CN102968400A - Multi-view three-dimensional data registration method based on spatial line recognition and matching - Google Patents
Multi-view three-dimensional data registration method based on spatial line recognition and matching Download PDFInfo
- Publication number
- CN102968400A CN102968400A CN2012103977527A CN201210397752A CN102968400A CN 102968400 A CN102968400 A CN 102968400A CN 2012103977527 A CN2012103977527 A CN 2012103977527A CN 201210397752 A CN201210397752 A CN 201210397752A CN 102968400 A CN102968400 A CN 102968400A
- Authority
- CN
- China
- Prior art keywords
- line
- straight line
- point
- cloud
- matrix
- 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.)
- Granted
Links
Abstract
The invention discloses a multi-view three-dimensional data registration method based on spatial line recognition and matching. The multi-view three-dimensional data registration method based on the spatial line recognition and matching is suitable for multi-view registration on data with obvious edge features. The multi-view three-dimensional data registration method based on the spatial line recognition and matching comprises the following main steps of: extracting line segments from three-dimensional point cloud data; realizing the matching of common line segments under different views according to the inherent constrained relationship of positions and angles of the extracted line segments; finding rotation invariant corresponding points under different views on the basis of the line segment matching; and obtaining an optimal global attitude transformation matrix based on the corresponding points to realize multi-view three-dimensional data registration.
Description
Technical field
The present invention relates to a kind of based on the multi-view angle three-dimensional data joining method of space line identification with coupling, it is a kind of method of space three-dimensional Point Cloud Processing, be specifically related to a kind of joining method of multi-view angle three-dimensional data, belong to three-dimensional measurement and technical field of computer vision.
Background technology
In three-dimensional measurement, often need to obtain the whole three-dimensional appearance model of workpiece, because the restriction that single visual angle is measured, a plurality of visual angles measurement result need to be stitched together just to obtain whole topographic data, so the selection of joining method seems relatively more crucial.
For the splicing that realizes that various visual angles are put cloud, method commonly used is by at measuring object surface sticking sign point, by to the transformation matrix between the definite different visual angles of sign recognition and location, realizes the coupling of multi-view angle three-dimensional data by transformation matrix.In addition, having does not need only to rely on cloud data to realize the method for splicing by monumented point yet, the splicing of the three dimensional point cloud of different visual angles realizes that by the Iterative matching adjustment in the coincidence zone in the visual angle method commonly used is ICP(iterative closet point) algorithm and improvement algorithm.
Can provide higher splicing precision but because will be at surperficial sticking sign point based on the method for model surface sticking sign point, can not contact during loaded down with trivial details and some model measurement of measuring process, limit the application of method.And the search corresponding point that need in the process of implementation iteration based on the joining method of cloud data to adjust relative attitude, need to consume the plenty of time, thereby and local extremum in calculating process, occurs easily converging to and can't realize aiming at.
Problem for the model three-dimensional data aligning that obvious border is arranged, seamed edge information can the primary expression model contour feature, the linear feature that seamed edge is corresponding can provide for the aligning of three-dimensional data stronger constraint, and because the relatively a large amount of cloud datas of straight line parameter greatly reduce in data volume, therefore bring very big facility can for the splicing with obvious border three-dimensional model based on the three dimensional point cloud joining method of linear feature.
Summary of the invention
The present invention proposes a kind ofly based on the space line identification multi-view angle three-dimensional data joining method with coupling, the three dimensional point cloud that is applicable to have obvious seamed edge feature splices from various visual angles.It splices by the line characteristic matching relation of utilizing the different visual angles point cloud model, compares with existing Point-clouds Registration method, and minimizing operand that can be a large amount of also can effectively prevent faulty convergence.
