CN115950427B - Robot navigation scanning matching algorithm based on combination of ICP and genetic optimization - Google Patents
Robot navigation scanning matching algorithm based on combination of ICP and genetic optimization Download PDFInfo
- Publication number
- CN115950427B CN115950427B CN202211666564.XA CN202211666564A CN115950427B CN 115950427 B CN115950427 B CN 115950427B CN 202211666564 A CN202211666564 A CN 202211666564A CN 115950427 B CN115950427 B CN 115950427B
- Authority
- CN
- China
- Prior art keywords
- icp
- point
- algorithm
- scan
- genetic
- 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.)
- Active
Links
- 230000002068 genetic effect Effects 0.000 title claims abstract description 41
- 238000005457 optimization Methods 0.000 title claims abstract description 22
- 238000000034 method Methods 0.000 claims abstract description 19
- 238000013519 translation Methods 0.000 claims abstract description 18
- 210000000349 chromosome Anatomy 0.000 claims abstract description 15
- 239000011159 matrix material Substances 0.000 claims abstract description 10
- 238000005259 measurement Methods 0.000 claims abstract description 4
- 230000035772 mutation Effects 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 5
- 230000033001 locomotion Effects 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 108090000623 proteins and genes Proteins 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 2
- 238000002921 genetic algorithm search Methods 0.000 claims 1
- 238000001514 detection method Methods 0.000 abstract description 13
- 238000013507 mapping Methods 0.000 abstract description 4
- 230000007547 defect Effects 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 13
- 230000000694 effects Effects 0.000 description 6
- 238000009826 distribution Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000036544 posture Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000004659 sterilization and disinfection Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- 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
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention relates to a robot navigation scanning matching algorithm based on combination of ICP and genetic optimization, which comprises the steps of running a genetic algorithm, finding (x, y) translation and rotation phi, obtaining optimal alignment between two different scans, encoding each parameter of scanning positions (x, y, phi) into a bit string in a chromosome, discretizing a problem space in a finite solution space, setting search area limit and resolution, and setting according to a ranging error model under the conditions of gesture tracking and ranging measurement; giving the pose result output by the genetic algorithm to an ICP algorithm, wherein the ICP algorithm minimizes an objective function; then searching for a corresponding point; optimizing the rotation matrix R and the translation matrix T by using a least square method; finally, iterative optimization is carried out continuously, so that accurate pose matching can be realized. In the technical scheme provided by the invention, the advantages of the genetic algorithm and the ICP algorithm are combined, the mutual defects are overcome, and the optimal initial value required by the ICP algorithm can be searched in a short period in the global search space, so that the high-efficiency and robust closed-loop detection is realized, and the laser SLAM also realizes better mapping.
Description
Technical Field
The invention relates to the field of robot navigation, in particular to a robot navigation scanning matching algorithm based on combination of ICP and genetic optimization.
Background
Due to the demands of modern society for improving production efficiency and the directions of people for intelligent life style, mobile robot technology is rapidly developed, and the frequency of occurrence in life is higher and higher, such as greeting robots in high-end hotels, disinfection robots in hospitals during epidemic situations, rescue robots for rescue and disaster relief and the like. In many application scenarios of mobile robots, the surrounding environment is not constant, sometimes even unknown. As a key technology for realizing autonomous navigation of a mobile robot in an unknown environment, SLAM (instant localization and mapping) is therefore particularly important.
The goal of SLAM is to solve the problems of building a map in an unknown environment, constantly updating the map, and locating objects in the map. The laser SLAM based on the graph optimization method is divided into a front end and a rear end. The front end is responsible for realizing laser mileage calculation, including interframe matching, loop detection and the like, and aims to construct an error function for rear end optimization according to the observation of a laser radar. The rear end is responsible for optimizing, such as optimizing the pose according to the constraint established by the front end; optimizing the pose graph according to the loop detection result; release maps, etc. The rear end is used for correcting the pose, the map and the like. After motion estimation, when the initial pose of the first frame image is fixed, the pose of each subsequent frame can be theoretically deduced, but this causes accumulated errors (the pose errors estimated for two adjacent frames are accumulated to the next frame so that the errors become larger). Therefore, loop-back detection is required to eliminate this accumulated error. In graph-based SLAM, closed loop detection is a key step in obtaining global pose constraints, and the real-time performance of this process ensures that the back-end optimization of the current frame can be successfully completed before the data at the next moment arrives.
The ICP (iterative closest point) algorithm is an inter-frame matching algorithm that can also be used for closed loop detection, and is the core of laser SLAM. Because of the influence of factors such as different acquisition devices, different shooting visual angles and the like, even the point clouds obtained by the same object have larger differences, mainly the rotation or translation changes. For two images in a group of image data sets, one image needs to be mapped to the other image by searching for a space transformation, so that points corresponding to the same position in space in the two images are in one-to-one correspondence, and the purpose of information fusion is achieved. However, using only this method does not eliminate errors caused by the completed partial matches, and accumulated errors may cause significant drift when constructing a large scale map. And the convergence speed is low, and when the initial postures of the source point cloud and the target point cloud are poor, the problems that the local optimal solution is easy to fall in exist.
NDT is an algorithm that converts a first frame point cloud to a grid map, each grid calculates a normal distribution of points therein, and thus converts the first frame point cloud to a piecewise continuously-derivable probability density function that is represented by the grids, and describes the local properties of the point cloud using the normal distribution probability density function. And then projecting the second frame of point cloud into the grid map, calculating the probability of the second frame of point cloud in the grid map, finding the best matching pose through a Newton method, maximizing the probability of the second frame of point cloud after being projected into the grid map, and finally completing point cloud registration. The specific flow is as follows: and establishing a grid map of the first frame point cloud, and enabling the probability obtained by projecting the second frame point cloud to the grid map to be maximized through Newton method iteration. However, for structured point clouds, NDT does not work well with a block grid gaussian distribution assumption, but instead the point-line ICP or point-plane ICP effect is better. For unstructured large-scale point clouds, NDT is faster, initial robustness depends on the size of the grid, the larger the accuracy is, the worse the initial robustness is, but the higher the initial robustness is, and otherwise, the higher the initial robustness is.
The biggest difference in ICP variation over ICP is that it improves the error equation. ICP is the point-to-point distance as an error and ICP variation is the distance from the point to its nearest two point connection. The working flow is as follows: projecting the current frame data to a reference frame coordinate system according to the initial pose; for the point of the current frame, find the nearest two points (j 1, j 2) in the reference frame using nearest neighbor rule; calculating errors and removing Outliers with overlarge errors; and solving a matrix R and a translation matrix according to the minimized error function. This approach causes significant computational expense as map scales grow until it is affordable. In addition, ICP and its variants are sensitive to initial values, and are not used alone, and are prone to falling into local circulation.
Cartographer handles larger spaces by creating many small submaps, optimizes the pose of all scans and subgraphs, and follows sparse pose adjustments. All created submap and current laser scan will be used as matches for loop-back detection. If the current scan and all created submap are sufficiently close in distance, then the closed loop will be found by some matching strategy. In order to reduce the calculation amount and improve the efficiency of real-time loop detection, the Cartographer uses a branch and db optimization method to perform optimization search. In this way, the memory occupation is larger and the algorithm volume is larger.
Disclosure of Invention
The invention aims to provide a robot navigation scanning matching algorithm based on combination of ICP and genetic optimization, which can realize efficient and robust closed-loop detection.
The technical scheme adopted by the invention is as follows.
The robot navigation scanning matching algorithm based on the combination of ICP and genetic optimization is characterized by comprising the following operations:
s10: running a genetic algorithm to find (x, y) translation and rotation phi, obtaining optimal alignment between two different scans, each parameter of a scan position (x, y, phi) being encoded in a chromosome as a bit string, the problem space being discretized in a finite solution space, setting search area limits and resolution, setting according to a ranging error model in case of gesture tracking and ranging measurements being possible;
s20: giving the pose result output by the genetic algorithm to an ICP algorithm, wherein the ICP algorithm minimizes an objective function; then searching for a corresponding point; optimizing the rotation matrix R and the translation matrix T by using a least square method;
s30: finally, iterative optimization is carried out continuously, so that accurate pose matching can be realized.
The detailed operation is as follows: in step S10: the search area restriction is set according to the problem at hand.
In step S10: the resolution is set according to the expansion of the search area and the number of bits per gene.
In step S20: the objective function is the sum of squares of the euclidean distances between all corresponding points.
The corresponding point identification criteria in step S20 are: as long as the distance between the two point clouds is smaller than the threshold value, the corresponding point can be identified. That is, as long as the distance between the two point clouds is minimized, the corresponding point can be recognized.
In the technical scheme provided by the invention, the advantages of the genetic algorithm and the ICP algorithm are combined, the mutual defects are overcome, and the optimal initial value required by the ICP algorithm can be searched in a short period in the global search space, so that the high-efficiency and robust closed-loop detection is realized, and the laser SLAM also realizes better mapping. Moreover, when large scanning errors occur very commonly in mobile robots, the proposed algorithm successfully overcomes the effects of large positioning errors.
Drawings
FIG. 1 is a schematic diagram of the present invention.
Detailed Description
The present invention will be specifically described with reference to examples below in order to make the objects and advantages of the present invention more apparent. It should be understood that the following text is intended to describe only one or more specific embodiments of the invention and does not limit the scope of the invention strictly as claimed.
As shown in fig. 1, the robot navigation scanning matching algorithm based on the combination of ICP and genetic optimization provided by the invention specifically operates as follows.
S10: first, a genetic algorithm is run to find (x, y) translation and rotation phi, an optimal alignment is obtained between two different scans, each parameter of the scan position (x, y, phi) is encoded as a bit string in the chromosome, the problem space is discretized in a finite solution space, the resolution is set according to the expansion of the search area and the number of bits per gene, the search area limit is set according to the problem at hand, and in case of gesture tracking and ranging measurements can be performed, the setting is performed according to the ranging error model.
The function of step S10 is to determine the optimum initial value of the ICP algorithm by using a genetic algorithm, and the detailed operation is as follows.
S11: a search window is determined.
For scan matching problems, using W as the search window, setting the linear and angular search window offsets W x ＝W y ＝8m，W θ =30°. The resolution of linear search is r t Selecting an angular search resolution r θ The method comprises the following steps:
where dmax is the maximum range of scan points.
Obtaining an integer number covering the steps of a given linear and angular search window size:
s12: chromosome coding.
Extending a two-dimensional linear search space into one-dimensional representation, i.e. conversion setLinear index->Is one-dimensional (I)>The range of (2) is 0 to 2Wx.2Wy. Let the coordinates of the search center be epsilon 0 (x 0 ，y 0 ) Obtaining the linear index epsilon of the point p (x p ，y p )：
UsingRepresentation->Maximum value of 2n x ·2n y . For the angle search window, the angle index +.>Is created to represent the rotation component and the maximum +.>
S13: in the encoding step, it is ensured that the chromosomes are globally optimized to represent all x, y and θ in the feasible solution.
Letting the length of the chromosomeFor l=l 1 +l 2 Wherein the length of the left part is l 1 The length on the right is l 2 . To ensure that the genetic algorithm can reach any position of the feasible solution, l 1 And l 2 Must meet the requirements ofAn adaptive function. For scan matching, the response value R v Is regarded as an adaptive function in genetic algorithm, gol i Probability value p being a scan point i And goalmax is the maximum response value of all scan points to surrounding objects.
S15: genetic operators.
Genetic operators include mutation operators, selection operators, and crossover operators. Conventional roulette strategies are used for select operations and two-point crossover is used for crossover operations. The mutation operation adopts a basic position mutation method. Considering the discontinuity of the matching process, reducing the crossover rate and expanding the mutation rate can prevent the genetic algorithm from falling into local iterations.
S16: chromosome decoding.
In the chromosomeAfter the crossover, mutation and selection, the corresponding linear and angular indices need to be calculated from the following equations, which is the chromosome decoding process.
S20: secondly, giving a pose result output by the genetic algorithm to an ICP algorithm, wherein the ICP algorithm minimizes an objective function, and the objective function is the square sum of Euclidean distances among all corresponding points; then searching corresponding points, wherein the corresponding points are identified as follows: as long as the distance between the two point clouds is smaller than a threshold value, the corresponding point can be identified; the rotation matrix R and the translation matrix T are then optimized using the least squares method.
S30: finally, iterative optimization is carried out continuously, so that accurate pose matching can be realized.
Step S30 is ICP closed loop detection based on genetic algorithm, and the detailed operation is.
