CN110097593A - A method of identifying cylindrical surface from multi-line laser radar point cloud data - Google Patents

A method of identifying cylindrical surface from multi-line laser radar point cloud data Download PDF

Info

Publication number
CN110097593A
CN110097593A CN201910300276.4A CN201910300276A CN110097593A CN 110097593 A CN110097593 A CN 110097593A CN 201910300276 A CN201910300276 A CN 201910300276A CN 110097593 A CN110097593 A CN 110097593A
Authority
CN
China
Prior art keywords
cylindrical surface
laser radar
coordinate
unit
point
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.)
Withdrawn
Application number
CN201910300276.4A
Other languages
Chinese (zh)
Inventor
张程
王建华
张山甲
赵明绘
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Maritime University
Original Assignee
Shanghai Maritime University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Maritime University filed Critical Shanghai Maritime University
Priority to CN201910300276.4A priority Critical patent/CN110097593A/en
Publication of CN110097593A publication Critical patent/CN110097593A/en
Withdrawn legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
    • G01S17/06—Systems determining position data of a target
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/87—Combinations of systems using electromagnetic waves other than radio waves
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/88—Lidar systems specially adapted for specific applications
    • G01S17/89—Lidar systems specially adapted for specific applications for mapping or imaging
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10028—Range image; Depth image; 3D point clouds
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10032—Satellite or aerial image; Remote sensing
    • G06T2207/10044—Radar image

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Electromagnetism (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Geometry (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Theoretical Computer Science (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Optical Radar Systems And Details Thereof (AREA)

Abstract

The method that the present invention relates to a kind of to identify cylindrical surface from multi-line laser radar point cloud data.This method sets cylindrical surface fit mathematics model first；Then it carries out cylindrical surface and is fitted pre-treatment step, the two-dimensional surface ellipse fitting coordinate of every line laser beam section is determined using RANSAC algorithm and spin matrix；Cylindrical surface fit procedure is carried out again, is solved oval optimization aim equation using least square method and Lagrange multiplier algorithm construction, is randomly selected match point and solve to obtain optimal fitted ellipse using least square short axle inclination angle constraint condition；Seven, cylindrical surface parameter is calculated finally by singular value decomposition and optimization.The present invention is able to solve in environment sensing field, multi-line laser radar laser radar during combining outer calibration with camera identifies cylindrical surface also precision with higher for the multi-line laser radar of low resolution using cylindrical surface as the cylindrical surface identification problem encountered when calibration object.

Description

A method of identifying cylindrical surface from multi-line laser radar point cloud data
Technical field
The invention belongs to environment sensing fields more particularly to a kind of from multi-line laser radar point cloud data identification cylindrical surface Method.
Background technique
Previous laser radar is not received extensive attention and is answered due to the limitation of purchase cost and use condition With.But in recent years, with the rapid development of laser radar technique, application field is constantly widened, and is especially led in environment sensing Domain is even more the main stream approach for becoming unmanned systems and using using the perception external environment that merges of laser radar and camera.Laser Radar is merged with the heterogeneous of camera firstly the need of using calibration object to carry out combining outer calibration, and the selection for demarcating object, which directly affects, melts Close effect.It is exactly chequered with black and white gridiron pattern that the most common calibration object in document is demarcated outside existing joint, but is made using gridiron pattern Need to carry out the outer calibration of offline joint in stable experimental situation for calibration object.
The angle that outer stated accuracy is combined in collateral security is set out, due to the presence of disturbing factor in actual environment, with the time Passage, combining outer stated accuracy inevitably reduces, this, which just needs to find one kind in actual environment, can be used as laser Radar carries out combining the outer calibration object demarcated online with camera.And it can not efficiently find in actual environment similar to gridiron pattern The object of feature, but there is the more objects with cylinder region feature in actual environment, such as cylindrical body, the bridge in building The objects such as pier, beacon waterborne and cylindrical surface type buoy.Using the cylindrical surface in actual environment as the mark in the outer calibration of joint Earnest is less to be referred to by people, and what mainly laser radar obtained is sparse incomplete cylindrical surface point cloud data, existing cylindrical surface The Research Literature of fitting is that the three-dimensional laser scanner based on high-accuracy high-resolution obtains substantially complete cylinder millet cake cloud number It is carried out on the basis of, the rare research of fitting for sparse endless integral point cloud cylindrical surface.In order to solve environment sensing field In, multi-line laser radar identifies problem for cylindrical surface as the cylindrical surface that encounters of calibration object combining external standard timing with camera, Invention is based on the above-mentioned fact, proposes a kind of for multi-line laser radar point cloud data identification cylindrical surface method.
Summary of the invention
The present invention to solve the above-mentioned problems, proposes a kind of side that cylindrical surface is identified from multi-line laser radar point cloud data Method, it is characterised in that comprise the steps of:
Step 1: setting cylindrical surface mathematical model
It can be obtained by the geometrical property on cylindrical surface, the distance of point to its axis on cylindrical surface is constantly equal to radius r0, it may be assumed that
Wherein, P (x, y, z)TFor any point on cylindrical surface, P0(x0,y0,z0)TFor on cylinder axis a bit, L (a, b, c)T For cylinder axis unit vector, r0For cylinder bottom radius of circle.A bit (x on cylinder axis0,y0,z0)T, cylinder axis unit vector (a,b,c)T, cylinder bottom radius of circle r0, this seven parameters can uniquely determine a cylindrical surface.
Step 2: cylindrical surface fitting pretreatment
The imperfect cylindrical surface point cloud data that laser radar scanning cylindrical surface is obtained carries out plane by RANSAC method Fitting, it is assumed that plane equation are as follows:
F (X, P)=XTP+d=0 (2)
In formula (2), X=(x, y, z)TFor arbitrary point cloud coordinate；P=(k, m, n)TFor planar unit normal vector；D is origin To the distance of plane, unit: mm.RANSA method is sought most according to measurement point to the principle that plan range quadratic sum meets threshold value Good plane equation parameter P and d, therefore error equation are as follows:
ei=F (Xi,P) (3)
Error equation is listed to all the points, using threshold decision condition (4), looks for and meets the most points of Rule of judgment, Then parameter P, d are determined；
ei<V (4)
In formula (4), that V is represented is range error threshold value, unit mm；It is also the flatness of last fit Plane, i point exists Coordinate (the xp of subpoint in planei,ypi,zpi)TAre as follows:
After finding out projection coordinate, the point cloud projection of each line laser radar is one section of three-dimensional elliptical arc, for three-dimensional elliptical Its equation complicated difficult is fitted effectively to solve, two-dimensional elliptic fitting can be converted by coordinate conversion, is turned by coordinate Elliptic arc equation its parameter changed will not change.
Coordinate conversion is carried out to subpoint, makes planar unit normal vector (k, m, n)TIt is transformed to be parallel to the vector of Z axisThat is (0,0,1)T.If spin matrix is R3x3, to make:
I.e.
By spin matrix and subpoint coordinate (xpi,ypi,zpi)TIt is multiplied, obtains new coordinate (x 'i,y′i,z′i)T, will be new X ' in coordinatei、y′iTake out (x ', y ')T。(x′,y′)TFor the coordinate being parallel in the plane of XOY plane, according to (x ', y′)TIt can carry out planar elliptical fitting.
Step 3: cylindrical surface fitting
Due to the influence of itself and site environment of laser radar, the inevitable area of cylindrical surface point cloud data of acquisition There is noise, these noises will affect the fitting precision on cylindrical surface, so needing to carry out noise reduction to obtained cylindrical surface point cloud data Processing.The noise of laser radar is similar to Gaussian noise, so point cloud noise-reduction method of the invention is removed far from cylinder first The shift point of main part, point cloud data remaining for each line uses moving average filtering method later.Calculation formula is as follows:
Wherein, R ' (j) is range points of the laser radar to cylindrical surface after j-th of filtering, unit: mm；R (j) is filter Measurement range points of j-th of laser radar to cylindrical surface, unit: mm before wave；Queue length J is set according to experiment value.Filtering A cloud coordinate (x, y, z) is calculated as by the angle initialization rule of laser radar both vertically as well as horizontally afterwardsT。
Since the point cloud of each line laser of radar on the cylinder is elliptic arc feature, and by laser radar after noise reduction Noise still exist, the present invention use a kind of elliptical effective approximating method for noise or circumstance of occlusion.Using swash The elliptic arc that optical radar is formed on cylindrical surface this symmetrical geometry on the direction of axis and the laser radar line of centres is special Property, take the least square short axle inclination angle ellipse fitting method based on algebraic distance that all the points are not all classified as to match point.
Two-dimensional surface ellipse fitting method step can be summarized are as follows:
1) according to this geometrically symmetric characteristic of elliptic arc, match point is randomly selected using following formula；
Wherein (x 'sel,y′sel)TIndicate the point set to be fitted chosen, (x 's,y′s)TPoint after s-th of rotation, tot are (x′,y′)TElement sum, behalf be choose interval.
2)(x′sel,y′sel)TIt is oval after being 1 by constant term naturalization in elliptic equation for the coordinate points for carrying out ellipse fitting Equation may be expressed as:
Ax2+Bxy+Cy2+ Dx+Ey+1=0 (10)
Enable U=[A, B, C, D, E, 1]T, V=[x2,xy,y2,x,y,1]T, then optimization aim are as follows:
Wherein
According to method of Lagrange multipliers, Lagrange factor λ is introduced, constructs Lagrangian
L (U)=UTVVTU-(UTHU-1) (12)
Peer-to-peer (12) derivation obtains zero
VVTU=λ HU (13)
The optimal solution of formula (11) is obtained by solving the generalized eigenvalue problem in equation (13).
To acquire elliptical geometrical characteristic parameter are as follows:
Elliptical center coordinate (unit: mm):
Long semi-minor axis square (unit: mm2):
Short axle inclination angle:
3) s is successively repeated into step 1)-step 2) according to 1,2,3,5,7,9 setting values, chooses min ‖ θ ‖2It is corresponding Oval coefficient U '=[A ', B ', C ', D ', E ', 1]TAs final ellipse fitting effect.
The cylinder point cloud data that line laser beam each for laser radar obtains execute above-mentioned step 1), 2), 3) after, The elliptical geometrical characteristic parameter of γ group can be obtained, singular value decomposition, available cylinder are carried out to γ elliptical center coordinate The axis unit vector (a, b, c) in faceT, a bit (x on cylinder axis0,y0,z0)TIt is set as γ -1 line laser Shu Yuanzhu millet cake cloud The intersection point of place plane and axis；Cylindrical surface bottom radius of circle is obtained by formula (17).
Wherein, γ is the laser harness number of the multi-line laser radar used；βiIt is the flat of the i-th line laser beam scanning formation Face normal vector and unit vector (0,0,1)TAngle, unit is rad；aciIt is the short axle that i-th of ellipse fitting is calculated, it is single Position is mm.
The present invention is simple and efficient real-time and realizability with higher, energy compared with existing cylindrical surface fitting technique Enough quickly identification cylindrical surfaces；Method proposed by the present invention, can be for using multi-thread to the identification on cylindrical surface precision with higher The unmanned systems of laser radar improve external environment sensing capability.
Detailed description of the invention:
Fig. 1 is a kind of flow chart of method that cylindrical surface is identified from multi-line laser radar point cloud data.
Specific embodiment:
It is the preferred embodiment of the present invention and in conjunction with attached drawing below, technical scheme of the present invention will be further described, But the present invention is not limited to embodiments.
As shown in Figure 1, the method for the invention for identifying cylindrical surface from multi-line laser radar point cloud data includes following step It is rapid:
Step 1: setting cylindrical surface mathematical model
It can be obtained by the geometrical property on cylindrical surface, the distance of point to its axis on cylindrical surface is constantly equal to radius r0, it may be assumed that
Wherein, P (x, y, z)TFor any point on cylindrical surface, P0(x0,y0,z0)TFor on cylinder axis a bit, L (a, b, c)T For cylinder axis unit vector, r0For cylinder bottom radius of circle.A bit (x on cylinder axis0,y0,z0)T, cylinder axis unit vector (a,b,c)T, cylinder bottom radius of circle r0, this seven parameters can uniquely determine a cylindrical surface.
Step 2: cylindrical surface fitting pretreatment
The imperfect cylindrical surface point cloud data that laser radar scanning cylindrical surface is obtained carries out plane by RANSAC method Fitting, it is assumed that plane equation are as follows:
F (X, P)=XTP+d=0 (2)
In formula (2), X=(x, y, z)TFor arbitrary point cloud coordinate；P=(k, m, n)TFor planar unit normal vector；D is origin To the distance of plane, unit: mm.RANSA method is sought most according to measurement point to the principle that plan range quadratic sum meets threshold value Good plane equation parameter P and d, therefore error equation are as follows:
ei=F (Xi,P) (3)
Error equation is listed to all the points, using threshold decision condition (4), looks for and meets the most points of Rule of judgment, Then parameter P, d are determined；
ei<V (4)
In formula (4), that V is represented is range error threshold value, unit mm；It is also the flatness of last fit Plane, i point exists Coordinate (the xp of subpoint in planei,ypi,zpi)TAre as follows:
After finding out projection coordinate, the point cloud projection of each line laser radar is one section of three-dimensional elliptical arc, for three-dimensional elliptical Its equation complicated difficult is fitted effectively to solve, two-dimensional elliptic fitting can be converted by coordinate conversion, is turned by coordinate Elliptic arc equation its parameter changed will not change.
Coordinate conversion is carried out to subpoint, makes planar unit normal vector (k, m, n)TIt is transformed to be parallel to the vector of Z axisThat is (0,0,1)T.If spin matrix is R3x3, to make:
I.e.
By spin matrix and subpoint coordinate (xpi,ypi,zpi)TIt is multiplied, obtains new coordinate (x 'i,y′i,z′i)T, will be new X ' in coordinatei、y′iTake out (x ', y ')T。(x′,y′)TFor the coordinate being parallel in the plane of XOY plane, according to (x ', y′)TIt can carry out planar elliptical fitting.
Step 3: cylindrical surface fitting
Due to the influence of itself and site environment of laser radar, the inevitable area of cylindrical surface point cloud data of acquisition There is noise, these noises will affect the fitting precision on cylindrical surface, so needing to carry out noise reduction to obtained cylindrical surface point cloud data Processing.The noise of laser radar is similar to Gaussian noise, so point cloud noise-reduction method of the invention is removed far from cylinder first The shift point of main part, point cloud data remaining for each line uses moving average filtering method later.Calculation formula is as follows:
Wherein, R ' (j) is range points of the laser radar to cylindrical surface after j-th of filtering, unit: mm；R (j) is filter Measurement range points of j-th of laser radar to cylindrical surface, unit: mm before wave；Queue length J is set according to experiment value.Filtering A cloud coordinate (x, y, z) is calculated as by the angle initialization rule of laser radar both vertically as well as horizontally afterwardsT。
Since the point cloud of each line laser of radar on the cylinder is elliptic arc feature, and by laser radar after noise reduction Noise still exist, the present invention use a kind of elliptical effective approximating method for noise or circumstance of occlusion.Using swash The elliptic arc that optical radar is formed on cylindrical surface this symmetrical geometry on the direction of axis and the laser radar line of centres is special Property, take the least square short axle inclination angle ellipse fitting method based on algebraic distance that all the points are not all classified as to match point.
Two-dimensional surface ellipse fitting method step can be summarized are as follows:
2) according to this geometrically symmetric characteristic of elliptic arc, match point is randomly selected using following formula；
Wherein (x 'sel,y′sel)TIndicate the point set to be fitted chosen, (x 's,y′s)TPoint after s-th of rotation, tot are (x′,y′)TElement sum, behalf be choose interval.
2)(x′sel,y′sel)TIt is oval after being 1 by constant term naturalization in elliptic equation for the coordinate points for carrying out ellipse fitting Equation may be expressed as:
Ax2+Bxy+Cy2+ Dx+Ey+1=0 (10)
Enable U=[A, B, C, D, E, 1]T, V=[x2,xy,y2,x,y,1]T, then optimization aim are as follows:
Wherein
According to method of Lagrange multipliers, Lagrange factor λ is introduced, constructs Lagrangian
L (U)=UTVVTU-(UTHU-1) (12)
Peer-to-peer (12) derivation obtains zero
VVTU=λ HU (13)
The optimal solution of formula (11) is obtained by solving the generalized eigenvalue problem in equation (13).
To acquire elliptical geometrical characteristic parameter are as follows:
Elliptical center coordinate (unit: mm):
Long semi-minor axis square (unit: mm2):
Short axle inclination angle:
3) s is successively repeated into step 1)-step 2) according to 1,2,3,5,7,9 setting values, chooses min ‖ θ ‖2It is corresponding Oval coefficient U '=[A ', B ', C ', D ', E ', 1]TAs final ellipse fitting effect.
The cylinder point cloud data that line laser beam each for laser radar obtains execute above-mentioned step 1), 2), 3) after, The elliptical geometrical characteristic parameter of γ group can be obtained, singular value decomposition, available cylinder are carried out to γ elliptical center coordinate The axis unit vector (a, b, c) in faceT, a bit (x on cylinder axis0,y0,z0)TIt is set as γ -1 line laser Shu Yuanzhu millet cake cloud The intersection point of place plane and axis；Cylindrical surface bottom radius of circle is obtained by formula (17).
Wherein, γ is the laser harness number of the multi-line laser radar used；βiIt is the flat of the i-th line laser beam scanning formation Face normal vector and unit vector (0,0,1)TAngle, unit is rad；aciIt is the short axle that i-th of ellipse fitting is calculated, it is single Position is mm.

Claims (1)

1. a kind of method for identifying cylindrical surface from multi-line laser radar point cloud data, it is characterised in that comprise the steps of:
Step 1: setting cylindrical surface mathematical model
It can be obtained by the geometrical property on cylindrical surface, the distance of point to its axis on cylindrical surface is constantly equal to radius r0, it may be assumed that
Wherein, P (x, y, z)TFor any point on cylindrical surface, P0(x0, y0, z0)TFor on cylinder axis a bit, L (a, b, c)TFor circle Mast axis unit vector, r0For cylinder bottom radius of circle；A bit (x on cylinder axis0, y0, z0)T, cylinder axis unit vector (a, b, c)T, cylinder bottom radius of circle r0, this seven parameters can uniquely determine a cylindrical surface；
Step 2: cylindrical surface fitting pretreatment
The imperfect cylindrical surface point cloud data that laser radar scanning cylindrical surface is obtained carries out plane fitting by RANSAC method, Assuming that plane equation are as follows:
F (X, P)=XTP+d=0 (2)
In formula (2), X=(x, y, z)TFor arbitrary point cloud coordinate；P=(k, m, n)TFor planar unit normal vector；D is origin to flat The distance in face, unit: mm；RANSA method is sought most preferably flat according to measurement point to the principle that plan range quadratic sum meets threshold value Face equation parameter P and d, therefore error equation are as follows:
ei=F (Xi, P) and (3)
Error equation is listed to all the points, using threshold decision condition (4), looks for and meets the most points of Rule of judgment, then Determine parameter P, d；
ei< V (4)
In formula (4), that V is represented is range error threshold value, unit mm；It is also the flatness of last fit Plane, i point is in plane On subpoint coordinate (xpi, ypi, zpi)TAre as follows:
After finding out projection coordinate, the point cloud projection of each line laser radar is one section of three-dimensional elliptical arc, and can be converted with coordinate will It is converted to two-dimensional elliptic fitting, and elliptic arc equation its parameter by coordinate conversion will not change；
Coordinate conversion is carried out to subpoint, makes planar unit normal vector (k, m, n)TIt is transformed to be parallel to the vector of Z axisThat is (0,0,1)T；If spin matrix is R3×3, to make:
I.e.
By spin matrix and subpoint coordinate (xpi, ypi, zpi)TIt is multiplied, obtains new coordinate (x 'i, y 'i, z 'i)T, by new coordinate In x 'i、y′iTake out (x ', y ')T；(x ', y ')TFor the coordinate being parallel in the plane of XOY plane, according to (x ', y ')TIt can Carry out planar elliptical fitting；
Step 3: cylindrical surface fitting
The shift point far from cylindrical body portion is removed, point cloud data remaining for each line uses moving average filtering later Method；Calculation formula is as follows:
Wherein, R ' (j) is range points of the laser radar to cylindrical surface after j-th of filtering, unit: mm；R (j) is filtering Measurement range points of preceding j-th of the laser radar to cylindrical surface, unit: mm；Queue length J is set according to experiment value；Lead to after filtering It crosses the angle initialization rule of laser radar both vertically as well as horizontally and is calculated as a cloud coordinate (x, y, z)T；
Steps are as follows:
1) according to this geometrically symmetric characteristic of elliptic arc, match point is randomly selected using following formula；
Wherein (x 'sel, y 'sel)TIndicate the point set to be fitted chosen, (x 's, y 's)TS-th rotation after point, tot be (x ', y′)TElement sum, behalf be choose interval；
2)(x′sel, y 'sel)TFor the coordinate points for carrying out ellipse fitting, after being 1 by constant term naturalization in elliptic equation, elliptic equation It may be expressed as:
Ax2+Bxy+Cy2+ Dx+Ey+1=0 (10)
Enable U=[A, B, C, D, E, 1]T, V=[x2, xy, y2, x, y, 1]T, then optimization aim are as follows:
Wherein
According to method of Lagrange multipliers, Lagrange factor λ is introduced, constructs Lagrangian
L (U)=UTVVTU-(UTHU-1) (12)
Peer-to-peer (12) derivation obtains zero
VVTU=λ HU (13)
The optimal solution of formula (11) is obtained by solving the generalized eigenvalue problem in equation (13)；
To acquire elliptical geometrical characteristic parameter are as follows:
Elliptical center coordinate (unit: mm):
Long semi-minor axis square (unit: mm2):
Short axle inclination angle:
3) s is successively repeated into step 1)-step 2) according to 1,2,3,5,7,9 setting value, chooses min | | θ | |2It is corresponding ellipse Circle coefficient U '=[A ', B ', C ', D ', E ', 1]TAs final ellipse fitting effect；
The cylinder point cloud data that line laser beam each for laser radar obtains execute above-mentioned step 1), 2), 3) after, can be with The elliptical geometrical characteristic parameter of γ group is obtained, singular value decomposition is carried out to γ elliptical center coordinate, available cylindrical surface Axis unit vector (a, b, c)T, a bit (x on cylinder axis0, y0, z0)TIt is set as where γ -1 line laser Shu Yuanzhu millet cake cloud The intersection point of plane and axis；Cylindrical surface bottom radius of circle is obtained by formula (17):
Wherein, γ is the laser harness number of the multi-line laser radar used；βiIt is the planar process that the i-th line laser beam scanning is formed Vector and unit vector (0,0,1)TAngle, unit is rad；aciIt is the short axle that i-th of ellipse fitting is calculated, unit is mm。
CN201910300276.4A 2019-04-15 2019-04-15 A method of identifying cylindrical surface from multi-line laser radar point cloud data Withdrawn CN110097593A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910300276.4A CN110097593A (en) 2019-04-15 2019-04-15 A method of identifying cylindrical surface from multi-line laser radar point cloud data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910300276.4A CN110097593A (en) 2019-04-15 2019-04-15 A method of identifying cylindrical surface from multi-line laser radar point cloud data

Publications (1)

Publication Number Publication Date
CN110097593A true CN110097593A (en) 2019-08-06

Family

ID=67444912

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910300276.4A Withdrawn CN110097593A (en) 2019-04-15 2019-04-15 A method of identifying cylindrical surface from multi-line laser radar point cloud data

Country Status (1)

Country Link
CN (1) CN110097593A (en)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111350214A (en) * 2020-03-23 2020-06-30 中交第三航务工程局有限公司江苏分公司 Multi-beam underwater steel pipe pile position measuring method
CN111562563A (en) * 2020-04-30 2020-08-21 北京国电富通科技发展有限责任公司 Laser radar rotary table calibration method and device and computer readable storage medium
CN112017233A (en) * 2020-07-10 2020-12-01 广东电网有限责任公司广州供电局 Method and device for measuring conical appearance of reaction force, computer equipment and system
CN112102474A (en) * 2020-09-01 2020-12-18 长春工程学院 Novel cylindrical three-dimensional reconstruction method and system
CN112465979A (en) * 2020-11-10 2021-03-09 国网内蒙古东部电力有限公司检修分公司 Cylindrical surface point cloud fitting method based on geometric characteristics of sectioning surface
CN112489123A (en) * 2020-10-30 2021-03-12 江阴市智行工控科技有限公司 Three-dimensional positioning method for surface target of truck in steel mill reservoir area
CN112652068A (en) * 2021-01-30 2021-04-13 上海汇像信息技术有限公司 Circular truncated cone fitting method of 3D model of rotating cylinder
CN112859054A (en) * 2021-04-21 2021-05-28 合肥工业大学 Vehicle-mounted multi-line laser radar external parameter automatic detection system and detection method
CN113495256A (en) * 2020-04-01 2021-10-12 北京京东乾石科技有限公司 Method and device for determining accuracy of calibration results among multiple laser radars
CN113865508A (en) * 2021-09-28 2021-12-31 南京航空航天大学 Automatic detection device and method for through hole rate of acoustic lining of honeycomb sandwich composite material
CN113865429A (en) * 2021-07-20 2021-12-31 中国人民解放军63921部队 Active measurement method and system for real-time drift amount of rocket takeoff
CN114140586A (en) * 2022-01-29 2022-03-04 苏州工业园区测绘地理信息有限公司 Indoor space-oriented three-dimensional modeling method and device and storage medium
CN114322802A (en) * 2021-12-30 2022-04-12 苏州中科行智智能科技有限公司 Line diameter measuring method based on 3D line laser camera
CN114346528A (en) * 2021-12-09 2022-04-15 成都卡诺普机器人技术股份有限公司 Method for high-precision identification of plate pipe welding seam and acquisition of welding track based on laser scanning

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111350214B (en) * 2020-03-23 2021-07-30 中交第三航务工程局有限公司江苏分公司 Multi-beam underwater steel pipe pile position measuring method
CN111350214A (en) * 2020-03-23 2020-06-30 中交第三航务工程局有限公司江苏分公司 Multi-beam underwater steel pipe pile position measuring method
CN113495256B (en) * 2020-04-01 2024-04-05 北京京东乾石科技有限公司 Method and device for determining accuracy of calibration result among multiple laser radars
CN113495256A (en) * 2020-04-01 2021-10-12 北京京东乾石科技有限公司 Method and device for determining accuracy of calibration results among multiple laser radars
CN111562563A (en) * 2020-04-30 2020-08-21 北京国电富通科技发展有限责任公司 Laser radar rotary table calibration method and device and computer readable storage medium
CN112017233A (en) * 2020-07-10 2020-12-01 广东电网有限责任公司广州供电局 Method and device for measuring conical appearance of reaction force, computer equipment and system
CN112102474A (en) * 2020-09-01 2020-12-18 长春工程学院 Novel cylindrical three-dimensional reconstruction method and system
CN112489123A (en) * 2020-10-30 2021-03-12 江阴市智行工控科技有限公司 Three-dimensional positioning method for surface target of truck in steel mill reservoir area
CN112465979A (en) * 2020-11-10 2021-03-09 国网内蒙古东部电力有限公司检修分公司 Cylindrical surface point cloud fitting method based on geometric characteristics of sectioning surface
CN112652068A (en) * 2021-01-30 2021-04-13 上海汇像信息技术有限公司 Circular truncated cone fitting method of 3D model of rotating cylinder
CN112652068B (en) * 2021-01-30 2022-03-11 上海汇像信息技术有限公司 Circular truncated cone fitting method of 3D model of rotating cylinder
CN112859054A (en) * 2021-04-21 2021-05-28 合肥工业大学 Vehicle-mounted multi-line laser radar external parameter automatic detection system and detection method
CN112859054B (en) * 2021-04-21 2023-09-29 合肥工业大学 Automatic detection system and detection method for external parameters of vehicle-mounted multi-line laser radar
CN113865429A (en) * 2021-07-20 2021-12-31 中国人民解放军63921部队 Active measurement method and system for real-time drift amount of rocket takeoff
CN113865429B (en) * 2021-07-20 2023-03-14 中国人民解放军63921部队 Method and system for actively measuring real-time drift amount of rocket takeoff
CN113865508A (en) * 2021-09-28 2021-12-31 南京航空航天大学 Automatic detection device and method for through hole rate of acoustic lining of honeycomb sandwich composite material
CN114346528A (en) * 2021-12-09 2022-04-15 成都卡诺普机器人技术股份有限公司 Method for high-precision identification of plate pipe welding seam and acquisition of welding track based on laser scanning
CN114322802A (en) * 2021-12-30 2022-04-12 苏州中科行智智能科技有限公司 Line diameter measuring method based on 3D line laser camera
CN114140586B (en) * 2022-01-29 2022-05-17 苏州工业园区测绘地理信息有限公司 Three-dimensional modeling method and device for indoor space and storage medium
CN114140586A (en) * 2022-01-29 2022-03-04 苏州工业园区测绘地理信息有限公司 Indoor space-oriented three-dimensional modeling method and device and storage medium

Similar Documents

Publication Publication Date Title
CN110097593A (en) A method of identifying cylindrical surface from multi-line laser radar point cloud data
Kang et al. Automatic targetless camera–lidar calibration by aligning edge with gaussian mixture model
CN109166140B (en) Vehicle motion track estimation method and system based on multi-line laser radar
Zheng et al. The obstacle detection method of uav based on 2D lidar
CN103424112B (en) A kind of motion carrier vision navigation method auxiliary based on laser plane
CN107632308B (en) Method for detecting contour of obstacle in front of vehicle based on recursive superposition algorithm
Wang et al. I can see the light: Attacks on autonomous vehicles using invisible lights
CN110888120B (en) Method for correcting laser radar point cloud data motion distortion based on integrated navigation system
TWI624170B (en) Image scanning system and method thereof
CN106127787B (en) A kind of camera calibration method based on Inverse projection
CN107289946B (en) A kind of high-precision laser positioning and air navigation aid based on double reflectors
CN113627473B (en) Multi-mode sensor-based water surface unmanned ship environment information fusion sensing method
CN108061897B (en) Underwater structured environment line feature extraction method based on forward-looking sonar
CN111079765A (en) Sparse point cloud densification and pavement removal method based on depth map
CN113050074A (en) Camera and laser radar calibration system and calibration method in unmanned environment perception
CN116087982A (en) Marine water falling person identification and positioning method integrating vision and radar system
CN110119671A (en) Underwater cognitive method based on artificial side line visual image
CN108544491B (en) Mobile robot obstacle avoidance method comprehensively considering two factors of distance and direction
CN117406234A (en) Target ranging and tracking method based on single-line laser radar and vision fusion
CN208314563U (en) A kind of visual identifying system for robotic tracking
Yang et al. Target recognition using rotating ultrasonic sensor for an amphibious ROV
CN110567395A (en) Laser spot size edge contour line drawing method
Yoon et al. A distortion model of laser sheet for a laser line scanner with large fan angle
Mount et al. Automatic coverage selection for surface-based visual localization
Shi et al. Dense normal based degeneration-aware 2-D lidar odometry for correlative scan matching

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
WW01 Invention patent application withdrawn after publication

Application publication date: 20190806

WW01 Invention patent application withdrawn after publication