CN206825428U - A kind of real-time acquisition device of three-dimensional point cloud - Google Patents
A kind of real-time acquisition device of three-dimensional point cloud Download PDFInfo
- Publication number
- CN206825428U CN206825428U CN201720526994.XU CN201720526994U CN206825428U CN 206825428 U CN206825428 U CN 206825428U CN 201720526994 U CN201720526994 U CN 201720526994U CN 206825428 U CN206825428 U CN 206825428U
- Authority
- CN
- China
- Prior art keywords
- robot
- real
- point cloud
- displacement sensor
- laser displacement
- 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.)
- Expired - Fee Related
Links
Landscapes
- Manipulator (AREA)
Abstract
The utility model discloses a kind of real-time acquisition device of three-dimensional point cloud, including robot, laser displacement sensor, robot real-time control system, the laser displacement sensor is arranged on the end of the robot by fixture, the robot real-time control system connects robot and laser displacement sensor by real-time industrial ethernet bus, for making the reading of laser displacement sensor and the pose of robot obtain synchronization, one-dimensional measurement is expanded into three-dimensional measurement, so as to obtain the three-dimensional point cloud of workpiece by scanning.The utility model robot real-time control system, EPA bus make the reading of laser displacement sensor and the pose of robot obtain synchronization, and one-dimensional measurement is expanded into three-dimensional measurement, so as to obtain the three-dimensional point cloud of workpiece by scanning.It is reliable and stable, precision is high, has real-time and flexibility characteristics, and cost simple in construction is low, disclosure satisfy that the processing of different workpieces.
Description
Technical field
The utility model should be related to the acquisition device of three-dimensional point cloud, more particularly to a kind of real-time collection dress of three-dimensional point cloud
Put.
Background technology
Now, the acquisition equipment of three-dimensional point cloud mainly has three-dimensional laser scanner, and two-dimensional line laser displacement sensor adds auxiliary
Motion is helped, above two method can obtain huge point off density cloud of counting.It can also be obtained in addition by three-coordinates measuring machine
Point cloud, the point cloud positional precision that this method obtains is high, for sparse less cloud of counting.These equipment prices are expensive so that
Point cloud procurement cost is higher, with the appearance of consumer level products, such as Kinect so that point cloud is developed faster.
But the point cloud that the above method obtains can be comprising the environmental information residing for target object, it is necessary to be filtered, go to original point cloud
Make an uproar, split and the processing such as registration, could obtain the point cloud of object, post-processing complexity.
Point cloud mainly applies to reverse-engineering, carries out the three-dimensionalreconstruction of existing object, and digital information storehouse is created for it;
Field of machine vision, point cloud can be used for the directions such as target identification, target three-dimensional values and map structuring, vision provided for robot
Information, aid in its decision-making and path planning.
Robot technology has been widely used in the fields such as welding, spraying, but the grinding workpieces of complicated shape are main still
By manual polishing and numerical control machine tooling, the former damages workers ' health because polishing noise, dust are big, and to the skill of worker
Art level requirement is higher;The latter's machining accuracy is high, but process equipment is expensive, and flexibility is poor to limit its extensive use.
Industrial robot has multiple degrees of freedom, and flexibility is good, is relatively adapted to apply to polishing field.At present, using robot
Off-line programming technique, robot technology is successfully applied into complex part polishing, but off-line programming technique requires right in advance
Instrument and workpiece coordinate system are demarcated, and the clamping error due to workpiece and processing inconsistency, cause robot polishing one
When criticizing workpiece, there is the inconsistent situation of precision, or even waster occur.
Utility model content
To solve the problems, such as the acquisition of three-dimensional point cloud and a cloud being applied into robot polishing field, the utility model provides
A kind of real-time acquisition device and robot polishing method for planning track of three-dimensional point cloud based on laser displacement sensor.Design is not
The workpiece of similar shape is tested, and planning robot scans track, based on bus communication protocol and real time operating system, by total
Line communication mode realizes the real-time Communication for Power of robot and laser displacement sensor, obtains workpiece three-dimensional point cloud, for by three-dimensional point cloud
Apply to robot polishing trajectory planning and provide external condition.
To realize above-mentioned purpose of the present utility model, the utility model provides a kind of real-time collection dress of three-dimensional point cloud
Put, including robot, laser displacement sensor, the robot real-time control system based on real time operating system, the laser position
Displacement sensor is arranged on the end of the robot, the artificial six degree of freedom articulated robot of machine, the machine by fixture
Device people real-time control system connects robot and laser displacement sensor by real-time industrial ethernet bus, for making laser position
The reading of displacement sensor and the pose of robot obtain synchronization, and one-dimensional measurement is expanded into three-dimensional measurement, so as to be obtained by scanning
Take the three-dimensional point cloud of workpiece.
Further, it is Millisecond between described robot real-time control system and robot and laser displacement sensor
Communication.
Compared with prior art, the utility model makes laser position by robot real-time control system, EPA bus
The reading of displacement sensor and the pose of robot obtain synchronization, and one-dimensional measurement is expanded into three-dimensional measurement, so as to be obtained by scanning
Take the three-dimensional point cloud of workpiece.It is reliable and stable, precision is high, has real-time and flexibility characteristics, and cost simple in construction is low, Neng Gouman
The processing of sufficient different workpieces.
Brief description of the drawings
Fig. 1 is of the present utility model cloud real-time acquisition device schematic diagram;
Fig. 2 is caliberating device schematic diagram.
In figure:1- robots, 2- laser displacement sensors;3- robots real-time control system;4- workpiece;5- demarcation top
Point.
Embodiment
To further understand the utility model, the utility model is described further with reference to the accompanying drawings and examples,
It should be understood that the scope of the requires of the utility model protection is not limited to the scope of embodiment statement.
Embodiment
As shown in figure 1, a kind of real-time acquisition device of three-dimensional point cloud, including robot 1, laser displacement sensor 2, it is based on
The robot real-time control system 3 of real time operating system, the laser displacement sensor 2 are arranged on the robot by fixture
1 end, the robot 1 are six degree of freedom articulated robot, the robot real-time control system 3 by real-time industrial with
Too network bus connection robot 1 and laser displacement sensor 2, for making the reading of laser displacement sensor 2 and the position of robot 1
Appearance obtains synchronization, and one-dimensional measurement is expanded into three-dimensional measurement, so as to obtain the three-dimensional point cloud of workpiece 4 by scanning;And it is based on
The three-dimensional point cloud information of collection carries out robot polishing trajectory planning.
Described robot real-time control system is with being that Millisecond communicates between robot 1 and laser displacement sensor 2.
During laser scanning, six axle crossmachine people are controlled to move in a manner of teaching playback, with laser displacement sensor 2
Workpiece 4 is scanned, the point on workpiece 4 is gathered in real time based on the utility model device, finally gives based on robot basis coordinates system
Workpiece three-dimensional point cloud.Point cloud real-time acquisition device is as shown in Figure 1.
The robot real-time control system 3 of this device is built based on real time operating system and EPA bussing technique,
The x of workpiece point cloud, y location are led to from robot end's pose, the z location of workpiece point cloud from laser displacement sensor 2
Real-time bus mechanics of communication is crossed, ensures the synchronization of x, y, z, one-dimensional measurement is expanded into three-dimensional measurement, and the control so as to realize
The communication cycle of system meets requirement of real-time, can use the three-dimensional point cloud of device scanning generation workpiece up to 1ms.
The shown end clamping laser displacement sensor 2 of robot 1, laser displacement is read by analog input and output module and passed
The analog quantity output signals of sensor are simultaneously connected with the pulse signal of six controlled motors of robot, then total based on EPA
All signal acquisitions are returned to carry out data processing, communication cycle energy by the real time operating system of line technology by bus communication mode
Reach Millisecond, so as to realize the synchronization of the reading of the pose of robot and laser displacement sensor, by the pose of robot 1 and laser
The distance of displacement transducer 2 is solved by robot transformation equation, obtains the workpiece three-dimensional point cloud under robot basis coordinates system.
The present embodiment utilizes the workpiece of the different surface configuration of SolidWorks Software for Design, for sweep test and polishing
Experiment.This experimental design S-shaped curve surface work pieces, using aluminium alloy, and are processed, for test experiments.
A kind of robot polishing method for planning track based on described device, including step:
(1) by the measuring coordinate system of multi-point calibration method Calibration of Laser displacement transducer 2 relative to the ending coordinates of robot 1
The transition matrix of system, in order to which the point cloud under measuring coordinate system is transformed under robot basis coordinates system, Calibration of Laser position must be shifted to an earlier date
The measuring coordinate system of displacement sensor 2 is relative to the transition matrix of robot end's coordinate system, abbreviation hand and eye calibrating;
(2) robot 1 drives laser displacement sensor 2 to scan workpiece, realizes one-dimensional measurement expanding to three-dimensional measurement, obtain
Take the three-dimensional point cloud of workpiece;
(3) by setting three-dimensional point cloud described in different adaptive weighting function pairs to carry out normal direction to each point in three-dimensional point cloud
Estimation and optimization, while the position when spatial positional information included based on three-dimensional point cloud and the polishing of normal direction information planning robot 1
Appearance and feed path, and corrected continuous polishing track as robot 1 to scan track.
Specifically, the step (1) specifically includes:
(11) demarcation top 5 is installed in the end of robot 1 first, control machine people 1 reaches a certain fixing point in space,
It is relative that from teaching machine the fixing point is read using robot 1 tool coordinates system, 5 standardizations, after the completion of tool coordinates system demarcation
In the locus of the basis coordinates system of robot 1Bp;
(12) and then laser is radiated at described in the end clamping laser displacement sensor 2 of robot 1, control machine people 1
Fixing point, read pose of the current end of robot 1 relative to basis coordinates systemAnd the registration of laser displacement sensor 2MP, institute
State the registration of laser displacement sensor 2MP and the locusBP transformational relation is:
In formulaBP --- coordinates of the spatial point p in the basis coordinates system of robot 1;
MP --- coordinates of the spatial point p in measuring coordinate system;
--- ending coordinates system of robot 1 relative to basis coordinates system homogeneous transform matrix;
--- measuring coordinate system relative to ending coordinates system of robot 1 homogeneous transform matrix, wherein,
(13) formula (1) is converted toIfThen (1) can represent as follows:
(14) in order to reduce calibrated error, based on multi-point calibration, formula (3) is changed into AX=k form, one is formed and surpasses
Determine equation group, and solved using least square, form is as follows:
MakeMinimum solution X*As over-determined systems AX=k least square solution;
(15) matrix is obtained after solving XObtain matrixAfterwards, you can so that the workpiece three-dimensional point cloud of scanning to be turned
Change under the basis coordinates system of robot 1.
Specifically, in step (2), described the step of one-dimensional measurement is expanded into three-dimensional measurement is real by robot 1
When control system, the x of workpiece point cloud, y location passes from the end pose of robot 1, the z location of workpiece point cloud by laser displacement
Sensor 2 is understood, by real-time bus mechanics of communication, ensures the synchronization of x, y, z, one-dimensional measurement is expanded into three-dimensional survey so as to realize
Amount.
By said apparatus obtain point cloud whether can really reflect workpiece surface shape facility, can be used as subsequently
The polishing trajectory planning data of robot 1, can be verified by following methods.
When carrying out replication experiment, its three-dimensional CAD model known to the workpiece 4 processed can be discrete by three-dimensional CAD model
Into a cloud as standard point cloud, because standard point cloud is based on workpiece coordinate system, workpiece can be demarcated in advance relative to machine
The homogeneous transform matrix of the basis coordinates system of people 1, coordinate of some points relative to workpiece coordinate system is first marked on workpiece during demarcation, then
Control machine people 1 arrives corresponding points, obtains coordinate of this in the basis coordinates system of robot 1.Obtain at same o'clock in two coordinate systems
Under expression, then the transition matrix for calculating two coordinate systems is connect by svd algorithm, i.e. workpiece coordinate system and the basis coordinates of robot 1
The transition matrix of system.Standard point cloud is transformed under basis coordinates system by can after the completion of demarcation, then with scanning gained workpiece point cloud
Contrasted, that is, obtain existing error between scanning element cloud and standard point cloud, so as to verify the feasibility of the device.Due to machine
The presence of the position error of device people 1, the measurement error and calibrated error of laser displacement sensor 2, should allow scanning element cloud and standard
Error be present between point cloud, but this error should be controlled within zone of reasonableness.
Specifically, described step (3) specifically includes step:
(31) spatial positional information of each point based on three-dimensional point cloud and by adaptive weighting function optimization adjust after
Normal direction information planning robot 1 polish pose, it is described polishing pose planning dependent on three-dimensional point cloud collection accuracy and
The degree of accuracy of point cloud normal estimation, and finally determine the polishing quality of workpiece;
(32) the processing node of defining point cloud, and processing nodal information is sent to robot successively to scan sequencing
1, robot 1 carries out inverse kinematics and moving interpolation and will process one continuous robot 1 of node series connection generation polishing rail
Mark.
Specifically, described step (31) specifically includes step:
(301) position of cutter during polishing is determined, because containing the sky that scanning workpiece surface is each put in three-dimensional point cloud
Between positional information, and the positional information for putting cloud is based on the basis coordinates system of robot 1, so as to utilize each point institute in three-dimensional point cloud
Comprising the spatial positional information based on the basis coordinates system of robot 1;Pass through the knife of 1 resolved motion control robot of robot 1
Tool reaches position during polishing;
(302) posture of cutter during polishing is determined, that is, passes through the pre- of each point of adaptive weighting function pair three-dimensional point cloud
The posture of cutter when normal direction information planning robot 1 after estimating normal direction information and optimizing and revising polishes, therefore accurately point cloud method
Polishing quality to estimation to workpiece has material impact.
Specifically, described step (302) specifically includes step:
(311) the closest points of k are searched for by improved k Neighborhood-region-search algorithms, i.e., using a point as body-centered, with
Fixed step size builds the square of itself, and k neighborhood search is then carried out in square, if finding k closest points, i.e.,
Search is terminated, otherwise expands the scope of square with fixed step size, k closest points are re-searched for, until in square
Inside search k closest point, this avoid search for each neighborhood of a point when will travel through a little, improve search
Efficiency;And traditional k Neighborhood-region-search algorithms, be traversal point cloud in institute a little, calculate corresponding Euclidean distance, and carry out ascending order
Arrangement, k closest points are taken out as k neighborhoods;This algorithm is required for traversal point cloud when calculating each neighborhood of a point
In each point, take it is longer, do not apply to large-scale point cloud situation;
(312) improved PCA algorithms are used, different weighting functions is assigned by the point in the k neighborhoods to certain point, uses
Least square fitting method fits a plane, the initial normal direction using the normal direction of institute's fit Plane as the point;It is next based on institute
State initial normal direction and adaptively pinpoint the weighting function of each point in cloud really, then initial normal direction is optimized;Traditional PCA algorithms
It is in the k neighborhoods of certain point, using least square fitting method, to fit a plane in a cloud, the normal direction of fit Plane is made
For the normal direction of the point, due to noise, exterior point and sharp features inevitably be present in a cloud, so this step is to PCA algorithms
It is improved, different weighting functions is assigned to the point in neighborhood, so as to improve the quality of fit Plane, the final normal direction that improves is estimated
The precision of meter.
(313) put cloud to redirect, i.e., the normal direction after optimization is adjusted using minimum spanning tree method so that point cloud normal direction
Point to unanimously, minimum spanning tree method is:First provide that the normal direction of any is pointed to, diffusion path then generated according to the weight of each point,
It is final to be diffused into whole point cloud space, complete the adjustment of point cloud normal direction.
Specifically, described step (312) specifically includes step:
(321) plane that be fitted PCA algorithms accurately reflects the normal direction at the point, should try one's best in guarantee fitting neighborhood
Point is on same dough sheet, but conventional measurement means can make the space exterior point for carrying measured deviation and noise spot scanning be present
In point cloud, therefore two weighting functions are introduced into PCA algorithms approximating method is had necessarily to the exterior point in a cloud and noise
Robustness, least square fitting is represented by:
S.t. | | n | |=1 (5)
ω in formulad(xi) --- Gauss weighting function so that effect of the point near apart from the point to fit Plane is a little big,
Otherwise influence is small, ωd(xi) be expressed as:σdFor apart from bandwidth, its initial value is according to the shape of workpiece
And the distribution situation of scanning element cloud is preset;
ωr(ri) --- the Gauss weighting function relevant with regression criterion, ωr(ri) be expressed as:σrFor
Regression criterion bandwidth, its initial value are preset according to the shape of workpiece and the distribution situation of scanning element cloud;
--- the m times iteration, xiThe regression criterion of point,It is represented by:
(322) formula (5) construction covariance matrix C is solved, is represented as follows:
In formula--- the center of neighborhood is represented, is expressed as:
K --- the points in neighborhood are represented,
By solving covariance matrix C, characteristic vector corresponding to its minimum characteristic value is the normal direction n of fit Plane,
Initial normal direction i.e. as the point, and
(323) a characteristic coefficient F is defined for each point in point cloudcIt is positioned at smooth region, sharp spy to quantify the point
Sign region or transitional region, characteristic coefficient are represented by:
S=2 in formula, represent when ratio is 0.5, then to judge that the point is located at sharp features region;
δl--- represent the average distance of the point 6 point closest with it;
(324) characteristic coefficient F is determinedcValue, the point cloud cluster in first defining point cloud in a certain vertex neighborhood, when certain point cloud
Points in cluster are more than the business of the points and the quantity of point cloud cluster in neighborhood, then it is assumed that this cloud cluster is principal point cloud cluster;In order to certainly
The quantity of dynamic acquisition point cloud cluster, sets an angle threshold, when 2 points of normal direction angles are more than threshold value, that is, thinks at this 2 points
In different point cloud clusters;A quantity of cloud cluster is obtained, so as to calculate characteristic coefficient F corresponding to each pointc, 2 points of a's, b
Normal direction angle is represented by:
D (a, b)=cos-1(na,nb); (8)
(325) F is determinedcValue after, it is adaptive calculate each point apart from bandwidth σd, regression criterion bandwidth σr, normal direction
Difference bandwidth σn, represent as follows:
σd=(1+Fc)·δl (9)
σr=rmax/3Fc (10)
σn=| | ni-n||max/3Fc; (11)
(326) because workpiece inevitably has the sharp features such as edge and turning so that the point in neighborhood is in multiple
On dough sheet, so as to cause larger normal estimation error, therefore the 3rd weighting function is introduced in PCA algorithms, then a most young waiter in a wineshop or an inn
Multiply fitting to be expressed as:
S.t. | | n | |=1 (12)
ω in formulan(ni) --- the weighting function relevant with normal direction difference, it is expressed as:σnFor normal direction
Difference bandwidth, niFor certain initial normal direction of point, gained is solved by formula (6);
(327) each point calculated by adaptometer in step (325) apart from bandwidth σd, regression criterion bandwidth σr, method
To difference bandwidth σnThe weighting function each put in point cloud is calculated, is replaced in covariance matrix C
Then formula (12) is solved by the covariance matrix C after replacement, obtains the optimization normal direction of each point.
When being solved using above-mentioned steps, the weighting function that influences each point is apart from bandwidth σ respectivelyd, regression criterion bandwidth σr
With normal direction difference bandwidth σn, above-mentioned steps are first according to the shape of workpiece and the distribution situation of scanning element cloud, preset σdAnd σr's
Initial value, initial normal direction is calculated using formula (6), and adaptively pinpoint the bandwidth parameter of each point in cloud really based on initial normal direction, then
Initial normal direction is optimized using formula (7).
Traditional algorithm is the global bandwidth parameter of artificial setting, and it is all that experiment proves that this parameter does not adapt to
Point.The difference in the utility model region residing for sets different bandwidth parameters to complete normal estimation.Using adaptive
Mode, according to the neighborhood characteristics of each point in a cloud, to determine corresponding bandwidth parameter, so as to complete the normal estimation of a cloud.
Specifically, the point cloud normal direction estimated by above-mentioned steps, has ambiguity, i.e. normal direction can be both pointed in workpiece
Portion, outer workpiece can also be pointed to, so the utility model is adjusted using minimum spanning tree method to normal direction so that point cloud method
Consistent to pointing to, its principle is by x in the cloud that sets up an officei、xjIt is apart from close 2 points, both dot products is calculated, if ni·nj≈
1, then it represents that 2 points of normal orientation is identical, otherwise xiOr xjShould be reverse, therefore, the step (313) specifically includes step:
(331) the maximum point of z coordinate as root node, adjusts its normal direction and is allowed to and vector (0,0,1) in selected point cloud
Dot product is more than 0, and this normal orientation after making adjustment is pointed on the outside of workpiece;
(332) weight c=1- is set | ni·nj|, this ensure that the nonnegative weights, and represent two neighbor points when c is smaller
Normal orientation it is more parallel;
(333) traveled through according to weight c size, if father node niWith child node njMeet ni·nj<0, then by njInstead
To;
(334) all nodes are traveled through, normal orientation is subjected to Spreading and diffusion, a threshold value e is set in communication process, works as c
During≤e, multiple spot in One Diffusion Process to neighborhood, work as c>During e, one point of One Diffusion Process, after the completion of judge whether to meet c≤e, then enter
Row multiple spot spreads, until being diffused into whole point cloud space;
(335) if failing to travel through all points, point cloud does not connect, and partial data forms local isolated island, if do not traveled through
Count number less than point cloud sum an one thousandth, then give up as noise, otherwise search in the data set not traveled through with
Closest point between having traveled through a little, as the starting point for continuing traversal and go to step (334).
The utility model obtains its normal direction information by carrying out normal estimation to a cloud, then will include normal direction information and position
For the three-dimensional point of confidence breath as processing node, then by trajectory planning, x, y, z, Rx, Ry, Rz information of node are processed in generation, and
Node will be processed with the sequencing of scanning and be synchronously sent to robot 1 successively, robot 1 carries out moving interpolation and inverse motion
Learn and solve, generation polishing posture, so as to complete to scan the polishing of workpiece.
Above-mentioned steps are completed, that is, complete the three-dimensional point cloud collection based on the utility model device, and according only to three-dimensional
The information of point cloud generates the polishing of robot 1 track.
Above-described embodiment of the present utility model is only intended to clearly illustrate the utility model example, and is not
Restriction to embodiment of the present utility model.For those of ordinary skill in the field, on the basis of described above
On can also make other changes in different forms.There is no necessity and possibility to exhaust all the enbodiments.
All made within spirit of the present utility model and principle all any modification, equivalent and improvement etc., should be included in this reality
Within new scope of the claims.
Claims (2)
- A kind of 1. real-time acquisition device of three-dimensional point cloud, it is characterised in that:Including robot, laser displacement sensor, based on reality When operating system robot real-time control system, the laser displacement sensor is arranged on the end of the robot by fixture End, the artificial six degree of freedom articulated robot of described machine;The robot real-time control system passes through real-time industrial ethernet Bus connects robot and laser displacement sensor, for making the reading of laser displacement sensor and the pose of robot obtain together Step, three-dimensional measurement is expanded to by one-dimensional measurement, so as to obtain the three-dimensional point cloud of workpiece by scanning.
- 2. the real-time acquisition device of three-dimensional point cloud according to claim 1, it is characterised in that:Described robot is controlled in real time System processed is with being that Millisecond communicates between robot and laser displacement sensor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201720526994.XU CN206825428U (en) | 2017-05-12 | 2017-05-12 | A kind of real-time acquisition device of three-dimensional point cloud |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201720526994.XU CN206825428U (en) | 2017-05-12 | 2017-05-12 | A kind of real-time acquisition device of three-dimensional point cloud |
Publications (1)
Publication Number | Publication Date |
---|---|
CN206825428U true CN206825428U (en) | 2018-01-02 |
Family
ID=60771395
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201720526994.XU Expired - Fee Related CN206825428U (en) | 2017-05-12 | 2017-05-12 | A kind of real-time acquisition device of three-dimensional point cloud |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN206825428U (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107127755A (en) * | 2017-05-12 | 2017-09-05 | 华南理工大学 | A kind of real-time acquisition device and robot polishing method for planning track of three-dimensional point cloud |
CN108582076A (en) * | 2018-05-10 | 2018-09-28 | 武汉库柏特科技有限公司 | A kind of Robotic Hand-Eye Calibration method and device based on standard ball |
CN109605390A (en) * | 2018-12-28 | 2019-04-12 | 芜湖哈特机器人产业技术研究院有限公司 | A kind of automobile washing machine people's control system |
CN110794762A (en) * | 2019-11-06 | 2020-02-14 | 北京轩宇智能科技有限公司 | Control method and system for polishing robot |
-
2017
- 2017-05-12 CN CN201720526994.XU patent/CN206825428U/en not_active Expired - Fee Related
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107127755A (en) * | 2017-05-12 | 2017-09-05 | 华南理工大学 | A kind of real-time acquisition device and robot polishing method for planning track of three-dimensional point cloud |
CN107127755B (en) * | 2017-05-12 | 2023-12-08 | 华南理工大学 | Real-time acquisition device of three-dimensional point cloud and robot polishing track planning method |
CN108582076A (en) * | 2018-05-10 | 2018-09-28 | 武汉库柏特科技有限公司 | A kind of Robotic Hand-Eye Calibration method and device based on standard ball |
CN109605390A (en) * | 2018-12-28 | 2019-04-12 | 芜湖哈特机器人产业技术研究院有限公司 | A kind of automobile washing machine people's control system |
CN110794762A (en) * | 2019-11-06 | 2020-02-14 | 北京轩宇智能科技有限公司 | Control method and system for polishing robot |
CN110794762B (en) * | 2019-11-06 | 2021-02-05 | 北京轩宇智能科技有限公司 | Control method and system for polishing robot |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107127755A (en) | A kind of real-time acquisition device and robot polishing method for planning track of three-dimensional point cloud | |
CN206825428U (en) | A kind of real-time acquisition device of three-dimensional point cloud | |
CN102494657B (en) | Measuring head radius compensation method for curve surface profile measuring and detecting | |
CN110202573B (en) | Full-automatic hand-eye calibration and working plane calibration method and device | |
CN109683552B (en) | Numerical control machining path generation method on complex point cloud model guided by base curve | |
CN106354094A (en) | Method for calibrating machine tool follow-up laser scanning coordinates on basis of space standard balls | |
CN111055293A (en) | High-precision constant-force grinding method for industrial robot based on curved surface self-adaption | |
CN104515478A (en) | Automatic three-dimensional measuring method and automatic three-dimensional measuring system for high-precision blade of aviation engine | |
CN101497279A (en) | Measuring and machining integrated laser three-dimensional marking method and device | |
CN106091931A (en) | A kind of adaptive scanning based on threedimensional model measures system and control method thereof | |
CN110059879B (en) | Automatic planning method for three-coordinate measurement of vehicle body | |
CN108827155A (en) | A kind of robot vision measuring system and method | |
CN114011608B (en) | Spraying process optimization system based on digital twinning and spraying optimization method thereof | |
CN110069041A (en) | A kind of Work piece processing method and system based on on-machine measurement | |
CN101973032A (en) | Off-line programming system and method of optical visual sensor with linear structure for welding robot | |
CN108994830A (en) | System calibrating method for milling robot off-line programing | |
CN104457645A (en) | Robot tool central point calibration method using two-dimensional measurement functional tablet | |
CN103948196A (en) | Human body data measuring method | |
CN109767416A (en) | The positioning system and method for mechanical equipment | |
CN109883336A (en) | Measuring system and measurement method during a kind of sheet fabrication towards ship surface | |
CN107564065A (en) | The measuring method of man-machine minimum range under a kind of Collaborative environment | |
CN107972027A (en) | The localization method and device of robot, robot | |
CN103236043A (en) | Plant organ point cloud restoration method | |
CN114523475A (en) | Automatic error calibration and compensation device and method for robot assembly system | |
Son et al. | Path planning of multi-patched freeform surfaces for laser scanning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180102 Termination date: 20190512 |