One point of the reference scan and one point of the new scan are set to sample the same point in the environment but from different locations. Consider a method in which a coordinate system (x 1 Point P in y) 1 (x 1 ，y 1 ) And a coordinate system (x 2 ，y 2 ) Point P in (a) 2 (x 2 ，y 2 ). The two coordinate systems differ in rotation θ and translation, the translation being determined by the offset vectorAnd (3) representing. Is composed of a series of points P 1 And a series of newly scanned points P obtained after a rotation and translation movement of the robot 2 Is composed of the components. Point P using rotation/translation operations 1 At (x) 2 ，y 2 ) The coordinate system is shown as follows:
if at the first scan midpoint P1 (x 1 ，y 1 ) Corresponding to point P in the second scan 2 (x 2 ，y 2 ) The following errors are defined:
the scan matching problem is to estimate the motion linearity index of the robot by finding two factorsAnd corner index->The two scans are made to overlap as much as possible. The optimal initial value required by the ICP algorithm is quickly searched in the global search space after the previous genetic algorithm is subjected to chromosome decoding>And->The problem that the ICP algorithm is easy to fall into local optimum is solved. And may then be performed iteratively as follows. The first operation of each iteration is for each point of the two sequences (corresponding problem). This pairing is performed by finding several points if the value +.>At a given actual value ∈ ->And->Lower minimized, then P i Corresponds to P j Where N is the number of points in the scan. This defines a corresponding index function I such that i=i (j). The second operation of the iteration is by solving +.>Two factors of refinement->And->The iteration proceeds until convergence. The final operation is to apply the conversion to the new scan. In this way, the two scans are aligned and a map of the environment is constructed.
The inter-frame matching algorithm directly influences the effect of the laser SLAM, the back-end optimization only eliminates errors accumulated in the process, and the more accurate the pose of inter-frame matching estimation is, the better the later image construction effect is. The ICP algorithm is a common interframe matching algorithm, can obtain very accurate registration effect, does not need to divide and extract features of the processed point set, and can obtain good algorithm convergence under the condition of good initial value. However, the ICP algorithm is greatly affected by the initial values, and when the initial postures of the source point cloud and the target point cloud are poor, the problem that the local optimal solution is easily trapped exists. The genetic algorithm is a heuristic algorithm which can find the optimal solution in the global search space through genetic iteration in a short time and is not easy to sink into a local minimum. And the pose result output by the genetic algorithm is given to the ICP algorithm by utilizing the global optimizing capability of the genetic algorithm to realize accurate pose matching, namely, the advantages of the genetic algorithm and the ICP algorithm are combined to realize a high-efficiency closed-loop detection function.
Therefore, the advantages of the genetic algorithm and the ICP algorithm are combined, the mutual defects are overcome, the optimal initial value required by the ICP algorithm can be searched in a short period in the global search space, and therefore high-efficiency and robust closed-loop detection is achieved, and the laser SLAM also achieves better mapping. Moreover, when large scanning errors occur very commonly in mobile robots, the proposed algorithm successfully overcomes the effects of large positioning errors.
Measures not described in detail in the present invention are implemented with reference to existing genetic algorithms and ICP algorithms.
The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which are intended to be comprehended within the scope of the present invention. Structures, mechanisms, and methods of operation not specifically described and illustrated in the present invention are implemented by conventional means in the art unless specifically described and limited.
Claims (5)
1. The robot navigation scanning matching algorithm based on the combination of ICP and genetic optimization is characterized by comprising the following operations:
s10: running a genetic algorithm to find (x, y) translation and rotation phi, obtaining optimal alignment between two different scans, each parameter of a scan position (x, y, phi) being encoded in a chromosome as a bit string, the problem space being discretized in a finite solution space, setting search area limits and resolution, setting according to a ranging error model in case of gesture tracking and ranging measurements being possible;
s20: giving the pose result output by the genetic algorithm to an ICP algorithm, wherein the ICP algorithm minimizes an objective function; then searching for a corresponding point; optimizing the rotation matrix R and the translation matrix T by using a least square method;
s30: finally, iterative optimization is carried out continuously, so that accurate pose matching can be realized;
step S10 includes the following operations:
s11: determining a search window;
setting the linear and angular search window offsets W using W as the search window x ＝W y ＝8m，W θ =30°, linear search resolution r t Selecting an angular search resolution r θ The method comprises the following steps:
where dmax is the maximum range of scan points;
obtaining an integer number covering the steps of a given linear and angular search window size:
s12: chromosome coding;
the two-dimensional linear search space is extended to one-dimensional representation,is in the range of 0 to 2Wx.2Wy, and the coordinates of the search center are epsilon 0 (x 0 ，y 0 ) Obtaining the linear index epsilon of the point p (x p ，y p )：
UsingRepresentation->Maximum value of 2n x ·2n y For the angle search window, the angle index +.>Is created to represent the rotation component and the maximum +.>
S13: in the encoding step, it is ensured that the chromosomes are globally optimized to represent all x, y and θ in the feasible solution;
let the length of the chromosome be l=l 1 +l 2 Wherein the length of the left part is l 1 The length on the right is l 2 To ensure that the genetic algorithm can reach any position of the feasible solution, l 1 And l 2 Must meet the requirements ofAdaptive function, response value R for scan matching v Is regarded as an adaptive function in genetic algorithm, gol i Probability value p being a scan point i And goalmax is the maximum response value of all scan points to surrounding objects,
s14: genetic operators;
genetic operators comprise mutation operators, selection operators and crossover operators;
s15: chromosome decoding;
in the chromosomeAfter the crossover, mutation and selection, the corresponding linear and angular indices need to be calculated from the following equations, which is the chromosome decoding process;
the specific operation of step S30 is:
a point of the reference scan and a point of the new scan are set to sample the same point in the environment but from different locations, taking into account a point in the coordinate system (x 1 Point P in y) 1 (x 1 ，y 1 ) And a coordinate system (x 2 ，y 2 ) Point P in (a) 2 (x 2 ，y 2 ) The two coordinate systems differ in rotation θ and translation, the translation being determined by the offset vectorIs represented by a series of points P 1 And a series of newly scanned points P obtained after a rotation and translation movement of the robot 2 Is composed of, using rotation/translation operation, point P 1 At (x) 2 ，y 2 ) The coordinate system is shown as follows:
if at the first scan midpoint P 1 (x 1 ，y 1 ) Corresponding to point P in the second scan 2 (x 2 ，y 2 ) The following errors are defined:
estimating a kinematic linearity index of a robot by finding two factorsAnd corner index->The two scans are overlapped as much as possible, and the genetic algorithm searches the optimal initial value required by the ICP algorithm in the global search space after chromosome decoding>And->And then iteratively performed as follows;
the first operation of each iteration is for each point of the two sequences, this pairing is performed by finding several points if the value isAt a given actual value ∈ ->And->Lower minimized, then P i Corresponds to P j Where N is the number of points in the scan;
defining a corresponding index function I such that i=i (j), the second operation of the iteration is by solving for
Two factors of refinement->And->
The iteration proceeds until convergence, the final operation is to apply the transformation to the new scan, in such a way that the two scans are aligned and a map of the environment is constructed.
2. The robot navigation scan matching algorithm based on a combination of ICP and genetic optimization according to claim 1, wherein in step S10: the search area restriction is set according to the problem at hand.
3. The robot navigation scan matching algorithm based on a combination of ICP and genetic optimization according to claim 1, wherein in step S10: the resolution is set according to the expansion of the search area and the number of bits per gene.
4. The robot navigation scan matching algorithm based on a combination of ICP and genetic optimization according to claim 1, wherein in step S20: the objective function is the sum of squares of the euclidean distances between all corresponding points.
5. The robot navigation scan matching algorithm based on a combination of ICP and genetic optimization of claim 1, wherein the corresponding point identification criteria in step S20 is: as long as there is a distance in the two point clouds that is less than the threshold, the corresponding point is identified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211666564.XA CN115950427B (en) | 2022-12-23 | 2022-12-23 | Robot navigation scanning matching algorithm based on combination of ICP and genetic optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211666564.XA CN115950427B (en) | 2022-12-23 | 2022-12-23 | Robot navigation scanning matching algorithm based on combination of ICP and genetic optimization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115950427A CN115950427A (en) | 2023-04-11 |
| CN115950427B true CN115950427B (en) | 2024-02-09 |
Family
ID=87287069
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211666564.XA Active CN115950427B (en) | 2022-12-23 | 2022-12-23 | Robot navigation scanning matching algorithm based on combination of ICP and genetic optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115950427B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600639A (en) * | 2016-12-09 | 2017-04-26 | 江南大学 | Genetic algorithm and adaptive threshold constraint-combined ICP (iterative closest point) pose positioning technology |
| CN109974707A (en) * | 2019-03-19 | 2019-07-05 | 重庆邮电大学 | A kind of indoor mobile robot vision navigation method based on improvement cloud matching algorithm |
| CN111784770A (en) * | 2020-06-28 | 2020-10-16 | 河北工业大学 | Three-dimensional attitude estimation method in disordered grabbing based on SHOT and ICP algorithm |
| CN113587933A (en) * | 2021-07-29 | 2021-11-02 | 山东山速机器人科技有限公司 | Indoor mobile robot positioning method based on branch-and-bound algorithm |
-
2022
- 2022-12-23 CN CN202211666564.XA patent/CN115950427B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600639A (en) * | 2016-12-09 | 2017-04-26 | 江南大学 | Genetic algorithm and adaptive threshold constraint-combined ICP (iterative closest point) pose positioning technology |
| CN109974707A (en) * | 2019-03-19 | 2019-07-05 | 重庆邮电大学 | A kind of indoor mobile robot vision navigation method based on improvement cloud matching algorithm |
| CN111784770A (en) * | 2020-06-28 | 2020-10-16 | 河北工业大学 | Three-dimensional attitude estimation method in disordered grabbing based on SHOT and ICP algorithm |
| CN113587933A (en) * | 2021-07-29 | 2021-11-02 | 山东山速机器人科技有限公司 | Indoor mobile robot positioning method based on branch-and-bound algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 基于遗传算法的三维曲面配准;刘纯国等;《计算机应用》;第111-112页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115950427A (en) | 2023-04-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111536964B (en) | Robot positioning method and device, and storage medium | |
| CN114862949B (en) | Structured scene visual SLAM method based on dot-line surface characteristics | |
| CN112230243B (en) | Indoor map construction method for mobile robot | |
| CN108253958A (en) | A kind of robot real-time location method under sparse environment | |
| Bosse | ATLAS: a framework for large scale automated mapping and localization | |
| CN113108773A (en) | Grid map construction method integrating laser and visual sensor | |
| CN114018248B (en) | Mileage metering method and image building method integrating code wheel and laser radar | |
| CN111540005A (en) | Loop detection method based on two-dimensional grid map | |
| CN108537263B (en) | Grid map fusion method based on maximum public subgraph | |
| CN116592897B (en) | Improved ORB-SLAM2 positioning method based on pose uncertainty | |
| CN116758153A (en) | Multi-factor graph-based back-end optimization method for accurate pose acquisition of robot | |
| He et al. | Online semantic-assisted topological map building with LiDAR in large-scale outdoor environments: Toward robust place recognition | |
| CN113532439B (en) | Synchronous positioning and map construction method and device for power transmission line inspection robot | |
| CN114299386A (en) | Laser SLAM method integrating laser odometer and loop detection | |
| CN115950427B (en) | Robot navigation scanning matching algorithm based on combination of ICP and genetic optimization | |
| CN117570968A (en) | Map construction and maintenance method and device based on visual road sign and storage medium | |
| CN112612034A (en) | Pose matching method based on laser frame and probability map scanning | |
| CN111951341A (en) | Closed loop detection improvement method based on RGB-D SLAM | |
| CN117053779A (en) | Tightly coupled laser SLAM method and device based on redundant key frame removal | |
| KR101054520B1 (en) | How to recognize the location and direction of the indoor mobile robot | |
| CN115962773A (en) | Method, device and equipment for synchronous positioning and map construction of mobile robot | |
| Peng et al. | Proximity-distance mapping and Jaya optimization algorithm based on localization for wireless sensor network | |
| CN112614162B (en) | Indoor vision rapid matching and positioning method and system based on space optimization strategy | |
| Zhi et al. | Research on Cartographer Algorithm based on Low Cost Lidar | |
| CN108090961B (en) | Rapid adjustment method in three-dimensional laser point cloud imaging |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |