CN108340371A - Target follows localization method and system a little - Google Patents

Target follows localization method and system a little Download PDF

Info

Publication number
CN108340371A
CN108340371A CN201810083785.1A CN201810083785A CN108340371A CN 108340371 A CN108340371 A CN 108340371A CN 201810083785 A CN201810083785 A CN 201810083785A CN 108340371 A CN108340371 A CN 108340371A
Authority
CN
China
Prior art keywords
target
coordinate
little
follows
millet cake
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201810083785.1A
Other languages
Chinese (zh)
Other versions
CN108340371B (en
Inventor
巫超
谈迎峰
张石
叶梦思
黄德立
李润泉
谭方杰
梁品聪
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.)
Zhimei Kangmin (Zhuhai) Health Technology Co., Ltd
Original Assignee
ZHUHAI JUNKAI MACHINERY TECHNOLOGY Co Ltd
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 ZHUHAI JUNKAI MACHINERY TECHNOLOGY Co Ltd filed Critical ZHUHAI JUNKAI MACHINERY TECHNOLOGY Co Ltd
Priority to CN201810083785.1A priority Critical patent/CN108340371B/en
Publication of CN108340371A publication Critical patent/CN108340371A/en
Application granted granted Critical
Publication of CN108340371B publication Critical patent/CN108340371B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1694Programme controls characterised by use of sensors other than normal servo-feedback from position, speed or acceleration sensors, perception control, multi-sensor controlled systems, sensor fusion
    • B25J9/1697Vision controlled systems
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J11/00Manipulators not otherwise provided for
    • B25J11/008Manipulators for service tasks
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1656Programme controls characterised by programming, planning systems for manipulators
    • B25J9/1664Programme controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning

Landscapes

  • Engineering & Computer Science (AREA)
  • Robotics (AREA)
  • Mechanical Engineering (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

Localization method and the system a little, this method is followed to include the present invention relates to a kind of target:It obtains the first point cloud data of the target object at the first moment and target follows the first coordinate a little, wherein point cloud data includes the coordinate of multiple appearance millet cake on target object, and target, which follows, to be a little located on target object;Each appearance millet cake, which is calculated separately, according to the first point cloud data and the first coordinate follows distance value a little to target;The second point cloud data for obtaining the target object at the second moment, according to the first point cloud data and the position offset of each appearance millet cake of second point cloud data acquisition;It follows the distance value selected, the position offset of each appearance millet cake and the first coordinate to obtain target according to each appearance millet cake to target and follows the second coordinate a little.Said program realizes that target follows accurate positionin a little, and accurately click-through line trace can be followed to target when target object position sends variation.

Description

Target follows localization method and system a little
Technical field
The present invention relates to technical field of computer vision, follows localization method a little more particularly to a kind of target and be System.
Background technology
Currently, target following is a popular research method of computer vision field, it is real-time to be applied to unmanned plane more The technical fields such as locating and tracking, video object detecting and tracking.But when the size of target following object and deviation post are smaller When, traditional target following technology is difficult to realize be accurately located target following object.
For example, as the improvement of people's living standards, the use of modality is more and more universal, modality automation, Intelligence is a main trend, and modality usually needs to position a certain acupuncture point of human body, and when acupuncture point shifts, physical therapy is set It is standby to be difficult to that the acupuncture point is accurately positioned again.
Invention content
Based on this, it is necessary to be directed to when the size of target following object and smaller deviation post, traditional target following Technology is difficult to realize the problem of being accurately located target object, provides a kind of target and follows localization method and system a little.
A kind of target follows localization method a little, includes the following steps:
It obtains the first point cloud data of the target object at the first moment and target follows the first coordinate a little, wherein institute The coordinate that point cloud data includes multiple appearance millet cake on the target object is stated, the target, which follows, is a little located at the target pair As upper;
According to first point cloud data and first coordinate calculate separately each appearance millet cake to the target with With the distance value of point;
The second point cloud data for obtaining the target object at the second moment, according to first point cloud data and described The position offset of each appearance millet cake of second point cloud data acquisition;
The distance value selected, the position offset of each appearance millet cake are followed according to each appearance millet cake to the target And first coordinate obtains the target and follows the second coordinate a little.
Above-mentioned target follows localization method a little, when by the point cloud data of the first moment target object of acquisition and second After the point cloud data for carving target object, the offset of former and later two moment target objects is calculated by point cloud data, to root Distance a little is followed to target according to the position offset and target object appearance millet cake of target object, target is estimated and follows a little In the coordinate position at the second moment, realize that target follows accurate positionin a little, it can be accurate when target object position sends and changes True follows click-through line trace to target.
It is described in one of the embodiments, that each institute is calculated separately according to first point cloud data and first coordinate The step of appearance millet cake follows distance value a little to the target is stated, is included the following steps:
If the target object is divided into according to the structure feature of the target object and first point cloud data Dry region, obtains the first nodal point coordinate of the regional center point in each region;
Each regional center point is calculated to the target according to each first nodal point coordinate and first coordinate Follow distance value a little;
The position according to first point cloud data and each appearance millet cake of the second point cloud data acquisition The step of offset, includes the following steps:
The target object is divided into several identical regions according to second point cloud data, and is obtained each described Second center point coordinate of the regional center point in region;
Each first nodal point coordinate and each second center point coordinate correspondence are subtracted each other, each region is obtained The position offset of central point;
It is described to follow the distance value selected, the position of each appearance millet cake inclined according to each appearance millet cake to the target Shifting amount and first coordinate obtain the step of target follows the second coordinate a little, include the following steps:
Follow distance value a little, the position of each regional center point inclined according to each regional center point to the target Shifting amount and first coordinate obtain the target and follow the second coordinate a little.
It is described in one of the embodiments, that distance value a little, each is followed according to each appearance millet cake to the target The position offset of the appearance millet cake and first coordinate obtain the step of target follows the second coordinate a little, packet Include following steps:
Each appearance millet cake to the target is followed to the position of the inverse ratio and each appearance millet cake of the distance value selected Offset, which corresponds to be multiplied, obtains corresponding target offset amount;
The target, which is obtained, according to each target offset amount and first coordinate follows the second coordinate a little.
It is described in one of the embodiments, that distance value a little, each is followed according to each appearance millet cake to the target The position offset of the appearance millet cake and first coordinate obtain the step of target follows the second coordinate a little, packet Include following steps:
The distance value selected, the position offset of each appearance millet cake are followed according to each appearance millet cake to the target And first coordinate calculates the target using following formula and follows the second coordinate a little
In formula,For the first coordinate of the target object;For the second coordinate of the target object; For the position offset of each appearance millet cake, whereinenx、eny、enzThe respectively described appearance millet cake Component of the position offset on x, y, z axis;Distance value a little is followed to the target for each appearance millet cake, Wherein,dnx、dny、dnzThe respectively described appearance millet cake to the target follow distance value a little x, Y, the component in z-axis.
In one of the embodiments, it is described obtain the first moment target object the first point cloud data and target with With point the first coordinate the step of, include the following steps:
The first point cloud data and the institute of the target object at first moment are obtained by binocular structure electro-optical device It states target and follows the first coordinate a little.
A kind of target follows positioning system a little, including:
Coordinate information acquisition module, the first point cloud data and target of the target object for obtaining for the first moment follow First coordinate of point, the point cloud data includes the coordinate of multiple appearance millet cake on the target object;
Distance value acquisition module, it is each described outer for being calculated separately according to first point cloud data and first coordinate Surface point follows distance value a little to the target, wherein the target, which follows, to be a little located on the target object;
Offset information acquisition module, the second point cloud data of the target object for obtaining for the second moment, according to institute State the position offset of the first point cloud data and each appearance millet cake of the second point cloud data acquisition;
Second coordinate obtaining module, for following distance value a little, each institute according to each appearance millet cake to the target It states the position offset of appearance millet cake and first coordinate obtains the target and follows the second coordinate a little.
Above-mentioned target follows positioning system a little, is obtained by coordinate information acquisition module and offset information acquisition module After the point cloud data of the point cloud data of first moment target object and the second moment target object, former and later two moment are calculated The offset of target object, the target to be calculated according to the position offset and distance value acquisition module of target object Objects looks millet cake follows distance a little to target, estimates target and follows coordinate position o'clock at the second moment, realizes target Accurate positionin a little is followed, can click-through line trace accurately be followed to target when target object position sends variation.
The second coordinate obtaining module includes target offset amount acquiring unit and the second coordinate in one of the embodiments, Computing unit;
The target offset amount acquiring unit, for each appearance millet cake to be followed distance value a little to the target Inverse ratio multiplication corresponding with each position offset of appearance millet cake obtains corresponding target offset amount;
Second coordinate calculating unit, for obtaining the mesh according to each target offset amount and first coordinate Mark follows the second coordinate a little.
Second coordinate obtaining module calculates the second coordinate using following formula in one of the embodiments,;
In formula,For the first coordinate of the target object;For the second coordinate of the target object; For the position offset of each appearance millet cake, whereinenx、eny、enzThe respectively described appearance millet cake Component of the position offset on x, y, z axis;Distance value a little is followed to the target for each appearance millet cake, Wherein,dnx、dny、dnzThe respectively described appearance millet cake to the target follow distance value a little x, Y, the component in z-axis.
A kind of mechanical arm, including robot body, moxibustion head, binocular add structure electro-optical device, controller and storage On the controller and the computer program that can be run on the controller;The binocular adds structure electro-optical device and the Chinese mugwort Moxibustion head is fixed on the robot body, and the binocular adds structure electro-optical device to be fixed with the moxibustion head relative position;
The binocular adds point cloud data and mesh of the structure electro-optical device for obtaining target object to follow the first coordinate a little Data;
It is realized when the controller executes the computer program as above-mentioned target follows localization method a little.
Above-mentioned mechanical arm realizes target and follows a little accurate by the computer program run on the controller Positioning, when target object position sends and changes, mechanical arm accurately can follow click-through line trace to target.
A kind of computer storage media, is stored thereon with computer program, is realized when which is executed by processor as above It states target and follows localization method a little.
Above computer storage medium realizes target and follows accurate positionin a little by the computer program of its storage, Can click-through line trace accurately be followed to target when target object position sends variation.
Description of the drawings
Fig. 1 is the flow chart that target follows localization method a little in one embodiment of the invention;
Fig. 2 is the flow chart that target follows localization method a little in another embodiment of the present invention;
Fig. 3 is the schematic diagram that target object is divided into several regions in one embodiment of the invention;
Fig. 4 is the structural schematic diagram that target follows positioning system a little in one embodiment of the invention;
Fig. 5 is the structural schematic diagram of coordinate information acquisition module in one embodiment of the invention.
Specific implementation mode
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to the accompanying drawings and embodiments, right The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and It is not intended to limit the scope of the invention.
Referring to Fig. 1, Fig. 1 is the flow chart that target follows localization method a little in one embodiment of the invention, the embodiment Middle target follows localization method a little, includes the following steps:
Step S110:It obtains the first point cloud data of the target object at the first moment and target follows a little first to sit Mark, wherein point cloud data includes the coordinate of multiple appearance millet cake on target object, and target, which follows, to be a little located on target object.
In this step, point cloud data is the set of one group of vector in a three-dimensional system of coordinate, and vector set can indicate The location information of the external surface shape of target object and target object in spatial distribution;Target, which follows, is a little located at target object On, it is a certain fixed point or the fixed position on target object;First coordinate is that target follows a little same with point cloud data One group in a three-dimensional system of coordinate vector represents target and follows location information a little.
Specifically, obtaining T1First point cloud data Q of moment target object1And target follow the first coordinate of object to Measure P1(x1, y1, z1), wherein the first point cloud data Q1It is the set of one group of vector, including the first multiple appearances of moment target object The coordinate vector of point on face.
Step S120:Appearance millet cake is calculated separately according to the first point cloud data with the first coordinate to follow at a distance from a little to target Value.
Pass through the coordinate vector of each appearance millet cake in the first point cloud data and the first coordinate vector P1(x1, y1, z1) calculate Each point follows distance a little to target on target object outer surface
Step S130:The second point cloud data for obtaining the target object at the second moment, according to the first point cloud data and Two point cloud datas obtain the position offset of appearance millet cake.
In this step, the second moment was a certain moment after the first moment, can be and the first time at intervals certain time At the time of, it can also the position of target object at the time of shift;Second point cloud data is used with the first point cloud data The same three-dimensional system of coordinate, it is same in the second point cloud data and the first point cloud data when the position of target object shifts The coordinate data of a appearance millet cake is different.
Step S140:According to each appearance millet cake to target follow the distance value selected, the position offset of each appearance millet cake with And first coordinate obtain target follow the second coordinate a little.
In above-described embodiment, the point cloud data of the first moment target object and the point cloud of the second moment target object are obtained After data, the offset of former and later two moment target objects is calculated by point cloud data, to according to the position of target object Offset and target object appearance millet cake follow distance a little to target, estimate target and follow o'clock coordinate at the second moment Position is realized that target follows accurate positionin a little, can accurately be followed a little to target when target object position sends variation Into line trace.
The point cloud data and mesh of target object can be obtained by binocular structure electro-optical device in one of the embodiments, Mark follows coordinate a little.
Step S140 follows distance value a little, each appearance according to each appearance millet cake to target in one of the embodiments, The position offset of millet cake and the first coordinate obtain the step of target follows the second coordinate a little, include the following steps:
Each appearance millet cake to target is followed into the inverse ratio for the distance value selected phase corresponding with each position offset of appearance millet cake Multiply and obtains corresponding target offset amount;Target, which is obtained, according to each target offset amount and the first coordinate follows the second coordinate a little.
In the present embodiment, the target that converts in proportion of distance value a little is followed to follow by target object appearance millet cake and target The offset target offset amount of point calculates acquisition target further according to target offset amount and the first coordinate and follows the second coordinate a little. Distance objective follows the proportion shared by the offset for selecting remoter appearance millet cake smaller, conversely, proportion is bigger, passes through target Offset acquires more accurately the second coordinate, is followed a little so as to be accurately tracked by target.
Further, distance value a little, each appearance are followed according to each appearance millet cake to target in one of the embodiments, The position offset of millet cake and the first coordinate obtain the step of target follows the second coordinate a little, include the following steps:
The distance value selected, the position offset and the first coordinate of each appearance millet cake are followed according to each appearance millet cake to target Target, which is calculated, using following formula (1) follows the second coordinate a little;
In formula,For the first coordinate of target object;For the second coordinate of target object;For each outer surface The position offset of point, whereinenx、eny、enzRespectively the position offset of appearance millet cake is in x, y, z axis On component;For each appearance millet cake distance value a little is followed to target, whereindnx、 dny、dnzRespectively appearance millet cake follows the component on x, y, z axis of distance value a little to target.
Specifically, each appearance millet cake to be followed to the inverse ratio of component of the distance value a little in x-axis to target, with each outer surface The position offset of point corresponds to be multiplied in the component of x-axis obtains the target offset amount of target object in the direction of the x axis, according to each mesh The target offset amount of mark object in the direction of the x axis and the x values of the first coordinate obtain the x values that target follows the second coordinate a little, lead to Cross y values and z values that same method obtains the second coordinate.
In view of point cloud data includes usually hundreds of thousands point up to a million, and operation is all carried out to all the points in point cloud data Processing often results in that processing operand is huge, and processing speed is greatly reduced, and difficulty of matching is big between point cloud data, for this purpose, can be with According to the structure feature of point cloud data and target object, the outer surface point of target object is divided into multiple regions, obtains this The central point in a little regions, using the equivalent point cloud data of center point coordinate, to reduce data processing operation amount and point cloud significantly Difficulty of matching between data improves processing speed so that can quickly be followed to target when target object position sends variation Point is repositioned.
Referring to Fig. 2, Fig. 2 is the flow chart that target follows localization method a little in another embodiment of the present invention.In this reality It applies in example, target follows localization method a little, includes the following steps:
Step S210:It obtains the first point cloud data of the target object at the first moment and target follows a little first to sit Mark, wherein point cloud data includes the coordinate of multiple appearance millet cake on target object, and target, which follows, to be a little located on target object.
In this step, T is obtained1First point cloud data Q of moment target object1And target follows the first coordinate of object Vectorial P1(x1, y1, z1), wherein the first point cloud data Q1It is the set of one group of vector, including the first moment target object is multiple outer The coordinate vector of point on surface.
Step S220:Target object is divided into several according to the structure feature of target object and the first point cloud data Region obtains the first nodal point coordinate of the regional center point in each region.
Point cloud data indicate target object external surface shape and target object spatial distribution location information, according to The outer surface point of target object is divided into multiple regions, for example, Fig. 3 is by the structure feature of point cloud data and target object Target object is divided into the schematic diagram in several regions, according to the structure of target object in Fig. 3 in one embodiment of the invention Feature, target object are divided into region 310, region 320, region 330, region 340, region 350 and region 360.
Regional center point is to be divided into multiple regions, physical centre's point in each region in target object;First center Point coordinates is in the first moment, coordinate position data of the regional center point in three-dimensional system of coordinate.
In this step, in T1Moment obtains the point cloud data Q of target object1Afterwards, according to the structure feature of target object and T1The point cloud data Q at moment1, target object is divided into n region, n>=2, and obtain first physical centre in each region Point q11、q12、…、q1n, and obtain the coordinate vector of these the first physical centre points.
Step S230:Each regional center point is calculated according to each first nodal point coordinate and the first coordinate to follow a little to target Distance value.
According to the physical centre in each region point q11、q12、…、q1nCoordinate vector and the first coordinate vector P1(x1, y1, z1), calculate separately these physical centre point q11、q12、…、q1nTo the first coordinate P1Distance
Step S240:The second point cloud data for obtaining the target object at the second moment, according to the second point cloud data by target Object is divided into several identical regions, and obtains the second center point coordinate of the regional center point in each region.
In this step, the second center point coordinate is in the second moment, coordinate bit of the regional center point in three-dimensional system of coordinate Set data.
Obtain T2Second point cloud data Q of moment target object2, the second point cloud data Q2Include in T2Moment target pair As the coordinate vector put on outer surface.After target object is divided into several identical regions, according to the second point cloud data Q2 Obtain the second physical centre point q in each region21、q22、…、q2n, and obtain the coordinate vector of these the second physical centre points.
Step S250:Each first nodal point coordinate and each second center point coordinate correspondence are subtracted each other, obtained in each region The position offset of heart point.
It is poor to make to the coordinate vector correspondence of physical centre's point of each region obtained twice, to obtain each region The position offset that physical centre's point is fastened in three-dimensional coordinate
Step S260:The position offset of distance value a little, each regional center point is followed according to each regional center point to target Amount and the first coordinate obtain target and follow the second coordinate a little.
In the present embodiment, according to the structure feature of point cloud data and target object, by the outer surface click and sweep of target object It is divided into multiple regions, and obtains physical centre's point in these regions and the coordinate of physical centre's point, using in several each physics The equivalent point cloud data of heart point coordinates calculates the offset of former and later two moment target objects by each physical centre's point coordinates, To follow distance a little to target according to the position offset and each physical centre point of target object, estimates target and follow Coordinate position o'clock at the second moment realizes that target follows accurate positionin a little, and significantly reduction data processing operation amount and Difficulty of matching between point cloud data improves processing speed so that can be quickly and accurately when target object position sends and changes Target is followed and is a little repositioned.
Further, in one of the embodiments, step S360 according to each regional center point to target follow a little away from It is obtained in the step of target follows the second coordinate a little from value, the position offset of each regional center point and the first coordinate, it can Target, which is calculated, using following formula (2) follows the second coordinate a little:
In formula,For the first coordinate of target object;For the second coordinate of target object;For each region The position offset of central point, whereinenx、eny、enzRespectively the position offset of regional center point x, Y, the component in z-axis;For each regional center point distance value a little is followed to target, whereindnx、dny、dnzRespectively each regional center point to target follow distance value a little on x, y, z axis Component.
In the present embodiment, the target that converts in proportion of distance value a little is followed to follow a little by each regional center point and target Offset target offset amount calculates acquisition target further according to target offset amount and the first coordinate and follows the second coordinate a little.Distance Target follows the proportion shared by the offset of a little remoter each regional center point smaller, conversely, proportion is bigger, passes through target Offset acquires more accurately the second coordinate, is followed a little so as to be accurately tracked by target.
Localization method a little is followed according to above-mentioned target, the present invention also provides a kind of targets to follow positioning system a little, with The target of the lower just present invention follows the embodiment of positioning system to be a little described in detail.
Referring to Fig. 4, Fig. 4 is the structural schematic diagram that target follows positioning system a little in one embodiment of the invention.This reality It applies in example, target follows positioning system a little, including:
Coordinate information acquisition module 410, the first point cloud data and target of the target object for obtaining for the first moment Follow the first coordinate a little, wherein point cloud data includes the coordinate of multiple appearance millet cake on target object, and target follows point In on target object;
Distance value acquisition module 420 is arrived for calculating separately each appearance millet cake with the first coordinate according to the first point cloud data Target follows distance value a little;
Offset information acquisition module 430, the second point cloud data of the target object for obtaining for the second moment, according to first The position offset of point cloud data and each appearance millet cake of second point cloud data acquisition;
Second coordinate obtaining module 440, for following the distance value selected, each appearance millet cake according to each appearance millet cake to target Position offset and the first coordinate obtain target follow the second coordinate a little.
Above-mentioned target follows positioning system a little, passes through coordinate information acquisition module 410 and offset information acquisition module After 430 obtain the point cloud data of the point cloud data of the first moment target object and the second moment target object, calculate front and back The offset of two moment target objects, to be counted according to the position offset and distance value acquisition module 420 of target object Obtained target object appearance millet cake follows distance a little to target, estimates target and follows o'clock coordinate bit at the second moment It sets, realizes that target follows accurate positionin a little, can click-through accurately be followed to target when target object position sends variation Line trace.
Distance value acquisition module 420 is according to the structure feature of target object and first point in one of the embodiments, Target object is divided into several regions by cloud data, obtains the first nodal point coordinate of the regional center point in each region;According to Each first nodal point coordinate and the first coordinate calculate each regional center point and follow distance value a little to target;
Target object is divided into several identical regions by offset information acquisition module 430 according to the second point cloud data, And obtain the second center point coordinate of the regional center point in each region;Each first nodal point coordinate and each second central point are sat Mark correspondence is subtracted each other, and the position offset of each regional center point is obtained.
Second coordinate obtaining module 440 follows distance value, each regional center point a little according to each regional center point to target Position offset and the first coordinate obtain target follow the second coordinate a little
It is the structural schematic diagram of coordinate information acquisition module in one embodiment of the invention referring to Fig. 5, Fig. 5.The present embodiment In, the second coordinate obtaining module 440 includes target offset amount acquiring unit 441 and the second coordinate calculating unit 442;
Target offset amount acquiring unit 441, for each appearance millet cake to target is followed the inverse ratio of distance value a little with it is each The position offset of appearance millet cake, which corresponds to be multiplied, obtains corresponding target offset amount;
Second coordinate calculating unit 442, for according to each target offset amount and the first coordinate obtain target follow a little the Two coordinates.
The second coordinate obtaining module 440 calculates the second coordinate using following formula (3) in one of the embodiments,;
In formula,For the first coordinate of target object;For the second coordinate of target object;For each appearance The position offset of millet cake, whereinenx、eny、enzRespectively the position offset of appearance millet cake is in x, y, z Component on axis;For each appearance millet cake distance value a little is followed to target, wherein dnx、dny、dnzRespectively appearance millet cake follows the component on x, y, z axis of distance value a little to target.
The target of the present invention follows the target of positioning system and the present invention a little that localization method a little is followed to correspond, Above-mentioned target follows the technical characteristic that the embodiment of localization method a little illustrates and advantage to be followed a little suitable for target Positioning system embodiment in, hereby give notice that.
In one of the embodiments, the present invention also provides a kind of mechanical arm, including robot body, moxibustion head, The computer program that binocular adds structure electro-optical device, controller and storage on the controller and can run on the controller;Binocular Structure electro-optical device is added to be fixed on robot body with moxibustion head, and binocular adds structure electro-optical device solid with moxibustion head relative position It is fixed;
Binocular adds point cloud data and mesh of the structure electro-optical device for obtaining target object to follow the first coordinate data a little;
Controller realizes that the target of any of the above-described embodiment follows localization method a little when executing computer program.
Specifically, when target object moves larger amplitude, target follows point location controller by realizing above-mentioned The target of one embodiment follows localization method a little, obtains the target after target object movement and follows coordinate information a little, simultaneously Control machinery arm moves so that moxibustion head can follow to target and follow new position a little automatically, realize that target follows a little oneself Dynamic to follow, the mechanical arm is simple in structure, and performance is stablized, and can accurately be followed after target object moves and be moved to target Follow new position a little.
In addition, one of ordinary skill in the art will appreciate that realize above-described embodiment method in all or part of flow, It is that relevant hardware can be instructed to complete by computer program, which can be stored in a non-volatile computer can It reads in storage medium, in the embodiment of the present invention, which can be stored in the storage medium of computer system, and by the meter At least one of calculation machine system processor executes, and includes that above-mentioned each target such as follows the embodiment of localization method a little to realize Flow.
A kind of storage medium is also provided in one of the embodiments, is stored thereon with computer program, wherein the journey Realize that any one target in the various embodiments described above such as follows localization method a little when sequence is executed by processor.Wherein, it stores Medium can be magnetic disc, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access Memory, RAM) etc..
The computer storage media, the computer program of storage include as above-mentioned each target follows a little by realizing The flow of the embodiment of localization method, to realize that target follows accurate positionin a little, when target object position sends and changes Can click-through line trace accurately be followed to target.
Each technical characteristic of above example can be combined arbitrarily, to keep description succinct, not to above-described embodiment In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance Shield is all considered to be the range of this specification record.
Only several embodiments of the present invention are expressed for above example, the description thereof is more specific and detailed, but can not Therefore it is construed as limiting the scope of the patent.It should be pointed out that for those of ordinary skill in the art, Under the premise of not departing from present inventive concept, various modifications and improvements can be made, these are all within the scope of protection of the present invention. Therefore, the protection domain of patent of the present invention should be determined by the appended claims.

Claims (10)

1. a kind of target follows localization method a little, which is characterized in that include the following steps:
It obtains the first point cloud data of the target object at the first moment and target follows the first coordinate a little, wherein the point Cloud data include the coordinate of multiple appearance millet cake on the target object, and the target, which follows, is a little located at the target object On;
Each appearance millet cake is calculated separately according to first point cloud data and first coordinate to follow a little to the target Distance value;
The second point cloud data for obtaining the target object at the second moment, according to first point cloud data and described second Point cloud data obtains the position offset of each appearance millet cake;
According to each appearance millet cake to the target follow the distance value selected, the position offset of each appearance millet cake and First coordinate obtains the target and follows the second coordinate a little.
2. target according to claim 1 follows localization method a little, which is characterized in that described according to first cloud Data and first coordinate calculate separately the step of each appearance millet cake follows distance value a little to the target, including with Lower step:
The target object is divided into several according to the structure feature of the target object and first point cloud data Region obtains the first nodal point coordinate of the regional center point in each region;
Each regional center point is calculated according to each first nodal point coordinate and first coordinate to follow to the target The distance value of point;
The position offset according to first point cloud data and each appearance millet cake of the second point cloud data acquisition The step of amount, includes the following steps:
The target object is divided into several identical regions according to second point cloud data, and obtains each region Regional center point the second center point coordinate;
Each first nodal point coordinate and each second center point coordinate correspondence are subtracted each other, each regional center is obtained The position offset of point;
It is described that the distance value selected, the position offset of each appearance millet cake are followed according to each appearance millet cake to the target And first coordinate obtains the step of target follows the second coordinate a little, includes the following steps:
The position offset of distance value a little, each regional center point is followed according to each regional center point to the target And first coordinate obtains the target and follows the second coordinate a little.
3. target according to claim 1 follows localization method a little, which is characterized in that described according to each outer surface The target is selected to follow described in the distance value selected, the position offset of each appearance millet cake and first coordinate acquisition Target follows the step of the second coordinate a little, includes the following steps:
Each appearance millet cake to the target is followed to the position offset of the inverse ratio and each appearance millet cake of the distance value selected Corresponding be multiplied of amount obtains corresponding target offset amount;
The target, which is obtained, according to each target offset amount and first coordinate follows the second coordinate a little.
4. target according to claim 3 follows localization method a little, which is characterized in that described according to each outer surface The target is selected to follow described in the distance value selected, the position offset of each appearance millet cake and first coordinate acquisition Target follows the step of the second coordinate a little, includes the following steps:
According to each appearance millet cake to the target follow the distance value selected, the position offset of each appearance millet cake and First coordinate calculates the target using following formula and follows the second coordinate a little;
In formula,For the first coordinate of the target object;For the second coordinate of the target object;For each institute State the position offset of appearance millet cake, whereinenx、eny、enzThe position of the respectively described appearance millet cake is inclined Move the component on x, y, z axis;For each appearance millet cake distance value a little is followed to the target, whereindnx、dny、dnzThe respectively described appearance millet cake to the target follow distance value a little in x, y, z axis On component.
5. target according to claim 1 follows localization method a little, which is characterized in that the mesh for obtaining for the first moment The step of the first point cloud data and target of mark object follow the first coordinate a little, includes the following steps:
The first point cloud data of the target object at first moment and the mesh are obtained by binocular structure electro-optical device Mark follows the first coordinate a little.
6. a kind of target follows positioning system a little, which is characterized in that including:
Coordinate information acquisition module, the first point cloud data and target of the target object for obtaining for the first moment follow a little First coordinate, the point cloud data include the coordinate of multiple appearance millet cake on the target object;
Distance value acquisition module, for calculating separately each outer surface according to first point cloud data and first coordinate Point follows distance value a little to the target, wherein the target, which follows, to be a little located on the target object;
Offset information acquisition module, the second point cloud data of the target object for obtaining for the second moment, according to described the The position offset of one point cloud data and each appearance millet cake of the second point cloud data acquisition;
Second coordinate obtaining module, for following distance value a little, each described outer according to each appearance millet cake to the target The position offset of surface point and first coordinate obtain the target and follow the second coordinate a little.
7. target according to claim 6 follows positioning system a little, which is characterized in that the second coordinate obtaining module includes Target offset amount acquiring unit and the second coordinate calculating unit;
The target offset amount acquiring unit, the inverse ratio for each appearance millet cake to be followed to distance value a little to the target Multiplication corresponding with each position offset of appearance millet cake obtains corresponding target offset amount;
Second coordinate calculating unit, for according to each target offset amount and first coordinate obtain the target with With the second coordinate of point.
8. target according to claim 7 follows positioning system a little, which is characterized in that second coordinate obtaining module The second coordinate is calculated using following formula;
In formula,For the first coordinate of the target object;For the second coordinate of the target object;For each institute State the position offset of appearance millet cake, whereinenx、eny、enzThe position of the respectively described appearance millet cake is inclined Move the component on x, y, z axis;For each appearance millet cake distance value a little is followed to the target, whereindnx、dny、dnzThe respectively described appearance millet cake to the target follow distance value a little in x, y, z axis On component.
9. a kind of mechanical arm, which is characterized in that add structure electro-optical device, controller including robot body, moxibustion head, binocular And it is stored in the computer program that can be run on the controller and on the controller;The binocular adds structure electro-optical device It is fixed on the robot body with the moxibustion head, and the binocular adds structure electro-optical device position opposite with the moxibustion head Set fixation;
The binocular adds point cloud data and mesh of the structure electro-optical device for obtaining target object to follow the first coordinate data a little;
Realize that the target as described in claim 1 to 5 any one follows a little when the controller executes the computer program Localization method.
10. a kind of computer storage media, is stored thereon with computer program, which is characterized in that the program is executed by processor Targets of the Shi Shixian as described in claim 1 to 5 any one follows localization method a little.
CN201810083785.1A 2018-01-29 2018-01-29 Target following point positioning method and system Active CN108340371B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810083785.1A CN108340371B (en) 2018-01-29 2018-01-29 Target following point positioning method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810083785.1A CN108340371B (en) 2018-01-29 2018-01-29 Target following point positioning method and system

Publications (2)

Publication Number Publication Date
CN108340371A true CN108340371A (en) 2018-07-31
CN108340371B CN108340371B (en) 2020-04-21

Family

ID=62960973

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810083785.1A Active CN108340371B (en) 2018-01-29 2018-01-29 Target following point positioning method and system

Country Status (1)

Country Link
CN (1) CN108340371B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110298889A (en) * 2019-06-13 2019-10-01 高新兴科技集团股份有限公司 A kind of video tab adjusting method, system and equipment
CN112509055A (en) * 2020-11-20 2021-03-16 浙江大学 Acupuncture point positioning system and method based on combination of binocular vision and coded structured light
CN112775955A (en) * 2019-11-06 2021-05-11 深圳富泰宏精密工业有限公司 Mechanical arm coordinate determination method and computer device
CN113171913A (en) * 2021-04-30 2021-07-27 哈尔滨工业大学 Spraying path generation method based on three-dimensional point cloud of seat furniture

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103630138A (en) * 2013-12-09 2014-03-12 天津工业大学 Unmanned aerial vehicle visual navigation method based on camera head calibration algorithm
CN104346811A (en) * 2014-09-30 2015-02-11 深圳市华尊科技有限公司 Video-image-based target real-time tracking method and device
CN104820978A (en) * 2015-01-28 2015-08-05 山西迪迈沃科光电工业有限公司 Reference positioning method of CCD (Charge Coupled Device) cameras
JP2015179453A (en) * 2014-03-19 2015-10-08 富士通株式会社 Positioning device, positioning method, and positioning computer program
CN107102647A (en) * 2017-03-30 2017-08-29 中国人民解放军海军航空工程学院青岛校区 Unmanned plane target tracking and controlling method based on image

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103630138A (en) * 2013-12-09 2014-03-12 天津工业大学 Unmanned aerial vehicle visual navigation method based on camera head calibration algorithm
JP2015179453A (en) * 2014-03-19 2015-10-08 富士通株式会社 Positioning device, positioning method, and positioning computer program
CN104346811A (en) * 2014-09-30 2015-02-11 深圳市华尊科技有限公司 Video-image-based target real-time tracking method and device
CN104820978A (en) * 2015-01-28 2015-08-05 山西迪迈沃科光电工业有限公司 Reference positioning method of CCD (Charge Coupled Device) cameras
CN107102647A (en) * 2017-03-30 2017-08-29 中国人民解放军海军航空工程学院青岛校区 Unmanned plane target tracking and controlling method based on image

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110298889A (en) * 2019-06-13 2019-10-01 高新兴科技集团股份有限公司 A kind of video tab adjusting method, system and equipment
CN112775955A (en) * 2019-11-06 2021-05-11 深圳富泰宏精密工业有限公司 Mechanical arm coordinate determination method and computer device
CN112775955B (en) * 2019-11-06 2022-02-11 深圳富泰宏精密工业有限公司 Mechanical arm coordinate determination method and computer device
CN112509055A (en) * 2020-11-20 2021-03-16 浙江大学 Acupuncture point positioning system and method based on combination of binocular vision and coded structured light
CN112509055B (en) * 2020-11-20 2022-05-03 浙江大学 Acupuncture point positioning system and method based on combination of binocular vision and coded structured light
CN113171913A (en) * 2021-04-30 2021-07-27 哈尔滨工业大学 Spraying path generation method based on three-dimensional point cloud of seat furniture
CN113171913B (en) * 2021-04-30 2022-04-22 哈尔滨工业大学 Spraying path generation method based on three-dimensional point cloud of seat furniture

Also Published As

Publication number Publication date
CN108340371B (en) 2020-04-21

Similar Documents

Publication Publication Date Title
CN108340371A (en) Target follows localization method and system a little
Nakai et al. Prediction of basketball free throw shooting by openpose
Lien et al. Model-based articulated hand motion tracking for gesture recognition
US9001129B2 (en) Processing apparatus and method for creating avatar
CN112525197B (en) Ultra-wideband inertial navigation fusion pose estimation method based on graph optimization algorithm
Zhang Application of intelligent virtual reality technology in college art creation and design teaching
CN106774901B (en) Remote PC body-sensing input method based on localization by ultrasonic
CN108427595A (en) The determination method and device of user interface controls display location in virtual reality
Shmuel et al. Active vision: 3d from an image sequence
JP2004265222A (en) Interface method, system, and program
CN110345936A (en) The track data processing method and its processing system of telecontrol equipment
CN101162525A (en) Human body multiple arthrosises characteristic tracking method based on shift Mean Shift and artificial fish school intelligent optimizing
Schröder et al. Design and evaluation of reduced marker layouts for hand motion capture
Jo et al. Tracking and interaction based on hybrid sensing for virtual environments
Herranz et al. Camera pose estimation using particle filters
Brückner et al. Evaluation of inertial sensor fusion algorithms in grasping tasks using real input data: Comparison of computational costs and root mean square error
Singh et al. Navigation and Speed Regulation Aimed at Travel through Immersive Virtual Environments: A Review
Luo et al. Adaptively adjusted EKF-based magnet tracking method for fast-moving object
CN108733211A (en) Tracing system, its operating method, controller and computer-readable recording medium
CN109343074A (en) Mobile humanoid robot and its map constructing method, device and readable storage medium storing program for executing
Lu et al. Binocular Vision‐Based Recognition Method for Table Tennis Motion Trajectory
Ning et al. MoiréTracker: Continuous Camera-to-Screen 6-DoF Pose Tracking based on Moiré Pattern
CN110263702A (en) A kind of real-time three-dimensional gesture method for tracing based on method of geometry
Wan et al. Research on Robot Monocular Vision-Based 6DOF Object Positioning and Grasping Approach Combined With Image Generation Technology
Cao et al. Perceptual MR Space: interactive toolkit for efficient environment reconstruction in mobile mixed reality

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
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20180831

Address after: 519000 A 2, two floor, No. 10 Futian Road, Xiangzhou District, Zhuhai, Guangdong.

Applicant after: Zhuhai Wanao te Health Technology Co., Ltd.

Address before: 519000 Zhuhai, Xiangzhou, Guangdong Futian Road, No. 10 plant 1 1 -3, 2, two factory floor

Applicant before: Zhuhai Junkai Machinery Technology Co., Ltd.

GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20201123

Address after: Room 1001, unit 3, building 1, No. 145, Mingzhu North Road, Xiangzhou District, Zhuhai City, Guangdong Province

Patentee after: Zhimei Kangmin (Zhuhai) Health Technology Co., Ltd

Address before: 519000 A 2, two floor, No. 10 Futian Road, Xiangzhou District, Zhuhai, Guangdong.

Patentee before: ZHUHAI WANNAOTE HEALTH TECHNOLOGY Co.,Ltd.