CN116563350A - Tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm - Google Patents
Tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm Download PDFInfo
- Publication number
- CN116563350A CN116563350A CN202310516584.7A CN202310516584A CN116563350A CN 116563350 A CN116563350 A CN 116563350A CN 202310516584 A CN202310516584 A CN 202310516584A CN 116563350 A CN116563350 A CN 116563350A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- algorithm
- point
- ndt
- tunneling roadway
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000005641 tunneling Effects 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 48
- 230000009466 transformation Effects 0.000 claims abstract description 26
- 239000011159 matrix material Substances 0.000 claims abstract description 23
- 238000001914 filtration Methods 0.000 claims abstract description 17
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 238000005457 optimization Methods 0.000 claims abstract description 6
- 230000007613 environmental effect Effects 0.000 claims abstract description 5
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 4
- 238000013519 translation Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 abstract description 5
- 230000008447 perception Effects 0.000 abstract description 4
- 230000005484 gravity Effects 0.000 description 5
- 238000002474 experimental method Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004040 coloring Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004387 environmental modeling Methods 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000000465 moulding Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000010186 staining Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/32—Determination of transform parameters for the alignment of images, i.e. image registration using correlation-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention provides a tunneling roadway point cloud registration method based on a 3D NDT-ICP algorithm, which is used for improving the modeling accuracy of the tunneling roadway environment and meeting the intelligent perception requirement of the tunneling environment. Firstly, preprocessing tunneling roadway point clouds by a Voxel Grid filtering method, and reducing the number of the point clouds while maintaining the overall structure of the point clouds; then, carrying out coordinate transformation solution on the tunneling roadway point cloud by using a 3D NDT algorithm, carrying out parameter optimization on the resolution of the algorithm unit grid by combining the environmental characteristics of the tunneling roadway, transmitting the coordinate transformation parameters obtained by the solution to an ICP algorithm, and initializing a point cloud seven-parameter coordinate matrix in the ICP algorithm; and finally, introducing KD-Tree into an ICP algorithm to perform point-to-point search, and optimizing algorithm nonlinear objective function solution by using a Gauss-Newton method to finish accurate registration of tunneling roadway point cloud.
Description
Technical Field
The invention relates to a point cloud registration method of an underground tunneling roadway, which is used for improving the modeling progress of the tunneling roadway environment and meeting the intelligent perception requirement of the tunneling roadway.
Background
In recent years, the mining unbalance caused by the low tunneling efficiency in China has become the main problem faced by the high-mining high-efficiency mine. Because the underground environment is complex, the environment perception of the tunneling equipment is difficult, the motion space of the tunneling equipment is difficult to restrict, and the molding quality of a tunnel is guaranteed, the intelligent level of a tunneling working face is improved, and the intelligent perception of the tunneling tunnel environment is realized early. The tunneling roadway is used as an underground independent airtight space, and the environmental modeling of the tunneling roadway is a key basis for realizing intelligent sensing of the environment, and is a primary difficulty in intelligent construction of a tunneling working face. However, the tunneling roadway point cloud map is greatly affected by the limiting factors in the roadway in the generation process. Therefore, in order to further improve the modeling precision of the environment of the tunneling tunnel, registration is required for the initial point cloud of the tunneling tunnel.
The patent number is CN107644433A, the patent of the invention with the publication date of 2018.01.30 is an improved closest point iteration point cloud registration method, the patent of the invention provides two improved closest point iteration point cloud registration methods, and the method belongs to the field of image processing and three-dimensional point cloud registration. The method for improving the convergence rate of the nearest point iterative algorithm is to add a link for constructing a rotation matrix on the basis of the traditional nearest point iterative algorithm, wherein the constructed rotation matrix is used for participating in the generation of a new point cloud to be registered in the iterative process. The method for constructing the matrix comprises the steps of constructing the matrix Ri obtained by the current iteration and constructing the matrix according to the difference of three-axis attitude angles obtained by two adjacent iterations by using an Euler angle formula. The method can improve algorithm accuracy, but the method is large in calculation amount and still has some defects in improving efficiency.
The patent number is CN111161327A, the patent number is 2020.05.15, the patent discloses a point cloud registration method combining a rotating platform and ICP, the patent discloses a point cloud registration method combining the rotating platform and ICP, the surface characteristics of point clouds are not needed to be relied on, only the rotating angle and the rotating shaft position between the point clouds are needed to be known, and then the registration result is used as the initial position of an ICP registration algorithm to conduct further fine registration, so that the problem of failure of point cloud registration due to the fact that the characteristics of all angles are similar or even the same is effectively solved, the patent simplifies the operation steps of point cloud registration, experiments are more convenient, but problems of low matching precision and low efficiency still exist, and the final iteration result falls into the condition of local optimum.
The patent number is CN112017219A, the patent of the invention with the publication date of 2020.12.01 is a laser point cloud registration method, the patent provides a laser point cloud registration method, after two frames of source three-dimensional point clouds and target three-dimensional point clouds to be registered are respectively subjected to downsampling and non-ground point clouds are segmented, the source local features and the target local features of local features corresponding to the source three-dimensional point clouds and the target three-dimensional point clouds are extracted and are subjected to feature matching, the pose transformation of the source three-dimensional point clouds relative to the target three-dimensional point clouds is estimated based on a matching result, and the pose transformation matrix between the source three-dimensional point clouds and the target three-dimensional point clouds is obtained to finish the registration of the source three-dimensional point clouds and the target three-dimensional point clouds.
In view of the above shortcomings, in order to consider two factors of accuracy and efficiency of tunneling roadway point cloud registration, the inventor provides a tunneling roadway point cloud registration method based on a 3D NDT-ICP algorithm through continuous research and design, and combines the characteristics of high accuracy of the ICP algorithm and high efficiency of the NDT algorithm, so that the point cloud registration method is more comprehensive.
Disclosure of Invention
The invention aims to provide the tunneling roadway point cloud registration method with high precision, high efficiency and small error, which constructs a comprehensive and complete roadway environment map, provides conditions for the following submerged activities and realizes intelligent sensing of the tunneling roadway environment early.
The technical scheme of the invention is as follows: tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm comprises the following steps:
s1, preprocessing tunneling roadway point clouds by a Voxel Grid filtering method, and reducing the number of the point clouds while maintaining the overall structure of the point clouds;
s2, carrying out coordinate transformation solution on the tunneling roadway point cloud by using a 3D NDT algorithm;
s3, parameter optimization is carried out on the resolution of the algorithm unit grid by combining the environmental characteristics of the tunneling roadway;
s4, transmitting the coordinate transformation parameters obtained by solving to an ICP algorithm, and initializing a point cloud seven-parameter coordinate matrix in the ICP algorithm;
s5, introducing KD-Tree in an ICP algorithm to perform point-to-point search;
and S6, optimizing the algorithm nonlinear objective function solution by using a Gauss-Newton method, and finishing accurate registration of the tunneling roadway point cloud.
Further, the step S4 specifically includes:
s41, setting any point P (x) in the reference point cloud M P ,y P ,z P ) Any point Q (x) in the point cloud N to be registered P ,y P ,z P ) And P, Q meets P epsilon (M N O), Q epsilon (N N O), the external conversion relation of two independent 3D coordinate systems is further described by adopting a seven-parameter space similarity transformation model through an ICP algorithm, namely the following formula is adopted:
wherein t is x ，t y ，t z For three components along the coordinate axis, α, β, γ are three angular parameters rotated about the coordinate axis, and w is an inter-coordinate scale transformation factor, which is generally defaulted to 1.
S42, transmitting the final result of the initialized change parameters obtained in S3 to an ICP algorithm, initializing a rotation matrix R (alpha, beta, gamma) =R and a translation matrix [ t ] in the formula (1) x t y t z ] T ＝T；
S43, setting two groups of point clouds of a tunneling roadway to generate m groups of corresponding point pairs, and when the formula (2) obtains an optimal solution, solving a spatial position conversion relation by the two groups of point clouds, wherein the formula (2) is as follows:
compared with the prior art, the invention has the beneficial effects that:
(1) According to the invention, the tunneling roadway point cloud data is preprocessed by adopting the Voxel Grid filtering method, so that the integrity of the tunneling roadway point cloud structure is ensured, the number of point clouds is reduced, and a foundation is laid for subsequent point cloud registration.
(2) According to the invention, the environmental characteristics of the tunneling roadway are combined, and the parameter optimization is performed aiming at an NDT algorithm. And initializing an ICP algorithm seven-parameter coordinate transformation matrix by utilizing coordinate transformation parameters obtained by solving an NDT algorithm, introducing KD-Tree into the ICP algorithm to perform point-to-point search, and optimizing and solving an objective function by adopting a Gauss-Newton method to realize rapid and accurate registration of tunneling roadway point clouds.
Drawings
FIG. 1 is a flow of a conventional 3D NDT-ICP algorithm;
FIG. 2 is a flow chart of a modified 3D NDT-ICP algorithm of the present invention;
FIG. 3 is a graph of the result of a Voxel Grid filtering process;
fig. 4 is a graph of the result of the straight-through filtering process;
FIG. 5 is a time chart for registration of a 3D NDT algorithm;
FIG. 6 is a graph of registration s (p) values for a 3D NDT algorithm;
fig. 7 is a graph of the result of point cloud registration of the experiment.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and all the inventions which make use of the inventive concept are protected by the spirit and scope of the present invention as defined and defined in the appended claims to those skilled in the art.
The flow of the traditional 3D NDT-ICP algorithm is shown in fig. 1, the initial input point cloud pose is corrected by the 3D NDT algorithm, and then the point cloud with the corrected pose is registered by the ICP algorithm.
The traditional 3D NDT-ICP algorithm improves the point cloud registration accuracy to a certain extent, but loses the continuity of registration, based on the point cloud registration accuracy, the traditional 3D NDT-ICP algorithm is optimized and improved, the algorithm flow is shown in figure 2, and the method comprises the following steps:
s1, preprocessing tunneling roadway point clouds by a Voxel Grid filtering method, and reducing the number of the point clouds while maintaining the overall structure of the point clouds;
s2, carrying out coordinate transformation solution on the tunneling roadway point cloud by using a 3D NDT algorithm;
s3, parameter optimization is carried out on the resolution of the algorithm unit grid by combining the environmental characteristics of the tunneling roadway;
s4, transmitting the coordinate transformation parameters obtained by solving to an ICP algorithm, and initializing a point cloud seven-parameter coordinate matrix in the ICP algorithm;
s5, introducing KD-Tree in an ICP algorithm to perform point-to-point search;
and S6, optimizing the algorithm nonlinear objective function solution by using a Gauss-Newton method, and finishing accurate registration of the tunneling roadway point cloud.
The method for filtering the Grid of the Voxel carries out the preprocessing of the point cloud data, converts the tunneling roadway point cloud model into the Grid model, then carries out filtering and screening on the point cloud data in each Grid, ensures the integrity of the point cloud structure, and simultaneously removes a large number of unnecessary point clouds. The step S1 specifically includes:
s11, according to the point cloud data coordinate set, obtaining the maximum value x of X, Y, Z on three coordinate axes max 、y max 、z max And a minimum value x min 、y min 、z min ；
S12, obtaining the side length l of the minimum bounding box of the point cloud according to the maximum value and the minimum value of the X, Y, Z coordinate axes x 、l y 、l z . The formula is as follows:
s13, setting a voxel small grid side length cell, equally dividing X, Y, Z coordinate axes into M, N, L parts, and dividing a minimum bounding box into M, N and L voxel small grids, wherein sum=M, N and L;
s14, numbering each voxel with a small grid number which is (i, j, k); determining a voxel small grid to which each data point belongs;
s15, performing point cloud reduced filtering. Calculating the gravity center of each voxel small grid, and replacing all points in the voxel small grid with the gravity center; if the center of gravity does not exist, replacing all points in the small grid with data points closest to the center of gravity in the small grid;
wherein c ijk 、p i K are the center of gravity, data point and point number of the voxel small grid respectively.
After preprocessing the point cloud, solving the point cloud coordinate transformation by using a 3D NDT algorithm, wherein the step S2 specifically comprises the following steps:
s21, uniformly dividing a datum point cloud M (x, y, z) data sample into a plurality of three-dimensional voxel units with the same size and regularity by using a 3D NDT algorithm, and then expressing probability distribution of each three-dimensional point cloud position in the three-dimensional voxel units by normal distribution, wherein the expression is shown in a general expression (4):
wherein C is a covariance matrix of the three-dimensional point cloud in each voxel unit, and q is a mean value of the three-dimensional point cloud in each voxel unit; where c is a constant. q and C are specifically defined as shown in formula (5) and formula (6):
m in the formula i (i=1.,. The term., (i.), n) represents all three-dimensional point cloud data in voxel unit
S22, initializing transformation parameters of the point clouds M and N:
s23, mapping each point cloud sample of the point clouds N (x, y, z) to be registered into a data sample coordinate system of the reference point clouds M (x, y, z) according to coordinate transformation parameters, obtaining mapped point clouds as N ', summing probability distribution of mapping of each three-dimensional point in the N', and evaluating parameters of coordinate transformation:
s24, optimizing S (p) through a Hessian matrix method, and maximizing the value of S (p).
After parameter optimization, taking a point cloud transformation matrix obtained by solving a 3D NDT algorithm as an initial matrix of an ICP algorithm, and guaranteeing continuity of registration, wherein the step S4 is as follows:
s41, setting any point P (x) in the reference point cloud M P ,y P ,z P ) Any point Q (x) in the point cloud N to be registered P ,y P ,z P ) And P, Q meets P epsilon (M N O), Q epsilon (N N O), the external conversion relation of two independent 3D coordinate systems is further described by adopting a seven-parameter space similarity transformation model through an ICP algorithm, namely the following formula is adopted:
wherein t is x ，t y ，t z For three components along the coordinate axis, α, β, γ are three angular parameters rotated about the coordinate axis, and w is an inter-coordinate scale transformation factor, which is generally defaulted to 1.
S42, transmitting the final result of the initialized change parameters obtained in S3 to an ICP algorithm, initializing a rotation matrix R (alpha, beta, gamma) =R and a translation matrix [ t ] in the formula (1) x t y t z ] T ＝T；
S43, setting two groups of point clouds of a tunneling roadway to generate m groups of corresponding point pairs, and when an optimal solution is obtained in a formula (10), solving a spatial position conversion relation by the two groups of point clouds, wherein the formula (10) is as follows:
and introducing KD-Tree into the ICP algorithm to perform point-to-point search, and then optimizing the algorithm nonlinear objective function solution by using a Gauss-Newton method to finish accurate registration of tunneling roadway point cloud.
Examples:
the two groups of laser point clouds are preprocessed by using a Voxel Grid filtering algorithm and a straight-through filtering algorithm respectively, the experimental results are shown in fig. 3 and 4, the number of the laser point clouds is reduced by the straight-through filtering method to be larger than that of the Voxel Grid filtering method, however, the tunneling roadway point cloud structure is easily damaged by the straight-through filtering method to lose integrity, and the number of the point clouds is reduced by the Voxel Grid filtering method while the structural integrity of the point clouds is ensured, so that the method is more suitable for preprocessing tunneling roadway point cloud data. And recording the two groups of laser point clouds obtained after the filtering of the Voxel Grid algorithm as experimental point clouds for subsequent experimental study.
Fig. 5 and 6 are time charts for registration and s (p) value charts when the 3D NDT algorithm is used in experiments, and the unit cell resolutions of the 3D NDT algorithm are set to 0.25m, 0.5m, 1m, 1.5m and 2m respectively. Based on two groups of experimental point clouds for experiments, algorithm parameter selection basis is provided for tunneling roadway environment modeling, as can be seen from fig. 5, 3D NDT algorithm is used for registering tunneling roadway point clouds, and when unit grid resolution is set to be 0.5m, algorithm time is shortest; as can be seen from fig. 6, the 3D NDT algorithm is used for registering the tunneling roadway point cloud, and when the unit grid resolution is set to 0.5m, the algorithm s (p) value reaches the maximum, and the registration accuracy is higher. With reference to fig. 5 and fig. 6, the cost and the precision requirement of the down-hole point cloud registration time are fully considered, the point cloud registration time is shortened as much as possible under the condition of ensuring the point cloud registration precision, the resolution of the 3D NDT algorithm unit is set to be 0.5m, at this time, the 3D NDT algorithm is shortest in registration time, and the registration precision is highest.
In order to verify the rapidness and accuracy of the algorithm, the experimental point cloud is registered by using an NDT algorithm, an ICP algorithm, a traditional 3D NDT-ICP algorithm and the 3D NDT-ICP algorithm proposed herein, and the registration result is shown in fig. 7, and as can be seen, the registration is performed aiming at the experimental point cloud, the registration result of the NDT algorithm is poor, and the point cloud contour is not clear enough; the ICP algorithm registration result is general, and the point cloud outline is clear; the traditional 3D NDT-ICP algorithm has the advantages that the registering result is good, the outline of the point cloud is clear, but the coloring degree of the point cloud is not high; compared with an NDT algorithm, an ICP algorithm and a traditional 3D NDT-ICP algorithm, the 3D NDT-ICP algorithm provided by the invention has relatively best registration result, clear point cloud outline and higher staining degree.
From the comparison, when the 3D NDT-ICP algorithm provided by the invention is used for calibrating the experimental point cloud, the point cloud registration error of a tunneling roadway is reduced, the point cloud registration time is saved, and the larger the experimental point cloud data volume is, the more obvious the property of the 3D NDT-ICP algorithm provided by the invention is.
Claims (6)
1. The tunneling roadway point cloud matching method based on the 3D NDT-ICP algorithm is characterized by comprising the following steps of:
s1, preprocessing tunneling roadway point clouds by a Voxel Grid filtering method, and reducing the number of the point clouds while maintaining the overall structure of the point clouds;
s2, carrying out coordinate transformation solution on the tunneling roadway point cloud by using a 3D NDT algorithm;
s3, parameter optimization is carried out on the resolution of the algorithm unit grid by combining the environmental characteristics of the tunneling roadway;
s4, transmitting the coordinate transformation parameters obtained by solving to an ICP algorithm, and initializing a point cloud seven-parameter coordinate matrix in the ICP algorithm;
s5, introducing KD-Tree in an ICP algorithm to perform point-to-point search;
and S6, optimizing the algorithm nonlinear objective function solution by using a Gauss-Newton method, and finishing accurate registration of the tunneling roadway point cloud.
2. The tunneling roadway point cloud registration method based on the 3D NDT-ICP algorithm of claim 1, wherein the step S2 specifically includes:
s21, firstly, uniformly dividing a datum point cloud M (x, y, z) data sample into a plurality of three-dimensional voxel units with the same size and regularity by using a 3D NDT algorithm, and then expressing probability distribution of each three-dimensional point cloud position in the three-dimensional voxel units by normal distribution, wherein the expression is as follows:
wherein C is a covariance matrix of the three-dimensional point cloud in each voxel unit, and q is a mean value of the three-dimensional point cloud in each voxel unit; where c is a constant. q and C are specifically defined as follows:
m in the formula i (i=1.,. The term., (i.), n) represents all three-dimensional point cloud data in the voxel unit.
S22, initializing transformation parameters of the point clouds M and N:
s23, mapping each point cloud sample of the point clouds N (x, y, z) to be registered into a data sample coordinate system of the reference point clouds M (x, y, z) according to coordinate transformation parameters, obtaining mapped point clouds as N ', summing probability distribution of mapping of each three-dimensional point in the N', and evaluating parameters of coordinate transformation:
3. the tunneling roadway point cloud matching method based on the 3D NDT-ICP algorithm of claim 2, wherein for step S23, S (p) is optimized by the Hessian matrix method to maximize the S (p) value.
4. The tunneling roadway point cloud matching method based on the 3D NDT-ICP algorithm according to claim 1, wherein the unit resolution of the 3D NDT algorithm is usually 0.5m-2m for step S3, and the modeling result is ideal for laser scanning equipment. The experimental result shows that when the resolution of the unit grid is set to be 0.5m, the algorithm time is shortest, the value of the algorithm s (p) is the largest, and the matching precision is the highest.
5. The tunneling roadway point cloud matching method based on the 3D NDT-ICP algorithm according to claim 1, wherein the step S4 is specifically as follows:
s41, setting any point P (x) in the reference point cloud M P ,y P ,z P ) Any point Q (x) in the point cloud N to be registered P ,y P ,z P ) And P, Q meets P epsilon (M N O), Q epsilon (N N O), the external conversion relation of two independent 3D coordinate systems is further described by adopting a seven-parameter space similarity transformation model through an ICP algorithm, namely the following formula is adopted:
wherein t is x ，t y ，t z For three components along the coordinate axis, α, β, γ are three angular parameters rotated about the coordinate axis, and w is an inter-coordinate scale transformation factor, which is generally defaulted to 1.
S42, transmitting the final result of the initialized change parameters obtained in S3 to an ICP algorithm, initializing a rotation matrix R (alpha, beta, gamma) =R and a translation matrix [ t ] in the above formula x t y t z ] T ＝T；
S43, setting two groups of point clouds of a tunneling roadway to generate m groups of corresponding point pairs, and when the optimal solution is obtained by the following formula, solving the spatial position conversion relation by the two groups of point clouds, wherein the formula is as follows:
6. the tunneling roadway point cloud matching method based on the 3D NDT-ICP algorithm, according to claim 1, wherein the rotation errors of the point cloud X axis, the Y axis and the Z axis are controlled within 0.6 degrees, and the translation errors of the point cloud X axis, the Y axis and the Z axis are controlled to be about 2.5 cm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310516584.7A CN116563350A (en) | 2023-05-09 | 2023-05-09 | Tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310516584.7A CN116563350A (en) | 2023-05-09 | 2023-05-09 | Tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116563350A true CN116563350A (en) | 2023-08-08 |
Family
ID=87487394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310516584.7A Pending CN116563350A (en) | 2023-05-09 | 2023-05-09 | Tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116563350A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117518197A (en) * | 2024-01-08 | 2024-02-06 | 太原理工大学 | Contour marking method for underground coal mine tunneling roadway |
-
2023
- 2023-05-09 CN CN202310516584.7A patent/CN116563350A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117518197A (en) * | 2024-01-08 | 2024-02-06 | 太原理工大学 | Contour marking method for underground coal mine tunneling roadway |
| CN117518197B (en) * | 2024-01-08 | 2024-03-26 | 太原理工大学 | Contour marking method for underground coal mine tunneling roadway |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113077497A (en) | Tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm | |
| CN112017220B (en) | Point cloud accurate registration method based on robust constraint least square algorithm | |
| CN111145227B (en) | Iterative integral registration method for space multi-view point cloud of underground tunnel | |
| CN116563350A (en) | Tunneling roadway point cloud registration method based on 3D NDT-ICP algorithm | |
| CN110796694A (en) | Fruit three-dimensional point cloud real-time acquisition method based on KinectV2 | |
| CN103673995A (en) | Calibration method of on-orbit optical distortion parameters of linear array push-broom camera | |
| CN113834625B (en) | Aircraft model surface pressure measuring method and system | |
| CN105118090A (en) | Adaptive point-cloud filtering method for complex terrain structure | |
| CN116204990B (en) | Three-dimensional measured data driven precise coordination repair method for large-scale framework of aircraft | |
| CN112465966A (en) | Cliff three-dimensional modeling method integrating oblique photogrammetry and three-dimensional laser scanning | |
| CN114659514A (en) | LiDAR-IMU-GNSS fusion positioning method based on voxelized fine registration | |
| CN107871139A (en) | A kind of neighborhood keeps the Method of Data with Adding Windows of embedded innovatory algorithm | |
| Song et al. | Curved buildings reconstruction from airborne LiDAR data by matching and deforming geometric primitives | |
| CN114279324B (en) | All-dimensional intelligent detection method for appearance quality of prefabricated part | |
| CN110047146B (en) | Error correction method based on single revolving body image 3D restoration | |
| CN114998395A (en) | Effective embankment three-dimensional data change detection method and system | |
| CN117315169A (en) | Live-action three-dimensional model reconstruction method and system based on deep learning multi-view dense matching | |
| CN114529703A (en) | Entropy increase optimization-based point cloud global matching method for large complex components | |
| CN111274732B (en) | Grid repairing method based on 'connection relation-position' iterative optimization | |
| CN115930800B (en) | Tunnel face displacement field monitoring method based on three-dimensional laser point cloud | |
| CN116401734A (en) | Brick-concrete building structure length calculation method based on point cloud | |
| CN116152512A (en) | Height measuring and calculating method based on building shadow restoration | |
| CN114821290A (en) | Aviation image building contour extraction method based on machine learning | |
| CN108387897B (en) | Projectile body positioning method based on improved Gauss Newton-genetic hybrid algorithm | |
| CN112381908B (en) | Topographic scanning point cloud boundary line extraction method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |