods of point cloud data can be accurately analyzed. Utilizing the stability of the spherical target further enhances registration accuracy, thereby effectively monitoring the linear deformation of the rock mass.

[0040] In this embodiment, optionally, the point cloud profile line is generated by fitting, including: Determine the connecting line segment between the two spherical targets based on their respective center coordinates; A cutting cross section is set on the connecting line segment according to the set sliding value; Based on all the set cut cross sections, the point cloud profile lines are generated by fitting using a random sampling consensus algorithm.

[0041] Specifically, when fitting and generating point cloud profile lines, a cutting cross-section can be set on the line segment connecting the centers of two spherical targets according to a set width. Then, based on all the set cutting cross-sections, the point cloud profile line is generated using a random sampling consensus algorithm. Specifically, first, representative points corresponding to the point cloud of each cutting cross-section are determined, and then, based on the representative points corresponding to all cutting cross-sections, a three-dimensional space curve is fitted using the RANSAC algorithm. Finally, after resampling, the final point cloud profile line is generated, as shown in the image. Figure 2 As shown.

[0042] In this embodiment, optionally, the displacement deviation spectrum of the selected key profile line is used as a constraint condition, including: Select the point cloud profiles that control the main structure of the unstable rock, pass through potential sliding surfaces or key joints from all the point cloud profiles as key profiles.

[0043] Specifically, since single-point deformation features can only reflect the deformation of discrete points on the stable structure of dangerous rocks, while the displacement deviation spectrum of point cloud profiles can macroscopically reflect the overall or local deformation trend of the stable slope structure, it lacks constraints on the deformation features of key structures, which may lead to inaccurate overall analysis results or insensitivity to deformation in key risk areas.

[0044] Therefore, after calculating the displacement deviation spectrum of the point cloud profiles between two different periods, multiple point cloud profiles that can control the main structure of the unstable rock mass, cross potential sliding surfaces, or key joints can be selected as key profiles based on the placement of the spherical target and the geological structure of the rock mass. The displacement deviation spectrum of the selected key profiles is then used as a constraint condition in the overall registration analysis of the point cloud data from two different periods, thereby achieving a comprehensive analysis integrating points, lines, and surfaces. The overall registration of the point cloud data includes: First, based on the calculated displacement of the center of each spherical target, the spherical target with the smallest displacement is selected as the reference target. Then, using the center coordinates of the reference target as the origin, a triorthogonal plane reference coordinate system containing the center coordinates is constructed. With the center coordinates as the corner points and the target radius as the side length, simulated triorthogonal plane point cloud data is generated using MATLAB.

[0045] Then, the simulated triorthogonal surface point cloud data is imported into point cloud data from two different periods. Using the displacement deviation spectrum of the selected key profile line as a constraint, the point cloud data from the two periods are registered holistically. Specifically, when searching for the rotation matrix and translation vector that minimizes the registration error function, it is required that the positions of points on the key profile line, after the overall transformation, must be consistent with the high-precision displacement results independently calculated in the line deformation analysis. This ensures that the deformation amount after overall registration of the key profile line is highly consistent with the results calculated by the line deformation analysis, guaranteeing the accuracy of the overall registration analysis results and its sensitivity to key areas.

[0046] After the point cloud data from two different periods are registered as a whole, the directed distances from the point clouds in the two periods to the imported triorthogonal planes can be calculated separately. Finally, by comparing the directed distances of the same point cloud in different periods, the displacement change of the point cloud can be determined, and the overall displacement deviation spectrum and volumetric deformation of the rock mass can be obtained.

[0047] In this embodiment, optionally, overall registration of point cloud data from different periods includes: Feature points were extracted from point cloud data from different periods, and a singular value decomposition registration algorithm was used to perform coarse registration of point cloud data from different periods based on the extracted feature points. The iterative nearest point algorithm is used to perform fine registration of point cloud data at different time periods after coarse registration.

[0048] Specifically, the overall registration method for point cloud data from different periods is the same as the registration method for point cloud profiles from different periods. First, points with significant features are extracted from the point cloud data from different periods. These points may have different feature descriptors in different point cloud data. Then, based on the extracted feature points, the initial transformation matrix is ​​calculated using the same method as for the coarse registration of point cloud profiles from different periods, transforming the point cloud data from different periods from their respective local coordinate systems to the global coordinate system.

[0049] Finally, using the same method as the fine registration of point cloud profiles from different periods, the optimal registration transformation matrix is ​​calculated by optimizing the objective function to minimize the distance difference between feature points.

[0050] Example 2 Example 2 is largely the same as Example 1, except that: spherical target extraction of the point cloud data includes: preprocessing the point cloud data using a Gaussian filtering algorithm.

[0051] A computer program product includes a computer program / instructions, which, when executed by a processor, implement the steps of the above-described rock mass deformation monitoring method.

[0052] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions 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, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for monitoring rock mass deformation, characterized in that, include: Spherical targets were extracted from point cloud data at different times to obtain the center coordinates and radius of each spherical target at different times. Based on the center coordinates and radius of the spherical target at different times, the displacement deviation of the center of each spherical target is calculated, and the displacement deviation spectrum of each point cloud profile is calculated using the profile point cloud registration algorithm. Using the spherical target with the smallest center displacement as the reference target, the point cloud data of the three orthogonal surfaces is simulated and generated based on the center coordinates of the reference target. The point cloud data of the three orthogonal planes were imported into the point cloud data of different periods, and the displacement deviation spectrum of the selected key profile line was used as a constraint to perform overall registration of the point cloud data of different periods. Based on the registered point cloud data, the directed distances from each point cloud to the three orthogonal planes at different times are calculated respectively. Based on the directional distances from the point cloud to the three orthogonal planes at different times, the displacement changes of each point cloud are calculated to determine the overall deformation of the rock mass.

2. The rock mass deformation monitoring method according to claim 1, characterized in that, Obtaining the coordinates of the center of the spherical target and the target radius includes: A spherical target extraction algorithm is used to extract spherical target point cloud data from the point cloud data; Based on the point cloud data of the spherical target, a fitting algorithm is used to calculate the center coordinates and radius of the spherical target.

3. The rock mass deformation monitoring method according to claim 2, characterized in that, Extracting the point cloud data of the spherical target includes: Based on the set geometric constraints, a random sampling consensus algorithm is used to extract point cloud data of spherical targets.

4. The rock mass deformation monitoring method according to claim 2, characterized in that, The coordinates of the center of the spherical target and the target radius are calculated by fitting using the least squares method.

5. The rock mass deformation monitoring method according to claim 1, characterized in that, Calculating the displacement deviation spectrum of the point cloud profile includes: Point cloud profiles for different periods are generated by fitting the corresponding sphere center coordinates, and the coordinates of multiple target base points corresponding to the point cloud profiles for different periods are determined by the sphere center coordinates. Using the coordinates of multiple target base points, the singular value decomposition registration algorithm is used to perform coarse registration of point cloud profiles at different times. The iterative nearest point algorithm is used to perform fine registration of the point cloud profiles at different times after coarse registration, and the displacement deviation spectrum of the point cloud profiles is obtained.

6. The rock mass deformation monitoring method according to claim 5, characterized in that, The process of fitting and generating the point cloud profile includes: Determine the connecting line segment between the two spherical targets based on their respective center coordinates; A cutting cross section is set on the connecting line segment according to the set sliding value; Based on all the set cut cross sections, the point cloud profile lines are generated by fitting using a random sampling consensus algorithm.

7. The rock mass deformation monitoring method according to claim 1, characterized in that, The displacement deviation spectrum of the selected key profile line is used as a constraint condition, including: Select the point cloud profiles that control the main structure of the unstable rock, pass through potential sliding surfaces or key joints from all the point cloud profiles as key profiles.

8. The rock mass deformation monitoring method according to claim 1, characterized in that, Overall registration of point cloud data from different periods, including: Feature points were extracted from point cloud data from different periods, and a singular value decomposition registration algorithm was used to perform coarse registration of point cloud data from different periods based on the extracted feature points. The iterative nearest point algorithm is used to perform fine registration of point cloud data at different time periods after coarse registration.

9. The rock mass deformation monitoring method according to claim 1, characterized in that, Extracting spherical targets from the point cloud data includes: preprocessing the point cloud data using a filtering algorithm.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, the steps of the rock mass deformation monitoring method as described in any one of claims 1-9 are implemented.

Images