ICP (Inductively Coupled Plasma) matching method and system based on concave-convex characteristics of mine cave

Abstract

The invention relates to an ICP (Inductively Coupled Plasma) matching method and system based on concave-convex characteristics of a mine cave. The method comprises the following steps: carrying out first screening on point clouds according to a first type curvature and a first type curvature threshold value; carrying out the second screening of the point clouds according to the second-class curvatures of the remaining point cloud points and a second-class curvature threshold value; calculating the main curvature of the remaining point cloud points after two times of screening, solving the curvature characteristics of the point cloud points in the local neighborhood, and screening the characteristic points based on the curvature characteristics of the local neighborhood; acquiring a first type curvature and a main curvature of the feature points in the annular neighborhoods with different radiuses; selecting the maximum value of the absolute values of the first type of curvature and the main curvature in the annular neighborhoods with different radiuses, and taking the radius of the annular neighborhood corresponding to the maximum value as a final feature; in the process of executing the ICP algorithm, the final feature residual error of the source and target point cloud frames is used as a constraint to be added into an ICP target function, and point cloud registration is carried out. The degradation problem of SLAM is effectively relieved through the characteristics of the concave and convex regions, and the positioning and mapping precision in a complex environment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of positioning and mapping, and particularly to an ICP matching method based on the concave and convex characteristics of a mine tunnel. Background Art

[0002] With the continuous development of autonomous driving technology, the Simultaneous Localization and Mapping (SLAM) technology, as one of the key technologies of autonomous driving, has attracted more and more attention. Currently, the mainstream SLAM systems are mainly divided into two types: laser SLAM and visual SLAM. Among them, laser SLAM is widely used in the field of autonomous driving due to its characteristics of being not affected by light, directly obtaining spatial positions, high accuracy, and good real-time performance. However, in the face of environments such as corridors and long straight mine tunnels, due to the lack of constraints in the forward direction, the lidar is prone to degradation, resulting in excessive SLAM mapping errors and affecting subsequent positioning and navigation. Zhang et al. introduced a degradation factor by analyzing the eigenvalues of the problem constraint coefficient matrix, separating the degradation and non-degradation directions, and partially solving the problem in the well-conditioned direction; Turcan, T et al. proposed the X-ICP algorithm, which groups by the projection of the eigenvectors in the three-axis translation and rotation directions, determines whether each direction can be located, and sets corresponding constraints for the non-locatable directions to alleviate degradation. These methods can alleviate the degradation problem to a certain extent, but in environments such as mines, the general direction of degradation is already clear, and there is a lack of research on the registration of similar point cloud features in the degradation direction. Summary of the Invention

[0003] In order to solve the above technical problems or at least partially solve the above technical problems, the present invention provides an ICP matching method and system based on the concave and convex characteristics of a mine tunnel.

[0004] In a first aspect, the present invention provides an ICP matching method based on the concave and convex characteristics of a mine tunnel, including: Calibrating and aligning the lidar and the inertial navigation unit, collecting the point cloud of the mine tunnel with the aligned lidar, and preprocessing the point cloud through filtering and voxelization to construct an octree; Screening out the points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold; Screening out the cylindrical structures, long corridors, and mine tunnel top corners according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold; For the point cloud points remaining after the two screenings, calculating their principal curvatures to obtain the final features of the subsequent feature points for subsequent matching; For the point cloud remaining after the two screenings, obtaining the curvature features within the local neighborhood of the point cloud points, and further screening the feature points based on the local neighborhood curvature features; Set a set number of rings with different radii as the neighborhoods of the filtered feature points, calculate the first type of curvature and the principal curvature of the feature points in the annular neighborhoods with different radii, and select the maximum absolute values of the first type of curvature and the principal curvature in the annular neighborhoods with different radii, as well as the radius of the annular neighborhood corresponding to the maximum value as the final feature of the feature points; During the execution of the ICP algorithm, the final feature residuals corresponding to the feature points in the source and target point cloud frames are added as constraints into the ICP objective function for point cloud registration.

[0005] Furthermore, the process of screening out the points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold includes: Use an octree to search for the local neighborhood of each point in the point cloud, analyze the local neighborhood through principal component analysis to calculate the covariance matrix, calculate the eigenvalues and eigenvectors of the covariance matrix, and take the ratio of the smallest eigenvalue to the sum of all eigenvalues as the first type of curvature of the point; set a first type of curvature threshold to compare with the first type of curvature of the point cloud points, and perform point cloud point screening to retain the point cloud points with the first type of curvature higher than the first type of curvature threshold.

[0006] Furthermore, the process of screening out the cylindrical structures, long corridors and mine roof corners according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold includes: Use an octree to search for the local neighborhood of each point in the point cloud, analyze the local neighborhood through principal component analysis to calculate the covariance matrix, calculate the eigenvalues and eigenvectors of the covariance matrix, take the eigenvector corresponding to the smallest eigenvalue as the normal vector n of the point cloud point, and unify the normal vector direction; construct a surface model and fit the local surface function by the least squares method using the coordinates of the point cloud points in the local neighborhood; Calculate the coefficients of the first fundamental form and the second fundamental form according to the local surface function for partial differential parameter calculation, and calculate the Gaussian curvature as the second type of curvature using the coefficients of the first fundamental form and the second fundamental form: Screen out the point cloud with the second type of curvature lower than the second type of curvature threshold through the preset second type of curvature threshold.

[0007] Furthermore, the local neighborhood of the point cloud point has a quadratic surface model in the form of: ; ; where and are the tangent plane projection coordinates, is the height along the normal vector direction, are the quadratic surface model parameters; The least squares fitting uses the coordinates of the point cloud points in the local neighborhood Solve the parameters of the quadratic surface model. The goal of least squares fitting is to minimize the residuals: ; Calculate the partial differential parameters according to the local surface function: Based on the points of the fitted point cloud of the local neighborhood of the local surface function Calculate the first-order partial derivatives of the local surface: ; where , ; Based on the points of the fitted point cloud of the local neighborhood of the local surface function Calculate the second-order partial derivatives of the local surface: ; Based on the local neighborhood of the point cloud point Calculate the coefficients of the first fundamental form from the first-order partial derivatives of the local surface function: ; Project the second-order partial derivatives of the local surface function of the local neighborhood of the point cloud point onto the normal vector direction to obtain the coefficients of the second fundamental form: ; Calculate the Gaussian curvature of the point cloud point according to the coefficients of the first fundamental form and the coefficients of the second fundamental form: .

[0008] Furthermore, the process of calculating the principal curvatures of the point cloud points includes: Calculate the mean curvature according to the coefficients of the first fundamental form and the coefficients of the second fundamental form of the local surface function of the local neighborhood of the point cloud point : ; Calculate its principal curvatures according to the Gaussian curvature and the mean curvature : .

[0009] Furthermore, for the point cloud remaining after two screenings, obtaining the curvature features within the local neighborhood of the point cloud points, and further screening the feature points based on the local neighborhood curvature features includes: For the point cloud points remaining after two screenings of the local neighborhood , extract the maximum first - type curvature and second - type curvature within the local neighborhood as curvature features: , , Among them, , belongs to the local neighborhood of the point cloud point Any point corresponding to the first - type curvature and second - type curvature, screen the maximum first - type curvature and second - type curvature in the local neighborhood, as the local neighborhood curvature features of the point cloud point , and at the same time delete the points within the local neighborhood range of the point cloud point to ensure that there is only one set of local neighborhood curvature features within the local neighborhood range of the point cloud point .

[0010] Furthermore, add the final feature residual constraint to the ICP objective function: ; Among them, is the IPC point - pair matching item, is the transformation relationship, including rotation and translation transformations, is the feature point i of the source point cloud frame s, is the matching feature point i of the target point cloud frame t; is the final feature residual constraint item, represents the global weight factor, balancing the contributions of the IPC point - pair matching item and the final feature residual constraint item, represents the final feature residual of the feature point with index i between the source point cloud frame s and the target point cloud frame t.

[0011] Furthermore, the final feature residual: ; Among them, represents the final feature residual of the feature point with index i between the source point cloud frame s and the target point cloud frame t, represents the weight corresponding to the l - th feature element in the final feature, The point represents the l - th feature element of the final feature of the feature point with index i in the source point cloud frame s and the target point cloud frame t.

[0012] Furthermore, after introducing the final feature residual constraint item, the final total Jacobian matrix of the transformation relationship T in the ICP algorithm is: ; Among them, , is the Jacobian matrix of the IPC point-to-point matching item, , is the Jacobian matrix of the final feature residual constraint term.

[0013] In a second aspect, the present invention provides an ICP matching system based on the concave and convex characteristics of a mine tunnel, including: at least one processing unit, which is connected to a storage unit and a collection unit through a bus unit. The storage unit stores computer programs and data collected by the collection unit. The processing unit realizes the ICP matching method based on the concave and convex characteristics of the mine tunnel by running the computer programs stored in the storage unit.

[0014] The above technical solutions provided by the embodiments of the present invention have the following advantages compared with the prior art: An ICP matching method based on the concave and convex characteristics of a mine tunnel proposed in this application screens point clouds based on two types of curvatures, and realizes targeted feature extraction for the concave and convex regions of the mine tunnel through a multi-scale circular neighborhood to obtain final features. The final feature residual is introduced into the final ICP objective function. The constraint ability of SLAM in the degradation direction is strengthened through the concave and convex region features, which can effectively alleviate the degradation problem of SLAM and improve the positioning and mapping accuracy in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with the present invention and used together with the specification to explain the principles of the present invention.

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 is a flowchart of an ICP matching method based on the concave and convex characteristics of a mine tunnel provided by an embodiment of the present invention; Figure 2 is a flowchart of screening points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold provided by an embodiment of the present invention; Figure 3 is a flowchart of screening out cylindrical structures, long corridors, and mine tunnel top corners according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold provided by an embodiment of the present invention; Figure 4 is a schematic diagram of an ICP matching system based on the concave and convex characteristics of a mine tunnel provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] It should be noted that in this document, the term "including", "comprising", or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article, or device including a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or device. Without further limitation, an element defined by the phrase "including one..." does not exclude the presence of additional identical elements in the process, method, article, or device including the said element.

[0020] Embodiment 1 Refer to Figure 1 As shown, the ICP matching method based on the concave-convex characteristics of a mine tunnel provided by this application includes: Step 1: Calibrate and align the lidar and the inertial navigation unit to ensure the synchronization of the point cloud data and the attitude information of the IMU, and ensure the accuracy of subsequent registration.

[0021] Collect the point cloud of the mine tunnel with the aligned lidar, and preprocess the point cloud through filtering and voxelization to construct an octree.

[0022] Step 2: Screen out the points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold.

[0023] As Figure 2 shown, in Step 2, the local neighborhood of each point in the point cloud is searched using the octree, and the covariance matrix is analyzed through principal component analysis to calculate the eigenvalues and eigenvectors of the covariance matrix. The ratio of the smallest eigenvalue to the sum of all eigenvalues is used as the first type of curvature of the point. Set a first type of curvature threshold to compare with the first type of curvature of the point cloud points, and perform point cloud point screening to retain the point cloud points with a first type of curvature higher than the first type of curvature threshold. The retained point cloud points are edge or corner features.

[0024] In the specific implementation process, the formula for calculating the covariance matrix of each point using the neighborhood points in the local neighborhood of the point is as follows: ; where is the point cloud point Local neighborhood The centroid of the midpoint cloud, and the centroid coordinates are within the local neighborhood The average value of the coordinates of the midpoints, For the point cloud points The covariance matrix; The formulas for calculating the eigenvalues and eigenvectors of the covariance matrix of the point cloud points are as follows: ; Where, Represents the point cloud point The covariance matrix The eigenvalues and eigenvectors. The eigen-decomposition of the covariance matrix will obtain three eigenvalues and the corresponding eigenvectors.

[0025] The ratio of the smallest eigenvalue to the sum of all eigenvalues is used as the first type of curvature of the point:

[0026] Where, Is the point cloud point The first type of curvature. The relationship of the eigenvalues is: , because the magnitude of the eigenvalues is related to the distribution direction of the points within the local neighborhood. For example, if the points are on a plane, the smallest eigenvalue will be very small, while the larger two eigenvalues are larger, and thus the curvature will be small. On the contrary, if the points are on the edge or corner, the curvature is higher.

[0027] Step 3: Screen out the cylindrical structure, long corridor, and the corner of the mine roof according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold.

[0028] As Figure 3 Shown, the process of Step 3 includes: using an octree to search for the local neighborhood of each point in the point cloud, analyzing the local neighborhood through principal component analysis to calculate the covariance matrix, calculating the eigenvalues and eigenvectors of the covariance matrix, taking the eigenvector corresponding to the smallest eigenvalue as the normal vector n of the point cloud point, and unifying the direction of the normal vector; using the normal vector to perform local surface function fitting on the remaining point cloud, and calculating the second type of curvature using the fitted local surface function.

[0029] In the specific implementation process, use an octree to search for the local neighborhood of each point in the point cloud, analyze the local neighborhood through principal component analysis to calculate the covariance matrix, calculate the eigenvalues and eigenvectors of the covariance matrix, which have been executed in Step 2, directly use the eigenvector corresponding to the smallest eigenvalue in Step 2 as the normal vector n of the point cloud point, and unify the direction of the normal vector. Construct a surface model and fit the local surface function by the least squares method using the coordinates of the point cloud points within the local neighborhood. In this example, a quadratic surface model is adopted, project the local neighborhood point set onto the tangent plane defined by the normal vector n, and fit the local surface structure with the quadratic surface model.

[0030] Point cloud point of the local neighborhood The quadratic surface model has the form: ; where, and are the tangent plane projection coordinates, is the height along the normal vector direction, are the quadratic surface model parameters.

[0031] Least squares fitting uses the coordinates of the point cloud points in the local neighborhood to solve for the quadratic surface model parameters. The goal of least squares fitting is to minimize the residual: ; Perform partial derivative parameter calculations according to the local surface function: Based on the fitted point cloud points in the local neighborhood of the local surface function calculate the first-order partial derivatives of the local surface: ; where, , .

[0032] Based on the fitted point cloud points in the local neighborhood of the local surface function calculate the second-order partial derivatives of the local surface: ; Based on the point cloud points in the local neighborhood of the local surface function, calculate the coefficients of the first fundamental form: ; Project the second-order partial derivatives of the local surface function of the local neighborhood of the point cloud points onto the normal vector direction to obtain the coefficients of the second fundamental form: .

[0033] Calculate the Gaussian curvature of the point cloud points based on the coefficients of the first fundamental form and the coefficients of the second fundamental form: ; The Gaussian curvature is the second type of curvature; Calculate the Gaussian curvature of each remaining point cloud point, and filter out the point cloud with the second curvature lower than the second curvature threshold through a preset second curvature threshold.

[0034] Step Four: For the point cloud points remaining after two screenings, calculate their principal curvatures as the final features of the points for subsequent matching.

[0035] The process of calculating the principal curvature of point cloud points includes: According to the point cloud points of the local neighborhood Calculate the mean curvature based on the coefficients of the first fundamental form and the coefficients of the second fundamental form of the local surface function: ; Calculate its principal curvature based on the Gaussian curvature and the mean curvature : .

[0036] Step Five: For the point cloud remaining after two screenings, obtain the curvature features within the local neighborhood of the point cloud points, and further screen the feature points based on the local neighborhood curvature features.

[0037] For the point cloud points remaining after two screenings of the local neighborhood , extract the maximum first curvature and the second curvature within the local neighborhood as the curvature features: , , where , belongs to the point cloud point of the local neighborhood any point corresponding to the first curvature and the second curvature, screen the maximum first curvature and the second curvature in the local neighborhood as the local neighborhood curvature features of the point cloud point , and at the same time delete the points within the local neighborhood range of the point cloud point to ensure that there is only one set of local neighborhood curvature features within the local neighborhood range of the point cloud point .

[0038] Step Six: Set a set number of rings with different radii as the neighborhoods of the screened feature points, calculate the first curvature and the principal curvature of the feature points in the annular neighborhoods with different radii, and select the maximum absolute values of the first curvature and the principal curvature in the annular neighborhoods with different radii, as well as the radius of the annular neighborhood corresponding to the maximum value as the final features of the feature points: ; where is the maximum value of the first type of curvature of the feature points in each annular neighborhood; is the principal curvature of the feature points in each annular neighborhood of the maximum value; where, is the set of feature points in the annular neighborhood; is the principal curvature of the feature points in each annular neighborhood of the maximum absolute value. is the radius of the annular neighborhood corresponding to the maximum value.

[0039] Step Seven: During the execution of the ICP algorithm, add the final feature residuals corresponding to the feature points in the source and target point cloud frames as constraints into the ICP objective function for point cloud registration.

[0040] Calculate the final feature residuals: ; where, represents the final feature residual of the feature point with index i between the source point cloud frame s and the target point cloud frame t, represents the weight corresponding to the l-th feature element in the final feature, represents the l-th feature element of the final feature of the feature point with index i in the source point cloud frame s and the target point cloud frame t.

[0041] Add the final feature residual constraint to the ICP objective function: ; where, is the IPC point pair matching term, is the transformation relationship, including rotation and translation transformations, is the feature point i of the source point cloud frame s, is the matching feature point i of the target point cloud frame t; is the final feature residual constraint term, represents the global weight factor that balances the contributions of the IPC point pair matching term and the final feature residual constraint term.

[0042] The final total Jacobian matrix of the corresponding transformation relationship T is: ; where, , is the Jacobian matrix of the IPC point pair matching term, , is the Jacobian matrix of the final feature residual constraint term.

[0043] Use the Gauss-Newton method to minimize the objective function and update T iteratively until the objective function converges.

[0044] Embodiment 2 Referring to Figure 4 as shown, an ICP matching system based on the concave and convex characteristics of a mine tunnel provided by an embodiment of the present invention includes: at least one processing unit, which is connected to a storage unit and a collection unit through a bus unit. The collection unit includes a lidar and an inertial navigation unit. The storage unit, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the software programs, computer-executable programs, and modules corresponding to an ICP matching method based on the concave and convex characteristics of a mine tunnel in an embodiment of the present invention, and can be used to store the data collected by the collection unit. The processing unit realizes the above-mentioned ICP matching method based on the concave and convex characteristics of a mine tunnel by running the software programs, computer-executable programs, and modules stored in the storage unit, including: Calibrate and align the lidar and the inertial navigation unit, collect the point cloud of the mine tunnel with the aligned lidar, and preprocess the point cloud by filtering and voxelization and construct an octree; Filter out the points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold; Filter out the cylindrical structures, long corridors, and mine tunnel top corners according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold; For the point cloud points remaining after two screenings, calculate their principal curvatures to obtain the final features of the subsequent feature points for subsequent matching; For the point cloud remaining after two screenings, obtain the curvature features within the local neighborhood of the point cloud points, and further screen the feature points based on the local neighborhood curvature features; Set a set number of rings with different radii as the neighborhoods of the screened feature points, calculate the first type of curvature and the principal curvature of the feature points in the annular neighborhoods with different radii, select the maximum absolute values of the first type of curvature and the principal curvature in the annular neighborhoods with different radii, and the radius of the annular neighborhood corresponding to the maximum value as the final features of the feature points; During the execution of the ICP algorithm, add the final feature residuals corresponding to the feature points in the source and target point cloud frames as constraints into the ICP objective function for point cloud registration.

[0045] Of course, the computer program stored in the storage unit in an ICP matching system based on the concave and convex characteristics of a mine tunnel provided by an embodiment of the present invention is not limited to the method operations as described above, and can also execute the related operations in an ICP matching method based on the concave and convex characteristics of a mine tunnel provided by any embodiment of the present invention.

[0046] Embodiment 3 An embodiment of the present invention provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program, and when the computer program is executed, it realizes the above-mentioned ICP matching method based on the concave and convex characteristics of a mine tunnel, including: Calibrate and align the lidar and the inertial navigation unit, collect the point cloud of the mine tunnel with the aligned lidar, preprocess the point cloud through filtering and voxelization, and construct an octree; Filter out the points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold; Filter out the cylindrical structures, long corridors and the corners of the mine tunnel roof according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold; For the point cloud points remaining after the two screenings, calculate their principal curvatures to obtain the final features of the subsequent feature points for subsequent matching; For the point cloud remaining after the two screenings, obtain the curvature features within the local neighborhood of the point cloud points, and further screen the feature points based on the local neighborhood curvature features; Set a set number of rings with different radii as the neighborhoods of the screened feature points, calculate the first type of curvature and the principal curvature of the feature points in the annular neighborhoods with different radii, select the maximum absolute values of the first type of curvature and the principal curvature in the annular neighborhoods with different radii, and the radius of the annular neighborhood corresponding to the maximum value as the final features of the feature points; During the execution of the ICP algorithm, add the final feature residuals corresponding to the feature points in the source and target point cloud frames as constraints into the ICP objective function for point cloud registration.

[0047] A computer-readable storage medium provided by an embodiment of the present invention stores a computer program that is not limited to the method operations described above, and can also execute related operations in a method for ICP matching based on the concave and convex characteristics of a mine tunnel provided by any embodiment of the present invention.

[0048] In the embodiments provided by the present invention, it should be understood that the disclosed structures and methods can be implemented in other ways. For example, the structural embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of structures or units can be in electrical, mechanical or other forms.

[0049] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0050] In addition, in each embodiment of the present invention, each functional unit may be integrated into a processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware or in the form of software functional units.

[0051] The foregoing is only a specific implementation manner of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather will conform to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. An ICP matching method based on the concave-convex characteristics of mine tunnels, characterized in that Including: Calibrate and align the lidar and inertial navigation unit, collect the point cloud of the mine tunnel with the aligned lidar, preprocess the point cloud through filtering and voxelization, and construct an octree; Filter out the points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold; Filter out the cylindrical structures, long corridors and the corners of the mine tunnel roof according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold; For the point cloud points remaining after two screenings, calculate their principal curvatures to obtain the final features of the subsequent feature points for subsequent matching; For the point cloud remaining after two screenings, obtain the curvature features within the local neighborhood of the point cloud points, and further screen the feature points based on the local neighborhood curvature features; Set a set number of rings with different radii as the neighborhoods of the screened feature points, obtain the first type of curvature and the principal curvature of the feature points in the annular neighborhoods with different radii, select the maximum absolute values of the first type of curvature and the principal curvature in the annular neighborhoods with different radii, and the radius of the annular neighborhood corresponding to the maximum value as the final features of the feature points; During the execution of the ICP algorithm, add the final feature residuals corresponding to the feature points in the source and target point cloud frames as constraints into the ICP objective function for point cloud registration.

2. The ICP matching method based on the concave and convex characteristics of the mine tunnel according to claim 1, characterized in that The process of filtering out the points in the flat area of the point cloud according to the first type of curvature and the set first type of curvature threshold includes: Use the octree to search for the local neighborhood of each point in the point cloud, analyze the local neighborhood through principal component analysis to calculate the covariance matrix, calculate the eigenvalues and eigenvectors of the covariance matrix, and take the ratio of the smallest eigenvalue to the sum of all eigenvalues as the first type of curvature of the point; set a first type of curvature threshold to compare with the first type of curvature of the point cloud points, perform point cloud point screening, and retain the point cloud points with the first type of curvature higher than the first type of curvature threshold.

3. The ICP matching method based on the concave-convex characteristics of a mine tunnel according to claim 1, characterized in that, The process of filtering out the cylindrical structures, long corridors and the corners of the mine tunnel roof according to the second type of curvature of the remaining point cloud points and the set second type of curvature threshold includes: Use the octree to search for the local neighborhood of each point in the point cloud, analyze the local neighborhood through principal component analysis to calculate the covariance matrix, calculate the eigenvalues and eigenvectors of the covariance matrix, take the eigenvector corresponding to the smallest eigenvalue as the normal vector n of the point cloud point, and unify the normal vector directions; construct a surface model and fit the local surface function by the least squares method using the coordinates of the point cloud points in the local neighborhood; Calculate the first fundamental form coefficients and the second fundamental form coefficients according to the local surface function through partial differential parameter calculation, and calculate the Gaussian curvature as the second type of curvature using the first fundamental form coefficients and the second fundamental form coefficients; Filter out the point cloud with the second type of curvature lower than the second type of curvature threshold through the preset second type of curvature threshold.

4. The ICP matching method based on the concave and convex characteristics of the mine tunnel according to claim 3, characterized in that, Point cloud point Local neighborhood of The quadratic surface model of has the form: ； Among them, and are the tangent plane projection coordinates, is the height along the normal vector direction, are the quadratic surface model parameters; Least squares fitting uses the coordinates of the point cloud points in the local neighborhood to solve for the parameters of the quadratic surface model. The goal of least squares fitting is to minimize the residual: ​ ； Perform partial differential parameter calculation according to the local surface function: Point cloud points based on fitting Local neighborhood Local surface function Calculate the first-order partial derivatives of the local surface: ； Among them, , ; Point cloud points based on fitting Local neighborhood Local surface function Calculate the second-order partial derivatives of the local surface: ； Based on the point cloud points of the local neighborhood Calculate the coefficients of the first fundamental form from the first-order partial derivatives of the local surface function: ； Project the second partial derivatives of the local surface function of the local neighborhood of the point cloud points onto the normal vector direction to obtain the second fundamental form coefficients: ； Calculating the Gaussian curvature of a point cloud point based on the coefficients of the first fundamental form and the second fundamental form : 。 5. The ICP matching method based on the concave and convex characteristics of the mine cave according to claim 4, wherein The process of calculating the principal curvature of the point cloud points includes: Based on the point cloud points of the local neighborhood Calculate the mean curvature according to the coefficients of the first fundamental form and the coefficients of the second fundamental form of the local surface function: ； Calculate its principal curvatures based on the Gaussian curvature and the mean curvature : 。 6. The ICP matching method based on the concave-convex characteristics of a mine cave according to claim 1, characterized in that The process of obtaining the curvature features within the local neighborhood of the point cloud points for the point cloud remaining after two screenings, and further screening the feature points based on the local neighborhood curvature features includes: For the point cloud points remaining after two screenings of the local neighborhood , extract the maximum first - type curvature and second - type curvature within the local neighborhood as curvature features: ， ， Among them, , belongs to the local neighborhood of the point cloud point The corresponding first type of curvature and second type of curvature are selected for the maximum first type of curvature and second type of curvature in the local neighborhood, and used as the local neighborhood curvature feature of the point cloud point At the same time, the points within the local neighborhood range of the point cloud point are deleted to ensure that there is only one set of local neighborhood curvature features within the local neighborhood range of the point cloud point .

7. The ICP matching method based on the concave and convex characteristics of a mine tunnel according to claim 1, characterized in that Add the final feature residual constraint to the ICP objective function: ； Among them, is an IPC point - to - point match item, is a transformation relationship, including rotation and translation transformations, is the feature point i of the source point cloud frame s, is the matched feature point i of the target point cloud frame t; is the final feature residual constraint term, represents the global weight factor, balancing the contributions of the IPC point - to - point match item and the final feature residual constraint term, represents the final feature residual of the feature point with index i between the source point cloud frame s and the target point cloud frame t.

8. The ICP matching method based on the concave-convex characteristics of a mine tunnel according to claim 7, characterized in that Final feature residual: ； Among them, represents the final feature residual of the feature point with index i between the source point cloud frame s and the target point cloud frame t, represents the weight corresponding to the l-th feature element in the final feature, The dot represents the l-th feature element of the final feature of the feature point with index i in the source point cloud frame s and the target point cloud frame t.

9. The ICP matching method based on the concave and convex characteristics of the mine cave according to claim 7, characterized in that, After introducing the final feature residual constraint term, the final total Jacobian matrix of the transformation relationship T in the ICP algorithm is: ； Among them, , is the IPC point-to-point matching item Jacobian matrix, , is the Jacobian matrix of the final feature residual constraint term.

10. An ICP matching system based on the concave and convex characteristics of a mine tunnel, comprising: At least one processing unit, the processing unit is connected to a storage unit and an acquisition unit through a bus unit, characterized in that the storage unit stores a computer program and the data acquired by the acquisition unit, and the processing unit realizes the ICP matching method based on the concave-convex characteristics of the mine tunnel according to any one of claims 1-9 by running the computer program stored in the storage unit.

Images