Technical scheme: for the multi-view angle three-dimensional point cloud of realizing obvious border splices fast, the present invention proposes a kind of based on the multi-view angle three-dimensional data joining method of space line identification with coupling., the three dimensional point cloud that is applicable to have obvious seamed edge feature splices from various visual angles.It is at first according to the straight line parameter in the extraction of straight line method extraction three dimensional point cloud, then can realize the coupling identification of common linear under the different visual angles according to the restriction relation of the position each other of extracting straight line and angle etc., then find the corresponding point of invariable rotary under the different visual angles on the basis of matching line, ask for the splicing that realizes three-dimensional data thereby carry out overall attitude optimum attitude based on corresponding point at last.
The present invention is a kind of based on the multi-view angle three-dimensional data joining method of space line identification with coupling, and the method concrete steps are:
Step 1: extract the linear feature parameter from original point cloud data: at first the amount of curvature according to cloud data finds the some cloud that belongs to the line feature, and postulated point converges and closes available P{ (x
1, y
1, z
1) ... (x
n, y
n, z
n) expression,
Be the average of coordinate, can make up matrix by average
The covariance matrix of P then:
X=[P-N]
T·[P-N]
According to the characteristic of eigenwert and proper vector as can be known:
X α=λ α establishes λ
1, λ
2, λ
3Be the eigenwert of X, and magnitude relationship there is λ
1＜λ
2＜λ
3, then the estimated value of curvature is:
Set up threshold value threth according to the size distribution of curvature, then the curvature valuation is considered to the line feature point cloud greater than threth.
On the basis of the line feature point cloud that extracts, the straight line parameter that the line feature point cloud comprises is asked for.To eigenwert and the proper vector analysis of line feature point cloud, α
1, α
2, α
3Be λ
1, λ
2, λ
3Characteristic of correspondence vector, then α
3Putting exactly the rectilinear direction of cloud estimates.The rectilinear direction of all some clouds is estimated that carrying out normalization obtains
Vector after the normalization is carried out the three-dimensional mapping in the Gaussian vectors ball, the direction of straight line can be thought in the center of the some cloud cluster in the Gaussian vectors ball that obtains, obtain the center of each cluster by cluster analysis, the institute that has also just obtained comprising in the some cloud linear feature might direction.
After obtaining the direction of 3 d-line, find again the locus of 3 d-line just can realize determining 3 d-line.To the plane, can obtain to converge at the line feature point cloud cluster of any along the direction projection of straight line, owing to may there be parallel space three-dimensional straight line, therefore converge to have on the plane and surpass one convergent point.Use clustering method can obtain the center of each cluster, thereby realize obtaining of straight line parameter.
Use the Household matrix to realize the line feature point cloud is carried out projection along self direction on the plane.If the direction availability vector of straight line is expressed as
Order
Then transition matrix is R=I-2bb
T, converge P for line feature point, through the some cloud P'=RP after the projection, if the artificial z axial coordinate with P' is made as zero, then can obtain in the plane projection result.Carry out cluster analysis and obtain cluster centre in planar range, then the projected position of straight line can obtain, thereby has realized determining the straight line parameter.
Step 2: the coupling that realizes the different visual angles associated straight lines: supposing has straight line l under certain visual angle
1: (x
1, y
1, z
1) cross a bit (a
1, b
1, c
1), straight line l
2: (x
2, y
2, z
2) cross a bit (a
2, b
2, c
2),
If: k
1=a
1(x
2-x
1)+b
1(y
2-y
1)+c
1(z
2-z
1)
k
2=a
2(x
2-x
1)+b
2(y
2-y
1)+c
2(z
2-z
1)
M=a
1a
2+b
1b
2+c
1c
2
N=a
1 2+b
1 2+c
1 2
P=a
2 2+bb
2 2+c
2 2
t=(k
2·N-k
1·M)/(M
2-P·N)
s=(k
2·M-k
1·P)/(M
2-P·N)
K1, k2, M, N, P, t, s are by straight line l in the formula
1: (x
1, y
1, z
1), straight line l
2: (x
2, y
2, z
2), point (a
1, b
1, c
1), point (a
2, b
2, c
2) the equivalent representation form of the algebraic expression that forms of data.
The distance of two space lines then
dist=|x
2-x
1+s·(a
2-a
1),y
2-y
1+s·(b
2-b
1),z
2-z
1+s·(c
2-c
1)|
Article two, the angle of space line:
Article two, the distance of space line and angle remain unchanged in the spatial alternation process, make up the description vectors of each bar straight line according to the formula of vector sum angle:
v
line=[dist
1…dist
n;angle
l…angle
n]
Every straight line is to there being different description vectors, have no relations owing to angle with apart from the isospace rotation, the description vectors of the associated straight lines under the different visual angles is consistent or is approximate, mates according to the similarity of vector, can realize the coupling of different visual angles line correspondence.
Step 3: find the spatial alternation corresponding point according to the straight line matching relationship: have 2 points on two of different surface beeline straight lines each other, their line is all vertical with two different surface beelines, distance between them is the different surface beeline distance, can know that by vertical relation have uniqueness at 2, point such under different visual angles is corresponding point.
Supposing has straight line l under certain visual angle
1: (x
1, y
1, z
1) cross a bit (a
1, b
1, c
1), straight line l
2: (x
2, y
2, z
2) cross a bit (a
2, b
2, c
2),
The same with step 3, if having: k
1=a
1(x
2-x
1)+b
1(y
2-y
1)+c
1(z
2-z
1)
k
2=a
2(x
2-x
1)+b
2(y
2-y
1)+c
2(z
2-z
1)
M=a
1a
2+b
1b
2+c
1c
2
N=a
1 2+b
1 2+c
1 2
P=a
2 2+b
2 2+c
2 2
t=(k
2·N-k
1·M)/(M
2-P·N)
s=(k
2·M-k
1·P)/(M
2-P·N)
Corresponding point then:
p
1=[x
1+s·a
1,y
1+s·b
1,z
1+s·c
1]
p
2=[x
2+s·a
2,y
2+s·b
2,z
2+s·c
2]
p
1, p
2Be expressed as two intersection points between the different surface beeline.
Can find reference point p on the straight line according to the matching relationship of straight line
1p
2, and the corresponding point q under their different visual angles
1, q
2In like manner obtain whole straight line corresponding point pair:
P:{p
1,p
2,…p
n},Q:{q
1，q
2,…q
n}。So just the matching line splicing is transformed into the right splicing of corresponding point.
Step 4: calculate transition matrix between the different visual angles based on corresponding point.Suppose:
P:{p
1, p
2... p
n, Q:{q
1, q
2... q
nBe corresponding point to collection, obtain optimum spatial alternation matrix R, T, satisfy:
The method of using SVD to decompose is calculated the equation optimum solution:
Order
Matrix H=P ' then
TQ' carries out svd [U, D, V]=SVD (H), and then final spatial alternation matrix is:
R=V·U
T,T=q-R·p
Symbol description is as follows in the following formula: R is rotation matrix, and T is translation matrix, and p is the arbitrfary point among the point set P, and q is the arbitrfary point among the point set Q, and H is the matrix that point set P, Q computing form, and U, D, V are the corresponding matrixes that the H svd is obtained.
Use R, T aims at original three dimensional point cloud as transformation relation, then can access the splicing result of cloud data under the different visual angles.
Beneficial effect: in the splicing of the three dimensional point cloud with obvious border, utilize based on the method for linear feature coupling and carry out the splicing of many visual fields cloud data, can improve splicing efficient.The present invention has provided a kind of based on the multi-view angle three-dimensional data joining method of space line identification with coupling, and its advantage is:
1, carries out the splicing of many visual field point clouds by extracting the linear feature structural parameters, can reduce in a large number operand, shorten computing time.
2, by splicing based on linear feature coupling, effectively avoided the local convergence problem that runs into based in the cloud process, ensured the true(-)running of algorithm.
3, the linear feature identification that relates to of splicing, the research of different surface beeline relative space position relation can be used for other related fields of Point Cloud Processing.
Description of drawings:
Fig. 1 is the process flow diagram of feature extracting method of the present invention
Embodiment: following according to the concrete embodiment of method step introduction shown in Figure 1.
The present invention is a kind of based on the multi-view angle three-dimensional data joining method of space line identification with coupling, and the method concrete steps are:
Step 1: extract the linear feature parameter from original point cloud data: at first the amount of curvature according to cloud data finds the some cloud that belongs to the line feature, and postulated point converges and closes available P{ (x
1, y
1, z
1) ... (x
n, y
n, z
n) expression,
Be the average of coordinate, can make up matrix by average
The covariance matrix of P then:
X=[P-N]
T·[P-N]
Characteristic according to eigenwert and proper vector has as can be known:
X α=λ α establishes λ
1, λ
2, λ
3Be the eigenwert of X, and magnitude relationship there is λ
1＜λ
2＜λ
3, then the estimated value of curvature is:
Set up threshold value threth according to the size distribution of curvature, then the curvature valuation is considered to the line feature point cloud greater than threth.
On the basis of the line feature point cloud that extracts, the straight line parameter that the line feature point cloud comprises is asked for.To eigenwert and the proper vector analysis of line feature point cloud, α
1, α
2, α
3Be λ
1, λ
2, λ
3Characteristic of correspondence vector, then α
3Putting exactly the rectilinear direction of cloud estimates.The rectilinear direction of all some clouds is estimated that carrying out normalization obtains
Vector after the normalization is carried out the three-dimensional mapping in Gaussian sphere, the direction of straight line is thought at the center of the some cloud cluster in the Gaussian vectors ball that obtains, obtain the center of each cluster by cluster analysis, the institute that has also just obtained comprising in the some cloud linear feature might direction.
After obtaining the direction of 3 d-line, find the locus of 3 d-line just can realize determining 3 d-line.To the plane, can obtain to converge at the line feature point cloud cluster of any along the direction projection of straight line, owing to may there be parallel space three-dimensional straight line, therefore converge to have on the plane and surpass one convergent point.Use clustering method can obtain the center of each cluster, thereby realize obtaining of straight line parameter.
Use the Household matrix to realize the line feature point cloud is carried out projection along self direction on the plane.If the direction availability vector of straight line is expressed as
Order
Then transition matrix is R=I-2bb
T, converge P for line feature point, through the some cloud P'=RP after the projection, if the artificial z axial coordinate with P' is made as zero, then can obtain in the plane projection result.Carry out cluster analysis and obtain cluster centre in planar range, then the projected position of straight line can obtain, thereby has realized determining the straight line parameter.
Step 2: the coupling that realizes the different visual angles associated straight lines: supposing has straight line l under certain visual angle
1: (x
1, y
1, z
1) cross a bit (a
1, b
1, c
1), straight line l
2: (x
2, y
2, z
2) cross a bit (a
2, b
2, c
2),
If: k
1=a
1(x
2-x
1)+b
1(y
2-y
1)+c
1(z
2-z
1)
k
2=a
2(x
2-x
1)+b
2(y
2-y
1)+c
2(z
2-z
1)
M=a
1a
2+b
1b
2+c
1c
2
N=a
1 2+b
1 2+c
1 2
P=a
2 2+b
2 2+c
2 2
t=(k
2·N-k
1·M)/(M
2-P·N)
s=(k
2·M-k
1·P)/(M
2-P·N)
K1, k2, M, N, P, t, s are by straight line l in the formula
1: (x
1, y
1, z
1), straight line l
2: (x
2, y
2, z
2), point (a
1, b
1, c
1), point (a
2, b
2, c
2) the equivalent representation form of the algebraic expression that forms of data.
The distance of two space lines then
dist=|x
2-x
1+s·(a
2-a
1),y
2-y
1+s·(b
2-b
1),z
2-z
1+s·(c
2-c
1)|
Article two, the angle of space line:
Utilize distance and the angle of two space lines in the spatial alternation process, to remain unchanged, make up the description vectors of each bar straight line according to the formula of vector sum angle:
v
line=[dist
1…dist
n;angle
l…angle
n]
Every straight line is to there being different description vectors, have no relations owing to angle with apart from the isospace rotation, the description vectors of the associated straight lines under the different visual angles is consistent or is approximate, mates according to the similarity of vector, coupling that can straight line different visual angles line correspondence.
Step 3: find the spatial alternation corresponding point according to the straight line matching relationship: have 2 points on two of different surface beeline straight lines each other, their line is all vertical with two different surface beelines, distance between them is the different surface beeline distance, can know that by vertical relation have uniqueness at 2, therefore such point is corresponding point under different visual angles.
Supposing has straight line l under certain visual angle
1: (x
1, y
1, z
1) cross a bit (a
1, b
1, c
1), straight line l
2: (x
2, y
2, z
2) cross a bit (a
2, b
2, c
2),
The same with step 3, if having: k
1=a
1(x
2-x
1)+b
1(y
2-y
1)+c
1(z
2-z
1)
k
2=a
2(x
2-x
1)+b
2(y
2-y
1)+c
2(z
2-z
1)
M=a
1a
2+b
1b
2+c
1c
2
N=a
1 2+b
1 2+c
1 2
P=a
2 2+b
2 2+c
2 2
t=(k
2·N-k
1·M)/(M
2-P·N)
s=(k
2·M-k
1·P)/(M
2-P·N)
Corresponding point then:
p
1=[x
1+s·a
1,y
1+s·b
1,z
1+s·c
1]
p
2=[x
2+s·a
2,y
2+s·b
2,z
2+s·c
2]
p
1, p
2Be expressed as two intersection points between the different surface beeline.
Find reference point p on the straight line according to the matching relationship of straight line
1p
2, and the corresponding point q under their different visual angles
1, q
2In like manner obtain whole straight line corresponding point pair:
P:{p
1,p
2,…p
n},Q:{q
1，q
2,…q
n}。So just the matching line splicing is transformed into the right splicing of corresponding point.
Step 4: calculate transition matrix between the different visual angles based on corresponding point.Suppose:
P:{p
1, p
2... p
n, Q:{q
1, q
2... q
nBe corresponding point to collection, obtain optimum spatial alternation matrix R, T, satisfy:
The method of using SVD to decompose is calculated the equation optimum solution:
Order
Matrix H=P ' then
TQ' carries out svd [U, D, V]=SVD (H), and then final spatial alternation matrix is:
R=V·U
T,T=q-R·p
Symbol description is as follows in the following formula: R is rotation matrix, and T is translation matrix, and p is the arbitrfary point among the point set P, and q is the arbitrfary point among the point set Q, and H is the matrix that point set P, Q computing form, and U, D, V are the corresponding matrixes that the H svd is obtained.
Use R, T carries out original three dimensional point cloud can obtain criterion the splicing result of cloud data under the different visual angles as transformation relation.
Claims (1)
1. one kind based on the space line identification multi-view angle three-dimensional data joining method with coupling, and it is characterized in that: the method concrete steps are as follows:
Step 1: extract the linear feature parameter from original point cloud data: at first the amount of curvature according to cloud data finds the some cloud that belongs to the line feature, and postulated point converges and share P{ (x
1, y
1, z
1) ... (x
n, y
n, z
n) expression,
Be the average of coordinate, make up matrix by average
The covariance matrix of P then:
X=[P-N]
T·[P-N]
According to the characteristic of eigenwert and proper vector,
X α=λ α establishes λ
1, λ
2, λ
3Be the eigenwert of X, and magnitude relationship there is λ
1＜λ
2＜λ
3, then the estimated value of curvature is:
Set up threshold value threth according to the size distribution of curvature, then the curvature valuation is considered to the line feature point cloud greater than threth;
On the basis of the line feature point cloud that extracts, the straight line parameter that the line feature point cloud comprises is asked for, to eigenwert and the proper vector analysis of line feature point cloud, α
1, α
2, α
3Be λ
1, λ
2, λ
3Characteristic of correspondence vector, then α
3Putting exactly the rectilinear direction of cloud estimates; The rectilinear direction of all some clouds is estimated that carrying out normalization obtains
Vector after the normalization is carried out the three-dimensional mapping in the Gaussian vectors ball, the direction of straight line is thought at the center of the some cloud cluster in the Gaussian vectors ball that obtains, obtain the center of each cluster by cluster analysis, the institute that has also just obtained comprising in the some cloud linear feature might direction;
After obtaining the direction of 3 d-line, find again the locus of 3 d-line just to realize determining 3 d-line; To the plane, obtain to converge at the line feature point cloud cluster of any along the direction projection of straight line, owing to may there be parallel space three-dimensional straight line, therefore converge to have on the plane and surpass one convergent point; Use clustering method to obtain the center of each cluster, thereby realize obtaining of straight line parameter;
Use the Household matrix to realize the line feature point cloud is carried out projection along self direction on the plane, if the direction vector representation of straight line is
Order
Then transition matrix is R=I-2bb
T, converge P for line feature point, through the some cloud P'=RP after the projection, if the artificial z axial coordinate with P' is made as zero, then obtain in the plane projection result; Carry out cluster analysis and obtain cluster centre in planar range, then the projected position of straight line obtains, thereby has realized determining the straight line parameter;
Step 2: the coupling that realizes the different visual angles associated straight lines: supposing has straight line l under certain visual angle
1: (x
1, y
1, z
1) cross a bit (a
1, b
1, c
1), straight line l
2: (x
2, y
2, z
2) cross a bit (a
2, b
2, c
2),
If: k
1=a
1(x
2-x
1)+b
1(y
2-y
1)+c
1(z
2-z
1)
k
2=a
2(x
2-x
1)+b
2(y
2-y
1)+c
2(z
2-z
1)
M=a
1a
2+b
1b
2+c
1c
2
N=a
1 2+b
1 2+c
1 2
P=a
2 2+b
2 2+c
2 2
t=(k
2·N-k
1·M)/(M
2-P·N)
s=(k
2·M-k
1·P)/(M
2-P·N)
K1, k2, M, N, P, t, s are by straight line l in the formula
1: (x
1, y
1, z
1), straight line l
2: (x
2, y
2, z
2), point (a
1, b
1, c
1), point (a
2, b
2, c
2) the equivalent representation form of the algebraic expression that forms of data;
The distance of two space lines then
dist=|x
2-x
1+s·(a
2-a
1),y
2-y
1+s·(b
2-b
1),z
2-z
1+s·(c
2-c
1)|
Article two, the angle of space line:
Article two, the distance of space line and angle remain unchanged in the spatial alternation process, make up the description vectors of each bar straight line according to the formula of vector sum angle:
v
line=[dist
1…dist
n;angle
l…angle
n]
Every straight line has no relations owing to angle with apart from the isospace rotation to different description vectors should be arranged, and the description vectors of the associated straight lines under the different visual angles is consistent or is approximate, mates according to the similarity of vector, realizes the coupling of different visual angles line correspondence;
Step 3: find the spatial alternation corresponding point according to the straight line matching relationship: have 2 points on two of different surface beeline straight lines each other, their line is all vertical with two different surface beelines, distance between them is the different surface beeline distance, can know that by vertical relation have uniqueness at 2, point such under different visual angles is corresponding point
Supposing has straight line l under certain visual angle
1: (x
1, y
1, z
1) cross a bit (a
1, b
1, c
1), straight line l
2: (x
2, y
2, z
2) cross a bit (a
2, b
2, c
2),
The same with step 3, if having: k
1=a
1(x
2-x
1)+b
1(y
2-y
1)+c
1(z
2-z
1)
k
2=a
2(x
2-x
1)+b
2(y
2-y
1)+c
2(z
2-z
1)
M=a
1a
2+b
1b
2+c
1c
2
N=a
1 2+b
1 2+c
1 2
P=a
2 2+b
2 2+c
2 2
t=(k
2·N-k
1·M)/(M
2-P·N)
s=(k
2·M-k
1·P)/(M
2-P·N)
Corresponding point then:
p
1=[x
1+s·a
1,y
1+s·b
1,z
1+s·c
1]
p
2=[x
2+s·a
2,y
2+s·b
2,z
2+s·c
2]
p
1, p
2Be expressed as two intersection points between the different surface beeline;
Can find reference point p on the straight line according to the matching relationship of straight line
1p
2, and the corresponding point q under their different visual angles
1, q
2, in like manner obtain whole straight line corresponding point pair:
P:{p
1, p
2... p
n, Q:{q
1, q
2... q
n, so just the matching line splicing is transformed into the right splicing of corresponding point,
Step 4: calculate transition matrix between the different visual angles based on corresponding point, suppose:
P:{p
1, p
2... p
n, Q:{q
1, q
2... q
nBe corresponding point to collection, obtain optimum spatial alternation matrix R, T, satisfy:
The method of using SVD to decompose is calculated the equation optimum solution:
Order
Matrix H=P ' then
TQ' carries out svd [U, D, V]=SVD (H), and then final spatial alternation matrix is:
R=V·U
T,T=q-R·p
Symbol description is as follows in the following formula: R is rotation matrix, and T is translation matrix, and p is the arbitrfary point among the point set P, and q is the arbitrfary point among the point set Q, and H is the matrix that point set P, Q computing form, and U, D, V are the corresponding matrixes that the H svd is obtained;
Use R, T aims at original three dimensional point cloud as transformation relation, then can access the splicing result of cloud data under the different visual angles.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210397752.7A CN102968400B (en) | 2012-10-18 | 2012-10-18 | A kind of based on space line identification and the multi-view three-dimensional data registration method of mating |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210397752.7A CN102968400B (en) | 2012-10-18 | 2012-10-18 | A kind of based on space line identification and the multi-view three-dimensional data registration method of mating |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102968400A true CN102968400A (en) | 2013-03-13 |
| CN102968400B CN102968400B (en) | 2016-03-30 |
Family
ID=47798548
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210397752.7A Active CN102968400B (en) | 2012-10-18 | 2012-10-18 | A kind of based on space line identification and the multi-view three-dimensional data registration method of mating |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102968400B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103955964A (en) * | 2013-10-17 | 2014-07-30 | 北京拓维思科技有限公司 | Ground laser point cloud splicing method based three pairs of non-parallel point cloud segmentation slices |
| CN104484648A (en) * | 2014-11-27 | 2015-04-01 | 浙江工业大学 | Variable-viewing angle obstacle detection method for robot based on outline recognition |
| CN104574273A (en) * | 2013-10-14 | 2015-04-29 | 鸿富锦精密工业（深圳）有限公司 | Point cloud registration system and method |
| CN106228603A (en) * | 2016-07-25 | 2016-12-14 | 武汉中观自动化科技有限公司 | A kind of three-dimensional model reconfiguration system and method based on Euclidean distance statistics splicing |
| CN109087244A (en) * | 2018-07-26 | 2018-12-25 | 贵州火星探索科技有限公司 | A kind of Panorama Mosaic method, intelligent terminal and storage medium |
| CN113534095A (en) * | 2021-06-18 | 2021-10-22 | 北京电子工程总体研究所 | Laser radar map construction method and robot autonomous navigation method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0594337A2 (en) * | 1992-10-22 | 1994-04-27 | International Business Machines Corporation | A data processing system |
| CN1928921A (en) * | 2006-09-22 | 2007-03-14 | 东南大学 | Automatic searching method for characteristic points cloud band in three-dimensional scanning system |
| CN1996387A (en) * | 2006-08-14 | 2007-07-11 | 东南大学 | Mark point matching method for point-cloud registration in 3D scanning system |
-
2012
- 2012-10-18 CN CN201210397752.7A patent/CN102968400B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0594337A2 (en) * | 1992-10-22 | 1994-04-27 | International Business Machines Corporation | A data processing system |
| CN1996387A (en) * | 2006-08-14 | 2007-07-11 | 东南大学 | Mark point matching method for point-cloud registration in 3D scanning system |
| CN1928921A (en) * | 2006-09-22 | 2007-03-14 | 东南大学 | Automatic searching method for characteristic points cloud band in three-dimensional scanning system |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104574273A (en) * | 2013-10-14 | 2015-04-29 | 鸿富锦精密工业（深圳）有限公司 | Point cloud registration system and method |
| CN103955964A (en) * | 2013-10-17 | 2014-07-30 | 北京拓维思科技有限公司 | Ground laser point cloud splicing method based three pairs of non-parallel point cloud segmentation slices |
| CN104484648A (en) * | 2014-11-27 | 2015-04-01 | 浙江工业大学 | Variable-viewing angle obstacle detection method for robot based on outline recognition |
| CN104484648B (en) * | 2014-11-27 | 2017-07-25 | 浙江工业大学 | Robot variable visual angle obstacle detection method based on outline identification |
| CN106228603A (en) * | 2016-07-25 | 2016-12-14 | 武汉中观自动化科技有限公司 | A kind of three-dimensional model reconfiguration system and method based on Euclidean distance statistics splicing |
| CN106228603B (en) * | 2016-07-25 | 2018-11-02 | 武汉中观自动化科技有限公司 | A kind of three-dimensional model reconfiguration system and method based on Euclidean distance statistics splicing |
| CN109087244A (en) * | 2018-07-26 | 2018-12-25 | 贵州火星探索科技有限公司 | A kind of Panorama Mosaic method, intelligent terminal and storage medium |
| CN109087244B (en) * | 2018-07-26 | 2023-04-18 | 深圳禾苗通信科技有限公司 | Panoramic image splicing method, intelligent terminal and storage medium |
| CN113534095A (en) * | 2021-06-18 | 2021-10-22 | 北京电子工程总体研究所 | Laser radar map construction method and robot autonomous navigation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102968400B (en) | 2016-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhu et al. | Single image 3D object detection and pose estimation for grasping | |
| CN106296693B (en) | Based on 3D point cloud FPFH feature real-time three-dimensional space-location method | |
| CN101645170B (en) | Precise registration method of multilook point cloud | |
| CN102968400A (en) | Multi-view three-dimensional data registration method based on spatial line recognition and matching | |
| CN101692257B (en) | Method for registering complex curved surface | |
| CN104200212B (en) | A kind of building external boundary line drawing method based on airborne LiDAR data | |
| CN109493372B (en) | Rapid global optimization registration method for product point cloud data with large data volume and few characteristics | |
| CN102799763B (en) | A kind of based on a cloud attitude standardized some cloud line feature extraction method | |
| Glasner et al. | aware object detection and continuous pose estimation | |
| CN110211129B (en) | Low-coverage point cloud registration algorithm based on region segmentation | |
| CN108022262A (en) | A kind of point cloud registration method based on neighborhood of a point center of gravity vector characteristics | |
| CN103236064A (en) | Point cloud automatic registration method based on normal vector | |
| CN103150747A (en) | Point cloud registration algorithm based on topological characteristic | |
| CN108830888B (en) | Coarse matching method based on improved multi-scale covariance matrix characteristic descriptor | |
| CN104318551B (en) | Gauss hybrid models point cloud registration method based on convex closure characteristic key | |
| CN106446773A (en) | Automatic robust three-dimensional face detection method | |
| CN102750449B (en) | Point cloud linear feature extraction method based on substep three-dimensional space and feature dimension mapping | |
| CN104899918A (en) | Three-dimensional environment modeling method and system for unmanned plane | |
| CN103400136A (en) | Target identification method based on elastic matching | |
| CN109255815A (en) | A kind of object detection and recognition methods based on order spherical harmonic | |
| CN106682575A (en) | Human eye point cloud feature location with ELM (Eye Landmark Model) algorithm | |
| Liu et al. | Coarse registration of point clouds with low overlap rate on feature regions | |
| Ma et al. | Efficient rotation estimation for 3D registration and global localization in structured point clouds | |
| Giachetti et al. | SHREC’14 track: automatic location of landmarks used in manual anthropometry | |
| CN106023314A (en) | B spline master curve fitting method based on rotary axis direction mapping |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |