CN105082158B - Method for grabbing and placing compressor based on image recognition - Google Patents
Method for grabbing and placing compressor based on image recognition Download PDFInfo
- Publication number
- CN105082158B CN105082158B CN201510437176.8A CN201510437176A CN105082158B CN 105082158 B CN105082158 B CN 105082158B CN 201510437176 A CN201510437176 A CN 201510437176A CN 105082158 B CN105082158 B CN 105082158B
- Authority
- CN
- China
- Prior art keywords
- compressor
- axis
- camera
- pallet
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention relates to a method for grabbing and placing a compressor based on image recognition and belongs to the robot technical field. The device is implemented by the aid of a system for grabbing and placing a compressor. The system comprise an X axis movement system, a Y axis movement system and a Z axis movement system arranged on a support frame and a manipulator arranged at the bottom of a vertical beam in the direction of a Z axis. One side of the manipulator is provided with a first camera. A second camera is arranged on a tray used for placing the compressor. The method for grabbing and placing mainly comprises the following steps that firstly, a photograph of the tray is taken, and the center coordinates of a tray placing point and the angle of the tray are obtained through calculation; then, a photograph of the compressor is taken to obtain the position of the compressor, the compressor is grabbed according to the obtained position, a photograph of the bottom of the compressor is taken, and the final placing position is worked out; and finally, the manipulator places the compressor in a target position accurately. According to the system and method for grabbing and placing the compressor based on image recognition, intelligent grabbing and placing can be conducted; the labor amount is reduced, and meanwhile, the work efficiency is improved.
Description
Technical field
The invention belongs to robotics, the method being specifically related to capture with place compressor, especially
It is the method captured and place compressor based on image recognition.
Background technology
In industrial production environment, gripper of manipulator is utilized to take the work being placed on workbench or conveyer belt
Part, has been sorting or the basic operation of fittage.In air-conditioning production line, by workbench
Compressor captures, and is assembled to compressor carry on the pallet of compressor, if this operation
Using artificial realization, not only labor intensity is big, and post mobility of people is strong, and work quality can not
Ensure, it may appear that the situation damaging compressor occurs, use programme-control machine people to capture and put
The standardization of work can be accomplished when putting, but this method comparison program, it is impossible to according to practical situation
Be adjusted, need artificial supervision, it may appear that sky is grabbed and vacant phenomenon, and capture occurring or
When placing information change, needing to be changed program, this just have impact on being smoothed out of production procedure,
Affecting work efficiency, therefore, urgent need one can be made corresponding adjustment according to practical situation thus be realized
Accurately capture and place.
Summary of the invention
Method in the present invention employs collected by camera compressor position and the angle figure of two same model
As information, and pass through location identification module and the position of compressor drop target in compressor home position
Image information is processed and captures compressor by identification module, then calculates and captures the former of compressor
Beginning position, the target location placing compressor and RA, for industrial improved efficiency
And alleviate in hand labor intensity significant.
The concrete technical scheme of the present invention is:
The system captured and place compressor based on image recognition, is used for capturing compressor and being put
Being placed on pallet, system includes control unit module, the mechanical hand driven by drive mechanism, closes
Key point is, is provided with for shooting compressor home position and the first of target placement location in this system
Camera, described control unit module includes that CPU and CPU carries out the compressor raw bits of signal transmission
The location identification module put and the location identification module of compressor drop target, the image letter of first camera
The transmitting terminal of breath respectively with location identification module and the position of compressor drop target in compressor home position
The signal receiving end putting identification module connects, the location identification module in compressor home position and compressor
The signal output part of the location identification module of drop target is connected with CPU, the control signal outfan of CPU
It is connected with the drive end of drive mechanism.
Described capture with place the system of compressor is additionally provided with for shoot compressor angle the
The location identification module of two cameras, the transmitting terminal of the image information of second camera and compressor drop target
Signal receiving end connect.
Described drive mechanism is X-axis, Y-axis, Z axis linear motion system and the Z being arranged on bracing frame
The four-axle linked motor system that axle rotary system is constituted, X-axis linear motion system includes being arranged on bracing frame
The parallel X-axis guide rail of upper end, X-axis guide rail is the Timing Belt around the axis of rolling, and X-axis guide rail upper end is fixed
Having crossbeam, axis of rolling one end connects X-axis motor and X-axis decelerator, is provided with along crossbeam side in crossbeam
To Y-axis leading screw, crossbeam side is provided with Z axis and fixes seat, and Z axis is fixed seat and is provided with Y-axis swivel nut and Y-axis
Leading screw matches, and Y-axis leading screw is rotated by the y-axis motor and Y-axis decelerator being arranged on crossbeam one end
Driving, Z axis is fixed seat opposite side and is provided with the Z axis guide block swivel nut and vertical beam being slidably matched, sets in vertical beam
Being equipped with Z axis leading screw to be slidably matched with Z axis swivel nut, vertical beam upper end is provided with Z axis motor and Z axis decelerator drives
Dynamic Z axis screw turns, Z axis motor upper end connects brake, and mechanical hand is subtracted by Z axis motor and Z axis
Speed device is connected to the lower end of vertical beam.
Described first camera and second camera are the camera of same model.
Described first camera is fixed on the lower end of vertical beam, and described second camera is fixed on pallet side.
System based on compressor crawl and placement is compressed machine crawl and includes following with the method placed
Step:
Step S1: image information is also sent to compressor drop target by first camera shooting pallet image
Location identification module, the location identification module of compressor drop target obtains pallet by image information
In the angle of coordinate and pallet at center of triangle of three set-points composition；
Step S2: first camera shooting is positioned at the compressor map picture in home position and image information is sent
To the location identification module in compressor home position, the location identification module in compressor home position is passed through
Image information obtains the original position-information of compressor and sends to CPU, and then CPU controls drive mechanism
Take compressor by gripper of manipulator and be delivered to above pallet；
Step S3: second camera shooting compressor ground plan picture the position sent to compressor drop target
Putting identification module, the location identification module of compressor drop target is compressed according to footprint image information
The angle information of machine also sends to CPU, CPU and drives machine according to the pallet Angle ambiguity obtained in step S1
Compressor is rotated by structure by mechanical hand so that compressor angle is identical with pallet angle, and root
3 points according to the pallet that triangle center coordinate and step S1 of compressor bottom mounting apertures composition obtain
Centre coordinate calculates the final placement location of compressor.
Step S4:CPU controls drive mechanism and is placed on by compressor and finally places position obtained by step S3
Put.
In described step S1, the location identification module of compressor drop target obtains tray position and angle
The step of information includes:
Step S11:CPU controls drive mechanism and moves to above pallet by first camera, shoots pallet figure
Picture；
Step S12: pallet image is carried out gradation conversion and generates gray level image, gradation conversion method formula
Representing gray level image for gray=1*R+0*G+0*B, gray, R, G, B represent red respectively, green,
The brightness of blue three color light；
Step S13: gray level image is carried out template matching (process of template matching), find three convex
Play the approximate location of set-point, limit the certain limit around each set-point as effective coverage；
Step S14: respectively three effective coverages are carried out binaryzation, and processed by binary morphology,
Obtain the center position of each protruding set-point；
Step S15: connect three central points and obtain triangle, calculate the center point coordinate and three of triangle
2 lines that in point, x coordinate is bigger and horizontal angle；
Step S16: calculate the distance of triangle center point coordinates and picture centre, and be transformed into world's seat
In mark system (disX0, disY0)；
Step S17: according to mechanical hand present co-ordinate position (xp1, yp1), and step S16 obtains
Distance, mechanical hand when calculating the triangle center of 3 compositions and first camera center superposition on pallet
Coordinate (x1, y1), formula is: x1=xp1+disX0, y1=yp1+disY0.
The location identification module in described compressor home position obtains the step bag in compressor home position
Include:
Step S21: first camera moved to above compressor, shoots compressor photo；
Step S22: compressor photo carrying out gradation conversion and generates gray level image, gradation conversion method is public
Formula is that gray=1*R+0*G+0*B, gray represent gray level image, and R, G, B represent red respectively, green,
The brightness of blue three color light；
Step S23: gray level image is carried out template matching (template and the details of coupling thereof), looks for
To matched position；
Step S24: calculate the distance of matched position and picture centre, and be transformed into world coordinate system
(disX1,disY1)；
Step S25: calculate the distance (disX2, disY2) of first camera and mechanical hand center；
Step S26: the distance calculated according to Current mechanical hand position coordinate (xp4, yp4), step S24
First camera and the distance at mechanical hand center in (disX1, disY1) and step S25
(disX2, disY2), (x, y), formula is: x=to obtain the accurate location of compressor
Xp4+disX1+disX2, y=yp4+disY1+disY2, CPU control drive mechanism mobile manipulator and enter
Row compressor captures and moves above pallet.
In described step S3, the location identification module of compressor drop target obtains compressor angle and
The step of whole placement location includes:
Step S31: first camera is moved to the picture centre of the shooting of second camera, note machine now
Tool hands coordinate (x2, y2)；
Step S32: move to above second camera by the compressor grabbed, shines bottom shooting compressor
Sheet；
Step S33: compressor bottom image is carried out gradation conversion and generates gray level image, gradation conversion side
Method formula is that gray=1*R+0*G+0*B, gray represent gray level image, and R, G, B represent red respectively,
Green, the brightness of blue three color light；
Step S34: gray level image is carried out template matching, finds three installing holes；
Step S35: according to the position of template matching, the border in location and installation hole, pass through method of least square
Determine the center of circle and the radius of more accurate installing hole；
Step S36: calculate line and the horizontal line angle of 3 middle x coordinate bigger 2, press from both sides according to this
Angle rotates with the pallet angle of acquisition in step S1 so that rotate latter two angle identical；
Step S37: use the method in step S35 to recalculate three installing holes postrotational image
The center of circle, connect three centers of circle and obtain triangle, calculate three dot center's coordinate positions, and calculate this position
With the distance of picture centre, it is transformed into world coordinate system；
Step S38: according to mechanical hand present co-ordinate position (xp0, yp0) and step S37 obtain away from
From (disX3, disY3), it is calculated the triangle center and second of compressor bottom surface three installing hole composition
Coordinate (x3, y3) when image center overlaps, formula is: x3=xp0+disX3, y3=yp0+disY3；
Step S39: according to the coordinate (x1, y1) obtained in step S18, the mechanical hand of record in step S31
In coordinate (x2, y2) and step S38, coordinates computed (x3, y3), is calculated compressor and finally places
Position (finalX, finalY), formula is: finalX=x1+ (x2-x3), finalY=y1+ (y2-y3).
The image collected in described step S1, step S2, step S3 is 24 true color images.
The invention has the beneficial effects as follows: present invention relates generally in air-conditioning production line, by production line
Compressor capture get up and compressor is placed into carrying compressor pallet on, shooting image use
First camera and second camera be same model, avoid different model camera to the full extent and cause
The difference gathering image, the location identification module in compressor home position and compressor drop target
Location identification module carries out the process of image information, obtains home position and the placement compressor of compressor
Position and angle, it is possible to according to practical situation adjust capture with place position and angle, it is not necessary to
Manual site supervises, it is to avoid the empty situation grabbing and grabbing mistake occurs, it is ensured that work the most stable enter
OK, and flow of personnel will be caused to affect the situation of the progress of work owing to labor intensity is big, equipment
Once putting into, life-time service, for comparing the most artificial long-term input, cost substantially reduces, this
System and method in invention is for industrial improved efficiency and alleviates tool in hand labor intensity
Significant.
Accompanying drawing explanation
Fig. 1 is the structural representation of the system capturing and placing compressor in the present invention.
In accompanying drawing, 1, brake, 2, Z axis motor, 3, Z axis decelerator, 4, vertical beam, 5, horizontal
Beam, 6, mechanical hand, 7, light fixture, 8, first camera, 9, bracing frame, 10, y-axis motor, 11,
Y-axis decelerator, 12, the axis of rolling, 13, X-axis guide rail, 14, X-axis motor, 15, pallet, 16,
Second camera.
Detailed description of the invention
The present invention relates to a kind of method captured and place compressor based on image recognition, the method is borrowed
Helping the system captured and place compressor to implement, this system is used for capturing compressor and being placed
On pallet 15, system includes control unit module, the mechanical hand 6 driven by drive mechanism, should
System is provided with the first camera 8 for shooting compressor home position and target placement location, described
Control unit module include CPU and CPU carry out the compressor home position of signal transmission position know
Other module and the location identification module of compressor drop target, the transmitting terminal of the image information of first camera 8
Respectively with location identification module and the location identification module of compressor drop target in compressor home position
Signal receiving end connect, the location identification module in compressor home position and compressor drop target
The signal output part of location identification module is connected with CPU, the control signal outfan of CPU and drive mechanism
Drive end connect, described crawl is additionally provided with for shooting compression with placing in the system of compressor
The second camera 16 of machine angle, the transmitting terminal of the image information of second camera 16 and compressor drop target
Location identification module signal receiving end connect.
Specific embodiment, as it is shown in figure 1, described drive mechanism be arranged on bracing frame 9 X-axis,
The four-axle linked motor system that Y-axis, Z axis linear motion system and Z axis rotary system are constituted, X-axis straight line
Motor system includes the parallel X-axis guide rail 13 being arranged on bracing frame 9 upper end, and X-axis guide rail 13 is around rolling
The Timing Belt of moving axis 12, X-axis guide rail 13 upper end is fixed with crossbeam 5, and the axis of rolling 12 one end connects X-axis electricity
Machine 14 and X-axis decelerator, be provided with the Y-axis leading screw along crossbeam 5 direction in crossbeam 5, crossbeam 5 side is arranged
Having Z axis to fix seat, Z axis is fixed seat and is provided with Y-axis swivel nut and matches with Y-axis leading screw, and Y-axis leading screw is by setting
Put the y-axis motor 10 in crossbeam 5 one end and Y-axis decelerator 11 carry out rotating and drives, Z axis fix seat another
Side is provided with the Z axis swivel nut and vertical beam 4 being slidably matched, and is provided with Z axis leading screw sliding with Z axis swivel nut in vertical beam 4
Dynamic cooperation, vertical beam 4 upper end is provided with Z axis motor 2 and Z axis decelerator 3 drives Z axis screw turns, Z axis electricity
Machine 2 upper end connects brake 1, and mechanical hand 6 is connected to vertical beam 4 by Z axis motor 2 and Z axis decelerator 3
Lower end, described first camera 8 and second camera 16 are the camera of same model, and first camera 8 is solid
Due to the lower end of vertical beam 4, second camera 16 is fixed on pallet 15 side.
The system captured and place compressor based on said structure is utilized to carry out capturing and place compressor
Method comprise the following steps:
Step S1:CPU controls drive mechanism and moves to above pallet 15 by first camera 8, shoots pallet
Image information is also sent the location identification module to compressor drop target by 15 images, and compressor is placed
Pallet 15 image is carried out gradation conversion and generates gray level image, gradation conversion by the location identification module of target
Method formula is that gray=1*R+0*G+0*B, gray represent gray level image, and R, G, B represent respectively
The brightness of red, green, blue three coloured light, carries out template matching to gray level image, finds three protruding placements
The approximate location of point, limits the certain limit around each set-point as effective coverage, respectively by three
Effective coverage carries out binaryzation, and is processed by binary morphology, obtains in each protruding set-point
Heart point position, connects three central points and obtains triangle, calculates center point coordinate and 3 points of triangle
Bigger 2 lines of middle x coordinate and horizontal angle, calculate triangle center point coordinates with in image
The distance of the heart, and it is transformed in world coordinate system (disX0, disY0), according to mechanical hand 6 changing coordinates
Position (xp1, yp1) and (disX0, disY0), calculate on pallet 15 in the trianglees of 3 compositions
The coordinate (x1, y1) of mechanical hand 6 when the heart and first camera 8 center superposition, formula is: x1=xp1+disX0,
Y1=yp1+disY0, the location identification module of compressor drop target obtains pallet 15 by image information
In the coordinate at center of triangle of three set-points composition and the angle of pallet 15 and send extremely
In CPU；
Step S2: move to above compressor by first camera 8, shoots compressor photo, by compressor
Photo carries out gradation conversion and generates gray level image, and gradation conversion method formula is gray=1*R+0*G+
0*B, gray represent gray level image, and R, G, B represent the brightness of red, green, blue three coloured light respectively, right
Gray level image carries out template matching, finds matched position, calculates the distance of matched position and picture centre,
And it is transformed into world coordinate system (disX1, disY1), calculate the distance of first camera 8 and mechanical hand 6 center
(disX2, disY2), according to Current mechanical hands 6 position coordinates (xp4, yp4), (disX1, disY1) with
And (disX2, disY2), (x, y), formula is: x=to obtain the accurate location of compressor
Xp4+disX1+disX2, y=yp4+disY1+disY2, CPU control drive mechanism mobile manipulator 6
The machine of being compressed captures and moves above pallet, and the location identification module in compressor home position is passed through
Image information obtains the original position-information of compressor and sends to CPU, and then CPU controls drive mechanism
Capture compressor by mechanical hand (6) and be delivered to pallet (15) top；
Step S3: first camera 8 moves to the picture centre of the shooting of second camera 16, remembers now
Mechanical hand 6 coordinate (x2, y2), moves to the compressor grabbed above second camera 16, shooting pressure
Image information is also sent the location identification module to compressor drop target, position by contracting motor spindle image
Compressor bottom diagram picture is carried out gradation conversion and generates gray level image by identification module, and gradation conversion method is public
Formula is that gray=1*R+0*G+0*B, gray represent gray level image, and R, G, B represent red respectively, green,
The brightness of blue three color light, carries out template matching to gray level image, finds three installing holes, according to template
The position of coupling, the border in location and installation hole, determine more accurate installing hole by method of least square
The center of circle and radius, calculate line and the horizontal line angle of 3 middle x coordinate bigger 2, press from both sides according to this
Angle rotates with pallet 15 angle of acquisition in step S1 so that rotate latter two angle identical, right
Postrotational image recalculates the center of circle of three installing holes, connects three centers of circle and obtains triangle, calculates
Three dot center's coordinate positions, and calculate the distance of this position and picture centre, it is transformed into world coordinate system,
According to mechanical hand 6 present co-ordinate position (xp0, yp0) and (disX3, disY3), it is calculated compression
The triangle center of machine bottom surface three installing hole composition and coordinate (x3, y3) during second camera center superposition, public
Formula is: x3=xp0+disX3, y3=yp0+disY3, according to the coordinate (x1, y1) obtained in step S1,
Mechanical hand 6 coordinate (x2, y2) and coordinate (x3, y3), be calculated the final placement location of compressor
(finalX, finalY), formula is: finalX=x1+ (x2-x3), finalY=y1+ (y2-y3), pressure
The location identification module of contracting machine drop target obtains the angle information of compressor also according to footprint image information
Transmission passes through mechanical hand 6 to CPU, CPU according to the pallet 15 Angle ambiguity drive mechanism obtained in step S1
Compressor is rotated so that compressor angle is identical with pallet 15 angle, and at the bottom of according to compressor
Three dot center's coordinates of the pallet 15 that the triangle center coordinate of portion's installing hole composition and step S1 obtain
Calculate the final placement location of compressor.
Step S4:CPU controls drive mechanism and is placed on by compressor and finally places position obtained by step S3
Put.
The image collected in described step S1, step S2, step S3 is 24 true color images.
The present invention by existing stylize crawl with laying method be improved to crawl based on image recognition with
Laying method, the method realizes by the system captured and place compressor based on image recognition,
This system can with autonomous classification compressor and the position of drop target and angle and capture and place,
Without manual site supervision and operation, save substantial amounts of manually, system can realize once putting into length
Phase uses, production cost relative reduction, and it can be avoided that the sky of compressor is grabbed and grab after mistake causes
Continuous produce chaotic situation, it is possible to ensure to capture with placement work smooth, stable, carry out accurately,
First camera and second camera for absorbing image use same model, it is possible to ensure two to greatest extent
The similarity of individual image shot by camera is maximum, gets rid of the photo caused due to camera difference and can not contrast ginseng
According to situation.
Claims (7)
1. based on image recognition capture with place compressor a method, the method by capture with
The system placing compressor is implemented, and described system is used for capturing compressor and being placed on pallet
(15), on, system includes control unit module, the mechanical hand (6) driven by drive mechanism,
This system is provided with the first camera (8) for shooting compressor home position and target placement location,
Described control unit module includes that CPU and CPU carries out the position in the compressor home position of signal transmission
Put identification module and the location identification module of compressor drop target, the image information of first camera (8)
Transmitting terminal respectively with location identification module and the position of compressor drop target in compressor home position
The signal receiving end of identification module connects, and location identification module and the compressor in compressor home position are put
The signal output part of the location identification module putting target is connected with CPU, the control signal outfan of CPU with
The drive end of drive mechanism connects；
Described capture with place the system of compressor is additionally provided with for shoot compressor angle the
Two cameras (16), the transmitting terminal of the image information of second camera (16) and compressor drop target
The signal receiving end of location identification module connects；
System above is utilized to be compressed the method that machine captures and places, it is characterised in that described
Compressor capture with placement method comprise the following steps:
Step S1: image information is also sent to compression by first camera (8) shooting pallet (15) image
The location identification module of machine drop target, the location identification module of compressor drop target is believed by image
Breath obtains coordinate and the pallet (15) at the center of the triangle of three set-point compositions in pallet (15)
Angle；
Step S2: first camera (8) shooting is positioned at the compressor map picture in home position and by image information
Send the location identification module to compressor home position, the location identification module in compressor home position
Obtaining the original position-information of compressor by image information and send to CPU, then CPU controls to drive
Mechanism captures compressor by mechanical hand (6) and is delivered to pallet (15) top；
Step S3: second camera (16) shooting compressor ground plan picture also sends to compressor drop target
Location identification module, the location identification module of compressor drop target obtains according to footprint image information
The angle information of compressor also sends to CPU, CPU according to pallet (15) angle obtained in step S1
Control drive mechanism by mechanical hand (6), compressor to be rotated so that compressor angle and pallet
(15) angle is identical, and according to compressor bottom mounting apertures composition triangle center coordinate and step
Three dot center's coordinates of the pallet (15) that rapid S1 obtains calculate the final placement location of compressor；
Step S4:CPU controls drive mechanism and is placed on by compressor and finally places position obtained by step S3
Put；
In described step S1, the location identification module of compressor drop target obtains pallet (15) position
Include with the step of angle information:
Step S11:CPU controls drive mechanism and first camera (8) moves to pallet (15) top,
Shooting pallet (15) image；
Step S12: pallet (15) image is carried out gradation conversion and generates gray level image, gradation conversion side
Method formula is that gray=1*R+0*G+0*B, gray represent gray level image, and R, G, B represent red respectively,
Green, the brightness of blue three color light；
Step S13: gray level image is carried out template matching, finds three protruding placements in pallet (15)
The approximate location of point, limits the certain limit around each set-point as effective coverage；
Step S14: respectively three effective coverages are carried out binaryzation, and processed by binary morphology,
Obtain the center position of each protruding set-point；
Step S15: connect three central points and obtain triangle, calculate the center point coordinate and three of triangle
2 lines that in point, x coordinate is bigger and horizontal angle；
Step S16: calculate the distance of triangle center point coordinates and picture centre, and be transformed into world's seat
In mark system (disX0, disY0)；
Step S17: according to mechanical hand (6) present co-ordinate position (xp1, yp1), and in step S16
The distance arrived, calculates in triangle center and the first camera (8) of upper 3 compositions of pallet (15)
The coordinate (x1, y1) of mechanical hand (6) when the heart overlaps, formula is: x1=xp1+disX0, y1=yp1+disY0.
The method captured and place compressor based on image recognition the most according to claim 1,
It is characterized in that: described drive mechanism is that the X-axis that is arranged on bracing frame (9), Y-axis, Z axis are straight
The four-axle linked motor system that line motor system and Z axis rotary system are constituted, X-axis linear motion system bag
Including the parallel X-axis guide rail (13) being arranged on bracing frame (9) upper end, X-axis guide rail (13) is cincture
The Timing Belt of the axis of rolling (12), X-axis guide rail (13) upper end is fixed with crossbeam (5), the axis of rolling (12)
One end connects X-axis motor (14) and X-axis decelerator, is provided with along crossbeam (5) in crossbeam (5)
The Y-axis leading screw in direction, crossbeam (5) side is provided with Z axis and fixes seat, and Z axis is fixed seat and is provided with Y-axis spiral shell
Set matches with Y-axis leading screw, and Y-axis leading screw is by being arranged on the y-axis motor (10) of crossbeam (5) one end
Carrying out rotating with Y-axis decelerator (11) and drive, Z axis is fixed seat opposite side and is provided with the Z axis being slidably matched
Swivel nut and vertical beam (4), be provided with Z axis leading screw in vertical beam (4) and be slidably matched with Z axis swivel nut, vertical beam
(4) upper end is provided with Z axis motor (2) and Z axis decelerator (3) drives Z axis screw turns, Z axis electricity
Machine (2) upper end connects has brake (1), mechanical hand (6) to be subtracted by Z axis motor (2) and Z axis
Speed device (3) is connected to the lower end of vertical beam (4).
The method captured and place compressor based on image recognition the most according to claim 1,
It is characterized in that: described first camera (8) and second camera (16) are the camera of same model.
The method captured and place compressor based on image recognition the most according to claim 2,
It is characterized in that: described first camera (8) is fixed on the lower end of vertical beam (4), described second
Camera (16) is fixed on pallet (15) side.
The method captured and place compressor based on image recognition the most according to claim 1,
It is characterized in that, the location identification module in described compressor home position obtains compressor home position
Step include:
Step S21: move to above compressor by first camera (8), shoots compressor photo；
Step S22: compressor photo carrying out gradation conversion and generates gray level image, gradation conversion method is public
Formula is that gray=1*R+0*G+0*B, gray represent gray level image, and R, G, B represent red respectively, green,
The brightness of blue three color light；
Step S23: gray level image is carried out template matching, finds matched position；
Step S24: calculate the distance of matched position and picture centre, and be transformed into world coordinate system
(disX1,disY1)；
Step S25: calculate the distance (disX2, disY2) of first camera (8) and mechanical hand (6) center；
Step S26: according to Current mechanical hands (6) position coordinates (xp4, yp4), step S24 calculate away from
First camera (8) and the distance at mechanical hand (6) center in (disX1, disY1) and step S25
(disX2, disY2), (x, y), formula is: x=to obtain the accurate location of compressor
Xp4+disX1+disX2, y=yp4+disY1+disY2, CPU control drive mechanism mobile manipulator (6)
The machine of being compressed captures and moves above pallet.
The method captured and place compressor based on image recognition the most according to claim 1,
It is characterized in that, in described step S3, the location identification module of compressor drop target obtains compressor
The step of angle and final placement location includes:
Step S31: first camera (8) is moved to the picture centre of the shooting of second camera (16),
Note mechanical hand (6) coordinate (x2, y2) now；
Step S32: the compressor grabbed is moved to second camera (16) top, shoots compressor
Bottom panels；
Step S33: compressor bottom image is carried out gradation conversion and generates gray level image, gradation conversion side
Method formula is that gray=1*R+0*G+0*B, gray represent gray level image, and R, G, B represent red respectively,
Green, the brightness of blue three color light；
Step S34: gray level image is carried out template matching, finds three installing holes；
Step S35: according to the position of template matching, the border in location and installation hole, pass through method of least square
Determine the center of circle and the radius of more accurate installing hole；
Step S36: calculate line and the horizontal line angle of 3 middle x coordinate bigger 2, press from both sides according to this
Angle rotates with pallet (15) angle of acquisition in step S1 so that rotate latter two angle identical；
Step S37: use the method in step S35 to recalculate three installing holes postrotational image
The center of circle, connect three centers of circle and obtain triangle, calculate three dot center's coordinate positions, and calculate this position
With the distance of picture centre, it is transformed into world coordinate system；
Step S38: obtain according in mechanical hand (6) present co-ordinate position (xp0, yp0) and step S37
Distance (disX3, disY3), be calculated the triangle center of compressor bottom surface three installing hole composition with
Coordinate (x3, y3) during second camera center superposition, formula is: x3=xp0+disX3, y3=yp0+disY3；
Step S39: according to the coordinate (x1, y1) obtained in step S18, the mechanical hand of record in step S31
(6) coordinates computed (x3, y3) in coordinate (x2, y2) and step S38, is calculated compressor final
Placement location (finalX, finalY), formula is: finalX=x1+ (x2-x3),
FinalY=y1+ (y2-y3).
The method captured and place compressor based on image recognition the most according to claim 1,
It is characterized in that, the image collected in described step S1, step S2, step S3 is 24 very
Coloured image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510437176.8A CN105082158B (en) | 2015-07-23 | 2015-07-23 | Method for grabbing and placing compressor based on image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510437176.8A CN105082158B (en) | 2015-07-23 | 2015-07-23 | Method for grabbing and placing compressor based on image recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105082158A CN105082158A (en) | 2015-11-25 |
| CN105082158B true CN105082158B (en) | 2017-01-11 |
Family
ID=54563928
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510437176.8A Active CN105082158B (en) | 2015-07-23 | 2015-07-23 | Method for grabbing and placing compressor based on image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105082158B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109760463A (en) * | 2018-12-29 | 2019-05-17 | 广东博智林机器人有限公司 | Wallpaper piece alignment means and the method for carrying out the alignment of wallpaper seam using it |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106040613B (en) * | 2016-05-24 | 2018-05-18 | 四川研宝科技有限公司 | A kind of automated power detection method |
| CN106091922B (en) * | 2016-05-25 | 2019-05-14 | 广州市思林杰自动化科技有限公司 | The method and device that a kind of pair of workpiece is detected |
| CN106960099B (en) * | 2017-03-28 | 2019-07-26 | 清华大学 | A kind of manipulator grasp stability recognition methods based on deep learning |
| CN106829469A (en) * | 2017-03-30 | 2017-06-13 | 武汉库柏特科技有限公司 | A kind of unordered grabbing device of robot based on double camera and method |
| CN107999330B (en) * | 2017-12-29 | 2020-06-30 | 科睿驰（深圳）医疗科技发展有限公司 | Automatic catheter coating equipment and method based on visual processing |
| CN109968237B (en) * | 2019-03-19 | 2021-11-23 | 珠海格力智能装备有限公司 | Mounting and positioning method and device of air conditioner |
| CN111240329A (en) * | 2020-01-17 | 2020-06-05 | 广东智源机器人科技有限公司 | Intelligent robot control method and device and intelligent robot |
| CN113997290B (en) * | 2021-11-17 | 2023-03-28 | 合肥联宝信息技术有限公司 | Control method and equipment for mechanical arm |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1882242A (en) * | 2005-06-02 | 2006-12-20 | 重机公司 | Element mounting device |
| CN101026952A (en) * | 2006-02-16 | 2007-08-29 | 重机公司 | Electronic component mounting method and device |
| CN101187549A (en) * | 2007-12-28 | 2008-05-28 | 谭建平 | Method and device for on-line monitoring multiple movable member center using double laser beam |
| CN101362332A (en) * | 2007-08-09 | 2009-02-11 | 富士通株式会社 | Part-delivery apparatus and method for the same |
| JP2010221381A (en) * | 2009-03-25 | 2010-10-07 | Fuji Xerox Co Ltd | Method for assembling part and device for assembling the part |
| CN204324384U (en) * | 2014-11-21 | 2015-05-13 | 河北科技大学 | Based on object localization and the crawl platform of opto-electronic pickup |
-
2015
- 2015-07-23 CN CN201510437176.8A patent/CN105082158B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1882242A (en) * | 2005-06-02 | 2006-12-20 | 重机公司 | Element mounting device |
| CN101026952A (en) * | 2006-02-16 | 2007-08-29 | 重机公司 | Electronic component mounting method and device |
| CN101362332A (en) * | 2007-08-09 | 2009-02-11 | 富士通株式会社 | Part-delivery apparatus and method for the same |
| CN101187549A (en) * | 2007-12-28 | 2008-05-28 | 谭建平 | Method and device for on-line monitoring multiple movable member center using double laser beam |
| JP2010221381A (en) * | 2009-03-25 | 2010-10-07 | Fuji Xerox Co Ltd | Method for assembling part and device for assembling the part |
| CN204324384U (en) * | 2014-11-21 | 2015-05-13 | 河北科技大学 | Based on object localization and the crawl platform of opto-electronic pickup |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109760463A (en) * | 2018-12-29 | 2019-05-17 | 广东博智林机器人有限公司 | Wallpaper piece alignment means and the method for carrying out the alignment of wallpaper seam using it |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105082158A (en) | 2015-11-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105082158B (en) | Method for grabbing and placing compressor based on image recognition | |
| CN107414832A (en) | A kind of mobile mechanical arm crawl control system and method based on machine vision | |
| CN107030693B (en) | A kind of hot line robot method for tracking target based on binocular vision | |
| CN109483531B (en) | Machine vision system and method for picking and placing FPC board by manipulator at fixed point | |
| CN103522291B (en) | The target grasping system of a kind of explosive-removal robot and method | |
| CN107253190A (en) | The device and its application method of a kind of high precision machines people trick automatic camera calibration | |
| CN108834576A (en) | A kind of citrus picking robot and its implementation based on binocular vision | |
| CN107590835A (en) | Mechanical arm tool quick change vision positioning system and localization method under a kind of nuclear environment | |
| CN107876269A (en) | The trinocular vision spraying profile extraction system and its method of work of footwear mould automatic glue-spraying | |
| CN205238065U (en) | Snatch many sucking discs manipulator of irregularity body | |
| CN1976445A (en) | Multi-camera monitoring system based on three-dimensional video frequency dynamic tracking and tracking method thereof | |
| CN110470241A (en) | A kind of refractory brick curvature detection system and method based on structure light vision | |
| CN104308530B (en) | The isolator star-wheel Automated assembly device of view-based access control model detection | |
| CN207206429U (en) | A kind of mobile mechanical arm crawl control system based on machine vision | |
| CN104760453B (en) | A kind of imprinter automatic loading and unloading device and its control method | |
| CN209174850U (en) | The device of big packet collector nozzle is positioned using machine vision | |
| CN206124463U (en) | Manipulator formula laser marking machine | |
| CN108174111A (en) | Crusing robot target image grasping means | |
| CN107263189A (en) | Auxiliary processing device, semi-automatic assistance processing method and automatic assistance processing method | |
| CN109848986B (en) | Positioning method for uncertain height of target object based on machine vision | |
| Madhira et al. | AGWallP—Automatic guided wall painting system | |
| CN207372549U (en) | Integrated handgrip grip device | |
| CN206728172U (en) | A kind of video monitoring apparatus | |
| CN207272673U (en) | Multifunctional zero component assembling machine | |
| CN208019761U (en) | It locks screw workbench and its five axis locks screw device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |