CN109375195B - Parameter quick calibrating method outside a kind of multi-line laser radar based on orthogonal normal vector - Google Patents
Parameter quick calibrating method outside a kind of multi-line laser radar based on orthogonal normal vector Download PDFInfo
- Publication number
- CN109375195B CN109375195B CN201811401717.1A CN201811401717A CN109375195B CN 109375195 B CN109375195 B CN 109375195B CN 201811401717 A CN201811401717 A CN 201811401717A CN 109375195 B CN109375195 B CN 109375195B
- Authority
- CN
- China
- Prior art keywords
- orthogonal
- scene
- normal vector
- laser radar
- line laser
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
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
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
- G01S7/497—Means for monitoring or calibrating
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Optical Radar Systems And Details Thereof (AREA)
Abstract
The invention discloses parameter quick calibrating methods outside a kind of multi-line laser radar based on orthogonal normal vector, are mainly concerned with pilotless automobile technical field of environmental perception, this method step are as follows: selection one has flat ground and has the scene in corner;The automatic driving vehicle for being mounted with multi-line laser radar is faced a wherein face wall to park;The radar data point set being radiated on three faces is chosen, P1, P2, P3 are labeled as;Each data point set is calculated, the normal vector I1, I2, I3 of point set are obtained;Calculate normal direction quantity set W1, W2, W3 of the metope under bodywork reference frame corresponding to data point set;It marks NI=[I1, I2, I3], NW=[W1, W2, W3];Singular value decomposition is carried out to H, obtains spin matrix R.The present invention does not need special calibration facility and stringent calibration scene, has the advantages that principle is simple and convenient to operate and stated accuracy is high.
Description
Technical field
Present invention relates generally to pilotless automobile technical field of environmental perception, specifically a kind of to be based on Orthogonal Method
The quick calibrating method of the outer parameter of the multi-line laser radar of vector.
Background technique
Multi-line laser radar is a kind of important sensor that pilotless automobile is equipped with, and is assumed responsibility for during automatic Pilot
Road Detection, positive and negative obstacle recognition and positioning in real time with build the vital tasks such as figure, have that measuring speed is fast, precision is high and surveys
Away from it is remote the advantages that, be widely applied in unmanned field.
In practical applications, in order to realize better environment sensing effect, or certain special applications demands are based on, it is multi-thread
Laser radar often presses a certain angle tilt installation.Therefore, the external parameters calibration of multi-line laser radar is exactly to set up laser
Transformation relation between radar fix system and bodywork reference frame, whether scaling method is simple, quick and accurate, to multi-thread laser
Application of the radar in pilotless automobile field is of great significance.
With the extensive use of multi-line laser radar, more and more laser radar scaling methods are studied, such as patent
201710009510.9 propose a kind of laser radar scaling method, by moving laser radar, in two different positions
The same scaling board is detected, the detection data of scaling board and the relative displacement data of radar itself are obtained, adjusts radar
Calibrating parameters.This method is needed using scaling board (detection target), and telecontrol equipment (has and carries out very small locomotivity),
It needs that inertia measurement unit, global positioning system or other modes are installed on telecontrol equipment and measures laser radar two and adopt
Relative displacement between the sample moment.Therefore, the calibration mode in practice to realize its demarcate purpose, need it is a set of complexity it is auxiliary
Assistant engineer's tool (one with position detecting function and can be carried out the telecontrol equipment of very small movement, and can accurately record its movement away from
From), need (at least twice) multiple to radar movable to realize entire calibration process.Patent ZL201710320952.5 is proposed
A kind of caliberating device and method of multi-line laser radar carry out suitably by designing a kind of caliberating device of complexity, and to device
Operation, the calibration process of Lai Shixian radar.Patent ZL201710515747.4 also proposed a kind of outer parameter of laser radar from
The method and apparatus of dynamic calibration, this method, which uses, places two class markers: for circular cone, the second class marker is first kind marker
Sphere obtains the radar data beaten in marker privileged site respectively, and the automatic Calibration for realizing radar is fitted by algorithm.
Existing multi-line laser radar scaling method requires special caliberating device or special marker to cooperate
Calibration, needs specifically to demarcate scene to place caliberating device or marker, these scaling methods are to pilotless automobile vehicle
It is for the use of load multi-line laser radar and inconvenient.
Summary of the invention
The present invention for existing unmanned field multi-line laser radar scaling method need special caliberating device or
The problem of person's marker, the invention proposes parameter Fast Calibration sides outside a kind of multi-line laser radar based on orthogonal normal vector
Method, the step are as follows:
(1) finding one has flat ground and has the scene in flat corner as calibration scene;Calibration scene be two or
Three orthogonal planes;When demarcating scene is two orthogonal faces, two orthogonal faces an are as follows: ground
Add two metopes in a metope or corner;When demarcating scene is three orthogonal planes, three orthogonal
Face is two metopes that a ground adds corner;
(2) automatic driving vehicle for being mounted with multi-line laser radar a face wall is faced to park;
(3) the radar data point set on three faces is chosen respectively, is labeled as P1, P2, P3;
(4) each data point set is calculated, obtains the normal vector I1, I2, I3 of point set;
(5) normal direction quantity set W1, W2, W3 of the plane under bodywork reference frame corresponding to data point set are calculated;
(6) NI=[I1, I2, I3] is marked, NW=[W1, W2, W3], NI and NW is respectively 3 × 3 matrix here, is calculated
Its product matrix H=NI × NW;
(7) singular value decomposition is carried out to H, obtained spin matrix R=svd (H), R is from multi-line laser radar coordinate system
The calibrating parameters converted to bodywork reference frame.
As a further improvement of the present invention:
Calibration scene in this method is two orthogonal planes, i.e. a ground adds a metope or corner
Two metopes, it is inevitable mutually orthogonal between the normal vector of two orthogonal planes, orthogonal normal vector as calibration feature,
Calibration result is obtained by singular value decomposition;Or calibration scene is selected as three orthogonal planes, i.e. a ground
Add two metopes in corner, inevitable mutually orthogonal, orthogonal normal vector conduct between the normal vector of three orthogonal planes
Feature is demarcated, calibration result is obtained by singular value decomposition.
This method can both be demarcated using two orthogonal flat surfaces, can also be orthogonal flat flat using three
Face is demarcated, and precision can meet automatic driving vehicle environment sensing demand, but with three orthogonal flat surfaces into
Rower is set to optimal scaling scheme.
Compared with the prior art, the advantages of the present invention are as follows, special caliberating device and stringent calibration environment are not needed,
Greatly reduce the requirement on hardware, reduce costs using radar point set normal vector and plane normal vector as feature into
Rower is fixed, has higher precision and better robustness than traditional radar points cloud scaling method, this method is not necessarily to special mark
Determine device, the multi-line laser radar on automatic driving vehicle is mounted on for any angle, as long as generally existing using indoor and outdoor
Ground and the natural scenes such as corner, the orthogonal normal vector obtained by orthogonal plane pass through surprise as calibration feature
Different value operation splitting, it will be able under multi-line laser radar calibration to bodywork reference frame;Operation of the present invention step is simple, precision
Height is used directly for multi-line laser radar in the external parameters calibration in pilotless automobile field, can generate very high economy
Benefit.
Detailed description of the invention
Fig. 1 is flow diagram of the invention;
Fig. 2 is present invention calibration scene operation schematic diagram.
Specific embodiment
The present invention is described in further details below with reference to Figure of description and specific embodiment.
As shown in Fig. 2, when multi-line laser radar to be calibrated according to the visual field need or mission requirements, at an angle
It is proposed by the present invention a kind of based on the multi-thread of orthogonal normal vector in conjunction with Fig. 1 after being installed on a pilotless automobile platform
Outside laser radar the step of parameter quick calibrating method are as follows:
(1) find that an indoor or outdoors are seen everywhere there are the scenes that three are mutually perpendicular to plane, such as spacious room
Interior, indoor garage, outdoor roadside the scenes such as corner, step;
(2) vehicle is faced one of plane to park, while ensures that a plurality of line of multi-line laser radar can
Three planes for calibration are irradiated to, range of exposures is wider, and effect is better;
(3) three numbers are formed from the radar data that selection is radiated in three planes respectively in the data of multi-line laser radar
Strong point collection is labeled as P1, P2, P3, as long as in principle in each data point set there are three radar data point not on one wire,
A plane can be calculated, in practical application, the area which includes is bigger, then precision is higher;
(4) data of each data point set are calculated, obtains the normal vector of point set, be labeled as I1, I2, I3;
(5) normal vector of the plane under bodywork reference frame corresponding to data point set is calculated, W1, W2, W3 are labeled as.Because
Pilotless automobile faces one of plane, therefore the normal direction of the plane and orthogonal plane under bodywork reference frame
Amount can directly obtain;
(6) NI=[I1, I2, I3] is marked, NW=[W1, W2, W3], NI and NW is respectively 3 × 3 matrix here, is calculated
Its product matrix H=NI × NW;
(7) singular value decomposition is carried out to H, obtained spin matrix R=svd (H), R is from multi-line laser radar coordinate system
The calibrating parameters converted to bodywork reference frame.
Pass through above step, it will be able under the calibration to bodywork reference frame of multi-line laser radar coordinate system.Since car body is sat
Translation matrix can be obtained by measurement between the origin of system and the installation origin of multi-line laser radar by marking.
From the foregoing, it will be observed that the present invention obtains external parameters calibration according to the orthogonal normal vector under two coordinate systems of singular value decomposition
Spin matrix this feature, propose a kind of method of multi-line laser radar Fast Calibration based on orthogonal normal vector, i.e., it is first
First finding one, there are two or three orthogonal planes, as calibration scene;Meanwhile just using automatic driving vehicle
Some plane against in calibration scene, so as to directly acquire normal direction of the orthogonal plane under bodywork reference frame
Amount;Then the data point set in Different Plane is radiated at by multi-line laser radar, each plane can also be calculated multi-thread
Normal vector under laser radar coordinate system;Using two groups of orthogonal normal vectors under two coordinate systems, pass through mature singular value
Decomposition method, so that it may obtain the transformational relation between two coordinate systems.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (2)
1. parameter quick calibrating method outside a kind of multi-line laser radar based on orthogonal normal vector, it is characterised in that step are as follows:
(1) it selects one to have flat ground and has the scene in corner as calibration scene;Calibration scene is that two or three are mutual
Vertical plane;When demarcating scene is two orthogonal faces, two orthogonal faces an are as follows: ground adds a wall
Two metopes in face or corner;When demarcating scene is three orthogonal planes, three orthogonal faces are one
Ground adds two metopes in corner;
(2) automatic driving vehicle for being mounted with multi-line laser radar a wherein face wall is faced to park;
(3) the radar data point set for choosing three faces respectively, is labeled as P1, P2, P3;
(4) each data point set is calculated, obtains the normal vector I1, I2, I3 of point set;
(5) normal direction quantity set W1, W2, W3 of three faces under bodywork reference frame corresponding to data point set are calculated, since nobody drives
It sails vehicle and faces some plane demarcated in scene, orthogonal plane can be directly acquired under bodywork reference frame
Normal vector;
(6) NI=[I1, I2, I3] is marked, NW=[W1, W2, W3], NI and NW is respectively 3 × 3 matrix here, calculates it and multiplies
Product matrix H=NI × NW;
(7) singular value decomposition is carried out to H, obtained spin matrix R=svd (H), R is from multi-line laser radar coordinate system to vehicle
The calibrating parameters of body coordinate system transformation;When demarcating scene is two orthogonal faces, the method for two orthogonal planes
Inevitable mutually orthogonal between vector, orthogonal normal vector obtains calibration result by singular value decomposition as calibration feature;Work as mark
It is inevitable mutually orthogonal between the normal vector of three orthogonal planes when to determine scene be three orthogonal planes, it is orthogonal
Normal vector as calibration feature, calibration result is obtained by singular value decomposition.
2. parameter quick calibrating method outside the multi-line laser radar according to claim 1 based on orthogonal normal vector, special
Sign is, when demarcating scene is two orthogonal faces, carries out calibration result using two orthogonal flat surfaces;Work as mark
When to determine scene be three orthogonal planes, calibration results are carried out using three orthogonal flat surfaces.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811401717.1A CN109375195B (en) | 2018-11-22 | 2018-11-22 | Parameter quick calibrating method outside a kind of multi-line laser radar based on orthogonal normal vector |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811401717.1A CN109375195B (en) | 2018-11-22 | 2018-11-22 | Parameter quick calibrating method outside a kind of multi-line laser radar based on orthogonal normal vector |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109375195A CN109375195A (en) | 2019-02-22 |
CN109375195B true CN109375195B (en) | 2019-07-16 |
Family
ID=65377350
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811401717.1A Active CN109375195B (en) | 2018-11-22 | 2018-11-22 | Parameter quick calibrating method outside a kind of multi-line laser radar based on orthogonal normal vector |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109375195B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112147598A (en) * | 2019-06-27 | 2020-12-29 | 武汉爱速达机器人科技有限公司 | Laser calibration method based on right-angle wall surface |
CN110320506B (en) * | 2019-08-06 | 2023-05-02 | 深圳市海梁科技有限公司 | Automatic calibration device and method for automotive laser radar |
CN112630751A (en) * | 2019-10-09 | 2021-04-09 | 中车时代电动汽车股份有限公司 | Calibration method of laser radar |
CN111121625B (en) * | 2019-12-27 | 2021-07-27 | 安徽意欧斯物流机器人有限公司 | Method for calibrating relative position of diagonally-arranged double laser radars |
CN111007485B (en) * | 2020-03-09 | 2020-10-27 | 中智行科技有限公司 | Image processing method and device and computer storage medium |
CN111190153B (en) * | 2020-04-09 | 2020-08-25 | 上海高仙自动化科技发展有限公司 | External parameter calibration method and device, intelligent robot and computer readable storage medium |
CN112485773B (en) * | 2020-11-09 | 2023-06-06 | 中国人民解放军军事科学院国防科技创新研究院 | External parameter information calibration method for laser radar and inclination angle sensor |
CN112560800B (en) * | 2021-01-12 | 2024-05-28 | 知行汽车科技(苏州)股份有限公司 | Road edge detection method, device and storage medium |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100235129A1 (en) * | 2009-03-10 | 2010-09-16 | Honeywell International Inc. | Calibration of multi-sensor system |
CN104656097B (en) * | 2015-01-28 | 2017-03-08 | 武汉理工大学 | Caliberating device based on rotary two-dimensional laser three-dimensional reconfiguration system and method |
US10359507B2 (en) * | 2016-12-30 | 2019-07-23 | Panosense Inc. | Lidar sensor assembly calibration based on reference surface |
CN107870324B (en) * | 2017-05-09 | 2021-06-25 | 吉林大学 | Calibration device and method for multi-line laser radar |
CN107179534B (en) * | 2017-06-29 | 2020-05-01 | 北京北科天绘科技有限公司 | Method and device for automatically calibrating laser radar parameters and laser radar |
-
2018
- 2018-11-22 CN CN201811401717.1A patent/CN109375195B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN109375195A (en) | 2019-02-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109375195B (en) | Parameter quick calibrating method outside a kind of multi-line laser radar based on orthogonal normal vector | |
EP2201532B1 (en) | Local positioning system and method | |
US7859655B2 (en) | Method involving a pointing instrument and a target object | |
CN110221623A (en) | A kind of air-ground coordination operating system and its localization method | |
US20170054954A1 (en) | System and method for visually displaying information on real objects | |
CN108919825A (en) | The unmanned plane indoor locating system and method for having barrier avoiding function | |
US20140160115A1 (en) | System And Method For Visually Displaying Information On Real Objects | |
CN108594245A (en) | A kind of object movement monitoring system and method | |
CN105184776A (en) | Target tracking method | |
CN109613584B (en) | UWB-based positioning and orientation method for unmanned card concentrator | |
CN105300375A (en) | Robot indoor positioning and navigation method based on single vision | |
CN110132130A (en) | Laser radar position calibration method, system and its data processing method, system | |
CN105445729A (en) | Unmanned plane three-dimensional flight track precision detection method and system | |
CN102998689B (en) | Region decision method based on virtual-sensor | |
CN108132029A (en) | A kind of accurate measurement method and device of satellite antenna development system assembling | |
CN105301621A (en) | Vehicle positioning device and intelligent driving exam system | |
Peng et al. | A measuring method for large antenna assembly using laser and vision guiding technology | |
CN112598756B (en) | Roadside sensor calibration method and device and electronic equipment | |
CN114035150A (en) | Radio frequency source direction finding device and positioning method based on unmanned aerial vehicle lift-off platform | |
CN105571491A (en) | Binocular vision-based automobile chassis data measuring system and method thereof | |
CN117310627A (en) | Combined calibration method applied to vehicle-road collaborative road side sensing system | |
Meissner et al. | Simulation and calibration of infrastructure based laser scanner networks at intersections | |
Wu et al. | A novel precise guiding method for visual guiding theodolite measurement in volume space | |
CN102798391A (en) | Laser image positioning navigation device | |
de Campos Porath et al. | Feasibility of measurement-assisted assembly of ship hull blocks |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |