CN110348473A - Non- cooperative Spacecraft autonomous classification method based on RANSAC - Google Patents
Non- cooperative Spacecraft autonomous classification method based on RANSAC Download PDFInfo
- Publication number
- CN110348473A CN110348473A CN201910448542.8A CN201910448542A CN110348473A CN 110348473 A CN110348473 A CN 110348473A CN 201910448542 A CN201910448542 A CN 201910448542A CN 110348473 A CN110348473 A CN 110348473A
- Authority
- CN
- China
- Prior art keywords
- point
- spacecraft
- points
- plane
- interior
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 230000011218 segmentation Effects 0.000 claims abstract description 8
- 239000013598 vector Substances 0.000 claims description 32
- 238000003032 molecular docking Methods 0.000 claims description 14
- 238000005070 sampling Methods 0.000 claims description 12
- 230000037237 body shape Effects 0.000 claims description 2
- 239000000203 mixture Substances 0.000 claims description 2
- 238000004088 simulation Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 244000291564 Allium cepa Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 238000009987 spinning Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The non-cooperative Spacecraft autonomous classification method based on RANSAC that the invention discloses a kind of, belong to non-cooperative Spacecraft identification and three-dimensional point cloud field, the spacecraft point cloud data of method of the invention based on laser radar scanning, according to fixed character existing for current in-orbit spacecraft, it is sequentially completed and randomly selects a cloud, for different component model of fit, segmentation best model, identify the process of spacecraft component.Recycle above-mentioned steps, each component until identifying non-cooperative Spacecraft completely.The method of the present invention does not need the prior informations such as the size that spacecraft is provided previously, and greatly strengthens versatility, the intelligence of non-cooperative Spacecraft autonomous classification.
Description
Technical field
The present invention relates to the identification of non-cooperative Spacecraft and three-dimensional point cloud field, specifically a kind of non-cooperations based on RANSAC
Spacecraft autonomous classification method.
Background technique
Spacecraft and space junk in-orbit at present is all noncooperative target: 1) being fitted without grabbing for mechanical arm capture
Hold mechanism (handle) and cooperation marker and characteristic block for subsidiary;2) target satellite motion state is unknown, Ke Nengwei
Rolling etc. under three-axis stabilization, spinning stability even runaway condition;3) without direct information between target satellite and tracking star
Exchange.Therefore, non-cooperative Spacecraft autonomous classification has become non-cooperative Spacecraft pose measurement, arrests docking, repairs and lengthen the life
The key technology of equal spatial operations.
RANSAC (stochastical sampling consistency algorithm) is the abbreviation of Random Sample Consensus, it is from one group
In sample data sets, the alternative manner of model parameter (models fitting) is estimated.It is in 1981 by Fischler and Bolles
It proposes, has been widely used in computer vision at first, such as solve of a pair of of camera simultaneously in stereoscopic vision field
Calculating with problem and fundamental matrix.
Traditional RACSAC algorithm, which is applied, can have following two points defect in non-cooperative Spacecraft autonomous classification: first is that model
Component identification is single, is blocked for the part of some spacecrafts, excalation, current goal are beyond special feelings such as field ranges
Condition is unable to satisfy the robustness demand of non-cooperative Spacecraft autonomous classification.Second is that precision problem, can only believe by spatial point coordinate
Breath carries out models fitting, does not distinguish iteratively sampled point cloud subset to model and describes surface equation, computationally intensive, time-consuming
Long, accuracy rate is low, is unable to satisfy the required precision of noncooperative target autonomous classification.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of general, intelligent non-cooperative Spacecraft autonomous classification methods.
The method can break through the limitation that the non-cooperative Spacecraft identification of tradition needs to be provided previously spacecraft shape and size, contain in data
It in the case where noise, still can accurately identify, spaceborne component of classifying.
Although may all be by identical the present invention is implemented as follows: then model is different for the component difference of spacecraft
What basic configuration was constituted, it is therefore desirable to the basic configuration based on fitting, further according to priori geometric knowledge, identification component, specific side
Steps are as follows for method:
Step 1: carrying laser radar by Simulation spatial service robot or tracking star, it is diversion to non-cooperative Spacecraft
Scanning obtains point cloud data;
Step 2:, according to the fixation shape feature of non-cooperative Spacecraft, being calculated using RANSAC based on obtained point cloud data
Method, stochastical sampling point set, for solar energy sailboard, spacecraft ontology, solar energy sailboard bracket, engine nozzle and satellite-rocket docking
Ring successively fits model of different shapes;After models fitting, model is divided, completes the identification of spacecraft component;
Step 3: successively identifying spaceborne component, spacecraft is completed after identification to be tested and is all identified;Institute
The inspection recognition methods stated are as follows:
1) subset is randomly selected in original point cloud data, and the point number that the subset chosen is included is by mould to be estimated
Type determines;It is assumed that subset only includes interior point, the model parameter to be fitted is estimated by all the points in subset, and by interior quantity
Set a preset threshold;
2) substitute point converge in left point, filter out the point for meeting the model parameter in left point, be denoted as interior point, count
In the model include all interior points;
If 3) the interior points that model of fit is included reach the preset threshold of the model, which is reasonable;In order to
More accurate achievees the effect that model of fit, according to all interior points model of fit again, determines model parameter;
4) if interior points contained by model of fit are not up to preset threshold, repeatedly step 1-3, by each model of fit institute
The interior points for including are as the standard of model accuracy is measured, until finding optimal model parameters as final result.
Further, in the step two, according to spacecraft solar array shape and spacecraft body shape, using RANSAC
Algorithm, three points of random acquisition, fit Plane;According to solar array bracket shape, using RANSAC algorithm, stochastical sampling two
Point and their corresponding normal vectors are fitted cylinder;It is adopted at random according to spacecraft engine nozzle shape using RANSAC algorithm
Three points of sample and their corresponding normal vectors are fitted circular cone;According to spacecraft satellite-rocket docking ring-shaped, using RANSAC algorithm,
Four points of stochastical sampling and their corresponding normal vectors are fitted annulus.
Further, the recognition methods of the solar energy sailboard are as follows:
1) from the acquisition point cloud data of acquisition, three points, fit Plane are randomly selected;
2) normal vector for calculating separately three points sets a differential seat angle threshold value, if between the normal vector of three points
Difference is both less than this threshold value, then the plane is reasonable enough;
3) remaining point in point cloud is calculated to set a square distance threshold value to the square distance of the plane, be less than threshold value
Point can be used as the interior point in plane, the parameter of plane is updated after statistics with all interior points;
4) it steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior plane put as optimal flat
Face has identified one piece accounted in the biggish solar array of spacecraft surface area;
5) it is partitioned into the plane of fitting, models fitting is carried out to remaining cloud again, fits another piece of optimal planar,
With the plane sizes such as substantially of first time fitting, that is, have identified another piece in solar array, the plane base of preceding twice fitting
This grade is big, and occupies 20% or more of total concurrent cloud number, that is, completes the identification of spacecraft solar array.
Further, the spacecraft ontology recognition methods are as follows: after completing solar energy sailboard identification, by solar energy sailboard
It splits;Based on left point, the step of applied solar energy windsurfing plane fitting again, after the segmentation for completing solar array,
It identifies the plane of the sizes such as six, and occupies 20% or more of total concurrent cloud number, that is, fitted six faces of cube, it will
The big facets such as six are successively split, and are completed the segmentation of cube, that is, are completed the identification of spacecraft ontology.
Further, the spacecraft ontology recognition methods are as follows:
1) after the identification for completing spacecraft ontology, spacecraft ontology is split, left point is based on, randomly selects two
A point m1And m2And the normal vector of the two pointsPass throughThe axial direction that can determine cylinder isThen
To cross this two o'clock andDirection axially projected to along cylinder with axially vertical same plane X, by the intersection point of linear projection
Bottom surface center of circle g as cylinder;
2) by g and m1Or m2Radius of the distance as cylinder between the subpoint on plane X calculates remaining in point cloud
Whether the angular deviation whether distance of point to cylindrical shaft is less than threshold epsilon and normal vector is less than threshold value μ, meets the note of condition
For point in cylindrical surface;After statistics, the parameter on cylindrical surface is updated with all interior points;
3) it steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior cylinder put as optimal column
Face;
4) due to solar array bracket at least there are two, so by cylinder all fitting segmentation finish, that is, complete the sun
The identification of windsurfing bracket.
Further, the recognition methods of the engine nozzle are as follows:
1) after the identification for completing solar array bracket, solar array bracket is split, is based on left point, it is random to take out
Take three point n1,n2,n3And their corresponding normal vectors;Each point normal vector corresponding with it can determine a plane,
Intersection point, that is, conical surface vertex of three planes;
2) by randomly select 3 points and conical surface vertex t, the other three point i.e. point can be determinedPointPointThe planar unit normal vector of this 3 points compositions is exactly the axial A of circular cone;
3) normal vector of these three points and the angle of circular cone axis are calculated again and takes its mean value, then half cone-apex angle ω of circular cone
Are as follows:
4) similarly, left point is successively substituted into the circular conical surface of fitting, judge whether to be interior point, counts all interior points, is updated
Circular conical surface parameter, steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior circular conical surface put as optimal
Circular conical surface is partitioned into circular conical surface, that is, completes the identification of engine nozzle.
Further, the recognition methods of the satellite-rocket docking ring are as follows:
1) after the identification for completing engine nozzle, engine nozzle is split;Based on left point, four are chosen
Point and the corresponding normal vector of four points, determine four straight lines, cross this four points respectively, and direction is the corresponding normal vector of point;
2) have one in this two straight lines it is found that extending vertically through at most only two straight lines of four straight lines by geometrical property
Item is the axis of annulus, respectively according to two annulus axis, determines different annulus, further according to four points chosen before other
Annulus;
3) in order to determine the internal diameter of annulus, these points are concentrated in the plane that one is pivoted, and pass through plane later
On three points can determine a circle, the distance of center of the circle to annulus axis is exactly outer diameter;
4) similarly, left point is successively substituted into the anchor ring of fitting, judge whether to be interior point, counts all interior points, is updated
Anchor ring parameter;It steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior anchor ring put as optimal
Anchor ring is partitioned into anchor ring, that is, completes the identification of satellite-rocket docking ring.
The beneficial effect of the present invention compared with prior art is:
Single problem is identified for model assembly, using stage know otherwise come identify it is spaceborne it is multiple not
Same component.Point cloud searching tree is established after having identified to current target component, the data reduction of identification component is divided it
Afterwards, to the target identification of next stage.After the identification to different components may be implemented in this method, while utilization is extracted every time
Segmentation can reduce the quantity of left point cloud, improve the efficiency and success rate of the identification of Remaining Stages point cloud component.It is low for precision
The problem of, the present invention is added to a cloud normal vector on the basis of traditional RANSAC algorithm, and the addition of the normal vector can be effective
Judge whether sampling subset is point in model, improve accuracy, reduces operand.
Detailed description of the invention
Fig. 1 is one of the non-cooperative Spacecraft of typical case being directed to used in the embodiment of the present invention;
Fig. 2 is general flow chart of the invention;
Fig. 3 is the flow chart based on RANSAC model of fit;
Fig. 4 is the effect picture that the present invention completes non-cooperative Spacecraft autonomous classification;
Fig. 5 is spacecraft component autonomous classification result of the present invention.
Specific embodiment
It is clear to keep the purpose of the present invention, technical solution and effect clearer, example is exemplified below to the present invention into one
Step is described in detail.It should be understood that specific implementation described herein is not intended to limit the present invention only to explain the present invention.
It is one of typical non-cooperative Spacecraft shown in as shown in Figure 1, current non-cooperative Spacecraft contains substantially
Fixed feature: spacecraft ontology, based on cylindrical body cuboid;Solar energy sailboard, based on rectangle;Solar energy sailboard bracket,
Based on cylindrical surface;Engine nozzle and satellite-rocket docking ring, based on circular cone, annulus.The different components of spacecraft have different shape, root
According to the difference of shape, the method for fitting will be distinguished, and process is main are as follows: according to spacecraft solar array shape and spacecraft sheet
Shape, using RANSAC algorithm, fit Plane.Three points of random acquisition, fit Plane.According to solar array bracket shape, answer
With RANSAC algorithm, it is fitted cylindrical surface.Two points of stochastical sampling and their corresponding normal vectors are fitted cylinder., according to space flight
Device engine nozzle shape is fitted circular conical surface using RANSAC algorithm.Three points of stochastical sampling and their corresponding normal direction
Amount is fitted circular cone.Anchor ring is fitted using RANSAC algorithm according to spacecraft satellite-rocket docking ring-shaped.Four points of stochastical sampling,
And their corresponding normal vectors, it is fitted annulus.
The flow chart of non-cooperative Spacecraft autonomous classification method based on RANSAC of the invention as shown in Fig. 2, step such as
Under:
Step 1, the acquisition of non-cooperative Spacecraft point cloud data;
Step 2, it identifies solar array, and divides;
Step 3, it is based on left point cloud, identifies spacecraft ontology, and divide;
Step 4, it is based on left point cloud, identifies solar array bracket, and divide;
Step 5, it is based on left point cloud, identifies engine nozzle, and divide;
Step 6, it is based on left point cloud, identifies satellite-rocket docking ring, and divide;
Step 7, judge whether success, "Yes" is then identified and finished, and "No" then switch mode re-recognizes, until being identified as
Function.Non- cooperative Spacecraft in-orbit at present has fixed feature, designs the different mode of several sets based on this, that is, can recognize absolutely mostly
Number spacecraft.Once being switched to corresponding mode, then reach best identified effect.
It is the flow chart based on RANSAC model of fit shown in Fig. 3, main flow is:
Step 1, according to the shape for the model to be fitted, s m point is randomly selected in N from converging to match point;
Step 2, according to the point model of fit of extraction;
Step 3, whether reasonable according to the modes such as normal angle judgment models;
Step 4, continue if rationally;Step 1 is gone back to if unreasonable, is extracted point set again, is fitted again;
Step 5, by remaining all the points, model is successively substituted into, according to the threshold value of setting, judges whether it is in model point (i.e.
Whether on model);
Step 6, number is put in statistics, and with all interior points, model of fit, keeps model more accurate again;
Step 7, it repeats above step T times, selects the model that point is most in contained, as optimal models, T meets:
In formula, η0It for confidence level, is usually arranged as in the range of [0.95,0.99], m is the number for the point randomly selected, δ
It is ratio of the interior point in all sample point set N.δ is clearly unknown under normal circumstances, therefore δ can take under worst case
The ratio of point, is arranged to the value of certain robusts.
It is the effect picture that the present invention completes non-cooperative Spacecraft autonomous classification shown in Fig. 4, all components are successively divided
It identifies, arrests the spatial operations such as docking, maintainable technology on-orbit, pose measurement for non-cooperative Spacecraft and provide technical support.
Component is carried out to multiple groups difference spacecraft the present invention is based on method of the invention and identifies autonomous classification, spacecraft portion
The partial results of part autonomous classification are shown in Fig. 5 (b) as shown in figure 5, be spacecraft main body recognition result shown in Fig. 5 (a)
Spacecraft nozzle recognition result.What box was chosen in 5 (a) is shown spacecraft main body, remaining is unidentified component.5(b)
What middle box was chosen is shown engine nozzle, remaining is unidentified component.It can be seen that the component identified all has
Certain integrality, and can be distinguished well with remaining unidentified component.In addition, the present invention is based on the knowledges of normal direction RANSAC
Other result is compared with tradition RANSAC, as a result as shown in table 1 below:
Table 1
RANSAC algorithm it can be seen from table in conjunction with normal vector is obviously good to the effect of spacecraft component autonomous classification
In traditional RANSAC algorithm.Engine nozzle more apparent for curved surface features and satellite-rocket docking ring, traditional RANSAC are calculated
Method can not almost identify, but the RANSAC algorithm of combination normal vector of the invention can still be known with higher discrimination
Not, while there is very strong robustness and robustness, average time-consuming differs also smaller with tradition RANSAC algorithm.
Claims (7)
1. a kind of non-cooperative Spacecraft autonomous classification method based on RANSAC, which is characterized in that the specific steps of the method
It is as follows:
Step 1: carrying laser radar by Simulation spatial service robot or tracking star, scanning of being diversion is carried out to non-cooperative Spacecraft
Obtain point cloud data;
Step 2: based on obtained point cloud data, according to the fixation shape feature of non-cooperative Spacecraft, using RANSAC algorithm,
Stochastical sampling point set, for solar energy sailboard, spacecraft ontology, solar energy sailboard bracket, engine nozzle and satellite-rocket docking ring
Successively fit model of different shapes;After models fitting, model is divided, completes the identification of spacecraft component;
Step 3: successively identifying spaceborne component, spacecraft is completed after identification to be tested and is all identified;Described
Examine recognition methods are as follows:
1) the point number that the subset for randomly selecting subset in original point cloud data, and choosing is included by model to be estimated Lai
It determines;It is assumed that subset only includes interior point, the model parameter to be fitted is estimated by all the points in subset, and interior quantity is set
One preset threshold;
2) substitute point converge in left point, filter out the point for meeting the model parameter in left point, be denoted as interior point, count the mould
In type include all interior points;
If 3) the interior points that model of fit is included reach the preset threshold of the model, which is reasonable;According to all
Interior point model of fit again, determines model parameter;
4) if interior points contained by model of fit are not up to preset threshold, repeatedly step 1-3, included by each model of fit
Interior points as the standard for measuring model accuracy, until find optimal model parameters as final result.
2. a kind of non-cooperative Spacecraft autonomous classification method based on RANSAC according to claim 1, which is characterized in that
In the step two, according to spacecraft solar array shape and spacecraft body shape, using RANSAC algorithm, random acquisition three
It is a, fit Plane;According to solar array bracket shape, using RANSAC algorithm, two points of stochastical sampling and they are corresponding
Normal vector is fitted cylinder;According to spacecraft engine nozzle shape, using RANSAC algorithm, three points of stochastical sampling and they
Corresponding normal vector is fitted circular cone;According to spacecraft satellite-rocket docking ring-shaped, using RANSAC algorithm, four points of stochastical sampling, with
And their corresponding normal vectors, it is fitted annulus.
3. a kind of non-cooperative Spacecraft autonomous classification method based on RANSAC according to claim 2, which is characterized in that
The recognition methods of the solar energy sailboard are as follows:
1) from the acquisition point cloud data of acquisition, three points, fit Plane are randomly selected;
2) normal vector for calculating separately three points sets a differential seat angle threshold value, if the difference between the normal vector of three points
Both less than this threshold value, then the plane is reasonable enough;
3) it calculates remaining point in point cloud and a square distance threshold value is set, less than the point of threshold value to the square distance of the plane
It can be used as the interior point in plane, update the parameter of plane after statistics with all interior points;
4) it steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior plane put as optimal planar, i.e.,
Have identified one piece accounted in the biggish solar array of spacecraft surface area;
5) it is partitioned into the plane of fitting, models fitting is carried out to remaining cloud again, fits another piece of optimal planar, with
The plane of the once fitting sizes such as substantially, that is, have identified another piece in solar array, and the plane of preceding twice fitting is substantially etc.
Greatly, and 20% or more of total concurrent cloud number, the i.e. identification of completion spacecraft solar array are occupied.
4. a kind of non-cooperative Spacecraft autonomous classification method based on RANSAC according to claim 2, which is characterized in that
The spacecraft ontology recognition methods are as follows: after completing solar energy sailboard identification, solar energy sailboard is split;Based on surplus
Yu Dian, the step of applied solar energy windsurfing plane fitting identifies the sizes such as six after the segmentation for completing solar array again
Plane, and occupy 20% or more of total concurrent cloud number, that is, fitted six faces of cube, by the big facets such as six according to
It is secondary to split, the segmentation of cube is completed, that is, completes the identification of spacecraft ontology.
5. a kind of non-cooperative Spacecraft autonomous classification method based on RANSAC according to claim 2, which is characterized in that
The spacecraft ontology recognition methods are as follows:
1) after the identification for completing spacecraft ontology, spacecraft ontology is split, left point is based on, randomly selects two points
m1And m2And the normal vector of the two pointsPass throughThe axial direction that can determine cylinder isThen it incited somebody to action
This two o'clock andDirection axially projected to along cylinder with axially vertical same plane X, using the intersection point of linear projection as
The bottom surface center of circle g of cylinder;
2) by g and m1Or m2Radius of the distance as cylinder between the subpoint on plane X calculates left point in point cloud and arrives
Whether the distance of cylindrical shaft is less than threshold epsilon and whether the angular deviation of normal vector is less than threshold value μ, and meet condition is denoted as circle
Point in cylinder;After statistics, the parameter on cylindrical surface is updated with all interior points;
3) it steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior cylinder put as optimal cylinder;
4) due to solar array bracket at least there are two, so by cylinder all fitting segmentation finish, that is, complete solar array
The identification of bracket.
6. a kind of non-cooperative Spacecraft autonomous classification method based on RANSAC according to claim 2, which is characterized in that
The recognition methods of the engine nozzle are as follows:
1) after the identification for completing solar array bracket, solar array bracket is split, left point is based on, randomly selects three
A point n1,n2,n3And their corresponding normal vectors;Each point normal vector corresponding with it can determine a plane, three
Intersection point, that is, conical surface vertex of plane;
2) by randomly select 3 points and conical surface vertex t, the other three point i.e. point can be determinedPointPointThe planar unit normal vector of this 3 points compositions is exactly the axial A of circular cone;
3) normal vector of these three points and the angle of circular cone axis are calculated again and takes its mean value, then half cone-apex angle ω of circular cone are as follows:
4) similarly, left point is successively substituted into the circular conical surface of fitting, judge whether to be interior point, counts all interior points, updates circular cone
Face parameter, steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior circular conical surface put as optimal circular cone
Face is partitioned into circular conical surface, that is, completes the identification of engine nozzle.
7. a kind of non-cooperative Spacecraft autonomous classification method based on RANSAC according to claim 2, which is characterized in that
The recognition methods of the satellite-rocket docking ring are as follows:
1) after the identification for completing engine nozzle, engine nozzle is split;Based on left point, four points are chosen, with
And the corresponding normal vector of four points, it determines four straight lines, crosses this four points respectively, direction is the corresponding normal vector of point;
2) one is in this two straight lines it is found that extending vertically through at most only two straight lines of four straight lines by geometrical property
The axis of annulus determines different annulus respectively according to two annulus axis, further according to other circles of four points chosen before
Ring;
3) in order to determine the internal diameter of annulus, these points are concentrated in the plane that one is pivoted, later by plane
Three points can determine a circle, and the distance of center of the circle to annulus axis is exactly outer diameter;
4) similarly, left point is successively substituted into the anchor ring of fitting, judge whether to be interior point, counts all interior points, updates annulus
Face parameter;It steps be repeated alternatively until and reach maximum number of iterations, select comprising the at most interior anchor ring put as optimal annulus
Face is partitioned into anchor ring, that is, completes the identification of satellite-rocket docking ring.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910448542.8A CN110348473A (en) | 2019-05-27 | 2019-05-27 | Non- cooperative Spacecraft autonomous classification method based on RANSAC |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910448542.8A CN110348473A (en) | 2019-05-27 | 2019-05-27 | Non- cooperative Spacecraft autonomous classification method based on RANSAC |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110348473A true CN110348473A (en) | 2019-10-18 |
Family
ID=68173980
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910448542.8A Pending CN110348473A (en) | 2019-05-27 | 2019-05-27 | Non- cooperative Spacecraft autonomous classification method based on RANSAC |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110348473A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111692997A (en) * | 2020-06-09 | 2020-09-22 | 西安交通大学 | Data-driven vector tail nozzle area in-situ measurement method |
CN111750870A (en) * | 2020-06-30 | 2020-10-09 | 南京理工大学 | Motion parameter estimation method for rocket body of space tumbling rocket |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106441311A (en) * | 2016-07-15 | 2017-02-22 | 上海宇航系统工程研究所 | Laser imaging radar-based non-cooperative aircraft relative position posture measuring method |
CN107655473A (en) * | 2017-09-20 | 2018-02-02 | 南京航空航天大学 | Spacecraft based on SLAM technologies is with respect to autonomous navigation system |
CN108917772A (en) * | 2018-04-04 | 2018-11-30 | 北京空间飞行器总体设计部 | Noncooperative target Relative Navigation method for estimating based on sequence image |
-
2019
- 2019-05-27 CN CN201910448542.8A patent/CN110348473A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106441311A (en) * | 2016-07-15 | 2017-02-22 | 上海宇航系统工程研究所 | Laser imaging radar-based non-cooperative aircraft relative position posture measuring method |
CN107655473A (en) * | 2017-09-20 | 2018-02-02 | 南京航空航天大学 | Spacecraft based on SLAM technologies is with respect to autonomous navigation system |
CN108917772A (en) * | 2018-04-04 | 2018-11-30 | 北京空间飞行器总体设计部 | Noncooperative target Relative Navigation method for estimating based on sequence image |
Non-Patent Citations (1)
Title |
---|
ZHIMING CHEN等: "A new pose estimation method for non-cooperative spacecraft based on point cloud", 《INTERNATIONAL JOURNAL OF INTELLIGENT COMPUTING AND CYBERNETICS》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111692997A (en) * | 2020-06-09 | 2020-09-22 | 西安交通大学 | Data-driven vector tail nozzle area in-situ measurement method |
CN111692997B (en) * | 2020-06-09 | 2021-08-13 | 西安交通大学 | Data-driven vector tail nozzle area in-situ measurement method |
CN111750870A (en) * | 2020-06-30 | 2020-10-09 | 南京理工大学 | Motion parameter estimation method for rocket body of space tumbling rocket |
CN111750870B (en) * | 2020-06-30 | 2023-12-26 | 南京理工大学 | Motion parameter estimation method for space rolling rocket body |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110473239A (en) | A kind of high-precision point cloud registration method of 3 D laser scanning | |
CN111553409B (en) | Point cloud identification method based on voxel shape descriptor | |
CN113436260A (en) | Mobile robot pose estimation method and system based on multi-sensor tight coupling | |
CN106373118A (en) | A complex curved surface part point cloud reduction method capable of effectively keeping boundary and local features | |
CN114972377B (en) | 3D point cloud segmentation method and device based on mobile least square method and super-voxel | |
CN106643492B (en) | A kind of aero-engine damaged blade 3-dimensional digital speckle formative method | |
JP2009093611A (en) | System and method for recognizing three-dimensional object | |
CN101839722A (en) | Method for automatically recognizing target at medium and low altitudes and positioning carrier with high accuracy | |
CN105654483A (en) | Three-dimensional point cloud full-automatic registration method | |
CN110807781A (en) | Point cloud simplification method capable of retaining details and boundary features | |
CN107680168A (en) | Lattice simplified method based on plane fitting in three-dimensional reconstruction | |
CN110348473A (en) | Non- cooperative Spacecraft autonomous classification method based on RANSAC | |
CN106570905A (en) | Initial attitude verification method of non-cooperative target point cloud | |
CN111932669A (en) | Deformation monitoring method based on slope rock mass characteristic object | |
CN110097598A (en) | A kind of three-dimension object position and orientation estimation method based on PVFH feature | |
CN114166211B (en) | Double-view-field star sensor star map identification method | |
CN109345571B (en) | Point cloud registration method based on extended Gaussian image | |
Kaushik et al. | Accelerated patch-based planar clustering of noisy range images in indoor environments for robot mapping | |
CN110288620A (en) | Image matching method and aircraft navigation method based on line segment geometrical characteristic | |
CN109934859A (en) | It is a kind of to retrace the ICP method for registering for stating son based on feature enhancing multi-dimension Weight | |
CN115760984B (en) | Non-cooperative target pose measurement method based on monocular vision for cube star | |
US20130332110A1 (en) | Non-iterative mapping of capped cylindrical environments | |
CN107742295A (en) | A kind of cube star docking reconstructing method of view-based access control model | |
Cao et al. | An end-to-end pose estimation network for multiscale space non-cooperative objects | |
CN108595373A (en) | It is a kind of without control DEM method for registering |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191018 |