CN104729406B - A kind of machine vision localization method of element pasted on surface - Google Patents

A kind of machine vision localization method of element pasted on surface Download PDF

Info

Publication number
CN104729406B
CN104729406B CN201510125262.5A CN201510125262A CN104729406B CN 104729406 B CN104729406 B CN 104729406B CN 201510125262 A CN201510125262 A CN 201510125262A CN 104729406 B CN104729406 B CN 104729406B
Authority
CN
China
Prior art keywords
coordinate
camera
feature point
tray platform
workpiece
Prior art date
Application number
CN201510125262.5A
Other languages
Chinese (zh)
Other versions
CN104729406A (en
Inventor
王德民
曹新
杨希
李继周
潘小弟
Original Assignee
上海瑞伯德智能系统股份有限公司
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 上海瑞伯德智能系统股份有限公司 filed Critical 上海瑞伯德智能系统股份有限公司
Priority to CN201510125262.5A priority Critical patent/CN104729406B/en
Publication of CN104729406A publication Critical patent/CN104729406A/en
Application granted granted Critical
Publication of CN104729406B publication Critical patent/CN104729406B/en

Links

Abstract

A kind of machine vision localization method of element pasted on surface, using the process of the image of camera collection surface mount components, wherein, the tray platform that setting can in the horizontal direction be translated and rotated, one camera is set above tray platform, view data Transfer pipe is set between camera and computer, a calibration steps is carried out first, then carry out the positioning step of workpiece to be processed.Using image shot by camera, collection volume of data, obtain a coordinate system being made up of the axial coordinate and visual pattern location of pixels of platform, and then measure side-play amount and calibrated, photo is sent to computer again, characteristic point particular location in the picture is found out by vision matching algorithm in the picture, sampled by 2 points to the certain distance on workpiece, measurement radix is expanded into an order of magnitude, making the error of image can almost ignore, and avoid the dependence of the shape and vision algorithm for characteristic point, applicable surface is wider, precision is higher.

Description

A kind of machine vision localization method of element pasted on surface

Technical field:

The present invention relates to electricity field, more particularly to electronic component surface mounting technology, particularly a kind of surface mount The machine vision localization method of element.

Background technology:

Surface mounting technology industry is high due to Product Precision, and workpiece situation is complicated and the reason such as changeable, is done with tradition positioning Method cannot be produced precisely and at high speed.In the prior art, element pasted on surface is improve using vision positioning method to be held in the palm in processing The location efficiency of disk platform.But, existing vision positioning technology analyzes the position for obtaining by template image, specifically, Exactly there is the vector graphics of certain identification by target workpiece, its angle and position is found out by vision computing. This method is not accurate enough, is calculated by pixel especially in the measuring and calculating of angle, when sampling due to vision, pixel There is no floating number this concept, so inevitably there is the missing of precision, cause the presence of error.

The content of the invention:

It is an object of the invention to provide a kind of machine vision localization method of element pasted on surface, described this surface The machine vision localization method of mount components will solve to analyze the vision positioning for obtaining position by template image in the prior art There is the technical problem of precision missing and error in technology.

The machine vision localization method of this element pasted on surface of the invention, an including utilization camera is gathered and analyzed The process of the image of element pasted on surface, wherein, the image of element pasted on surface is gathered and analyzed in described utilization camera During, a tray platform that can in the horizontal direction translate and rotate is set, fixed in the top of described tray platform One camera is set, view data Transfer pipe is set between described camera and a computer, a school is carried out first Quasi- step, then carries out the positioning step of workpiece to be processed, in described calibration steps, a circle is made in tray platform Shape characteristic point is simultaneously placed into viewing field of camera center, and then tray platform is carried out the flat of X-direction and Y-direction respectively Move, using the image of the circular feature point of camera collection two or more position, put down in conjunction with the mobile pallet of each image correspondence The coordinate of platform, the location of pixels obtained using the coordinate position and correspondence of each positioning calculates location of pixels and seat in photo Target proportionate relationship, then follow the steps below:

Step one, viewing field of camera center is placed on by circular feature point,

Step 2, by tray platform to rotating clockwise or counter-clockwise 1 °,

Step 3, then in X direction with the translational tray platform of Y-direction, make circular feature point be located at image center location,

Step 4, the current X of recoding tray platform, the coordinate of Y-direction,

Step 5, repeat step two arrives step 4, stops when being more than 60 ° to the accumulative deflection of tray platform,

The tray platform coordinate sequence that step one to step 5 is collected calculates centre coordinate by least square method, The coordinate is the actual coordinate in the tray platform corresponding to camera image center,

In the positioning step of described workpiece to be processed, workpiece to be processed is placed in standard Working position, to be processed One fisrt feature point and a second feature point are set on workpiece, and mobile tray platform makes fisrt feature point appear in camera In visual field, photograph is at this moment gathered, regarded according to the camera obtained in the location of pixels of fisrt feature point in photo, and calibration flow Field centre coordinate, draws the standard machining coordinate of fisrt feature point, then moves tray platform, second feature point is appeared in phase In machine visual field, photograph is at this moment gathered, according to the camera obtained in the location of pixels of second feature point in photo, and calibration flow The coordinate of field of view center, draws the standard machining coordinate of second feature point;According to reality processing coordinate and standard machining coordinate Deviation is repositioned, and reaches standard Working position, and new workpiece to be processed positions two fact characteristic points in visual field Location of pixels, the corresponding coordinate for calculating is compared by the standard machining coordinate with two characteristic points, directly calculates new work The differential seat angle of Working position residing for part and standard Working position, then by the seat after the correction angle for calculating new workpiece Mark, draws x, and the translation distance of y is repositioned, and reaches standard Working position.

The present invention and prior art compare, and its effect is positive and obvious.The present invention utilizes industrial camera shooting figure Picture, gathers volume of data, obtains a coordinate system being made up of the axial coordinate and visual pattern location of pixels of platform, and then Accurately measure side-play amount and calibrated, then be sent to photo by including the view data Transfer pipe including Ethernet Computer, finds out characteristic point particular location in the picture, by workpiece in the picture by vision matching algorithm 2 points of set a distance are sampled, and measurement radix is expanded into an order of magnitude, and making the error of image can almost ignore, and The dependence of the shape and vision algorithm for characteristic point is avoided, applicable surface is wider, and precision is higher.

Brief description of the drawings:

Fig. 1 is a kind of schematic diagram of the machine vision localization method of element pasted on surface of the invention.

Specific embodiment:

Embodiment 1:

A kind of machine vision localization method of element pasted on surface of the present invention, an including utilization camera is gathered and analytical table The process of the image of face mount components, wherein, in described utilization camera collection and the mistake of the image of analysis element pasted on surface Cheng Zhong, sets a tray platform 1 that can in the horizontal direction translate and rotate, and is fixed in the top of described tray platform 1 One camera 2 is set, view data Transfer pipe is set between described camera 2 and a computer (not shown), it is first A calibration steps is first carried out, the positioning step of workpiece to be processed 3 is then carried out, it is flat in pallet in described calibration steps A circular feature point is made in platform 1 and is placed into the field of view center position of camera 2, tray platform 1 then being carried out into X respectively Direction and the translation of Y-direction, using the image of the circular feature point of the collection two or more of camera 2 position, in conjunction with each image The coordinate of the mobile tray platform of correspondence, the location of pixels obtained using the coordinate position and correspondence of each positioning, calculates photograph The proportionate relationship of location of pixels and coordinate in piece, then follow the steps below:

Step one, the field of view center position of camera 2 is placed on by circular feature point,

Step 2, by tray platform 1 to rotating clockwise or counter-clockwise 1 °,

Step 3, then in X direction with the translational tray platform 1 of Y-direction, make circular feature point be located at image center location,

Step 4, the current X of recoding tray platform 1, the coordinate of Y-direction,

Step 5, repeat step two arrives step 4, stops when being more than 60 ° to the accumulative deflection of tray platform 1,

The coordinate sequence of tray platform 1 that step one to step 5 is collected calculates center and sits by least square method Mark, the coordinate is the actual coordinate in the tray platform 1 corresponding to the picture centre of camera 2,

In the positioning step of described workpiece to be processed 3, workpiece to be processed 3 is placed in standard Working position, to be added An one fisrt feature point B1 and second feature point B2 is set on work workpiece 3, and mobile tray platform 1 makes fisrt feature point B1 Appear in the visual field of camera 2, at this moment gather photograph, according to the location of pixels of fisrt feature point B1 in photo, and calibration flow In the coordinate of circular feature point that obtains, it can be deduced that the standard machining coordinate of fisrt feature point B1.Then tray platform is moved 1, in second feature point B2 is appeared in the visual field of camera 2, photograph is at this moment gathered, according to the pixel position of second feature point B2 in photo Put, and calibrate the coordinate of the circular feature point obtained in flow, draw the standard machining coordinate of second feature point B2.

In actual production, new workpiece is placed on the Working position on platform, mobile platform, goes out fact characteristic point B1 In the present visual field of camera 2, photograph is at this moment gathered, obtained according in the location of pixels of characteristic point B1 in photo, and calibration flow Circular feature point coordinate, it can be deduced that the reality processing coordinate of characteristic point B1.Then mobile platform, makes fact characteristic point B2 Appear in the visual field of camera 2, at this moment gather photograph, obtained according in the location of pixels of characteristic point B2 in photo, and calibration flow The field of view center coordinate for arriving, draws the reality processing coordinate of characteristic point B2.

The coordinate of the standard Working position of known two characteristic points of workpiece again, can according to the coordinate of reality processing position and The coordinate of standard Working position, obtains X to, Y-direction and anglec of rotation deviation.Platform is according to front and rear deviation, left-right deviation and rotation After gyration is adjusted position, workpiece can be moved to standard Working position.

Specifically, in the present embodiment, fisrt feature point B1 and second feature point B2 uses cross pattern.

Claims (1)

1. a kind of machine vision localization method of element pasted on surface, an including utilization camera is gathered and analysis surface mount unit The process of the image of part, it is characterised in that:In described utilization camera collection and the process of the image of analysis element pasted on surface In, a tray platform that can in the horizontal direction translate and rotate is set, it is fixedly installed above described tray platform One camera, sets view data Transfer pipe between described camera and a computer, and a calibration step is carried out first Suddenly, the positioning step of workpiece to be processed is then carried out, in described calibration steps, a circular spy is made in tray platform Viewing field of camera center is levied a little and be placed into, then tray platform is carried out the flat of X-direction and Y direction respectively Move, using the image of the circular feature point of camera collection two or more position, put down in conjunction with the mobile pallet of each image correspondence The coordinate of platform, the location of pixels obtained using the coordinate position and correspondence of each positioning calculates location of pixels and seat in photo Target proportionate relationship, then follow the steps below:
Step one, viewing field of camera center is placed on by circular feature point,
Step 2, by tray platform to rotating clockwise or counter-clockwise 1 °,
Step 3, then along X-direction and Y direction translational tray platform, circular feature point is located at image center location,
Step 4, the current X of recoding tray platform, the coordinate of Y direction,
Step 5, repeat step two arrives step 4, stops when being more than 60 ° to the accumulative deflection of tray platform,
The tray platform coordinate sequence that step one to step 5 is collected calculates centre coordinate by least square method, the seat The actual coordinate in the tray platform corresponding to camera image center is designated as,
In the positioning step of described workpiece to be processed, workpiece to be processed is placed in standard Working position, in workpiece to be processed Upper to set a fisrt feature point and a second feature point, mobile tray platform makes fisrt feature point appear in viewing field of camera In, photograph is at this moment gathered, according in the viewing field of camera obtained in the location of pixels of fisrt feature point in photo, and calibration flow Heart coordinate, draws the standard machining coordinate of fisrt feature point, then moves tray platform, second feature point is appeared in camera and regards In, photograph is at this moment gathered, according to the viewing field of camera obtained in the location of pixels of second feature point in photo, and calibration flow The coordinate at center, draws the standard machining coordinate of second feature point;According to reality processing coordinate and the deviation of standard machining coordinate Repositioned, reached standard Working position, new workpiece to be processed is positioned pixel of two fact characteristic points in visual field Position, the corresponding coordinate for calculating is compared by the standard machining coordinate with two characteristic points, directly calculates new workpiece institute The differential seat angle of place's Working position and standard Working position, then by the coordinate after the correction angle for calculating new workpiece, obtains Go out x, the translation distance of y is repositioned, and reaches standard Working position.
CN201510125262.5A 2015-03-23 2015-03-23 A kind of machine vision localization method of element pasted on surface CN104729406B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510125262.5A CN104729406B (en) 2015-03-23 2015-03-23 A kind of machine vision localization method of element pasted on surface

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510125262.5A CN104729406B (en) 2015-03-23 2015-03-23 A kind of machine vision localization method of element pasted on surface

Publications (2)

Publication Number Publication Date
CN104729406A CN104729406A (en) 2015-06-24
CN104729406B true CN104729406B (en) 2017-06-13

Family

ID=53453534

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510125262.5A CN104729406B (en) 2015-03-23 2015-03-23 A kind of machine vision localization method of element pasted on surface

Country Status (1)

Country Link
CN (1) CN104729406B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105354858B (en) * 2015-12-09 2018-04-24 北京中电科电子装备有限公司 Scribing machine searches the method, apparatus and scribing machine of same block characteristic point
CN105643265B (en) * 2016-01-12 2018-05-22 济南大学 A kind of detection method for the cooperation of large-scale workpiece fitting surface
CN106931912B (en) * 2016-12-30 2019-03-05 西安航天精密机电研究所 A method of for examining torquer
CN108243602A (en) * 2017-01-06 2018-07-03 广东华志珹智能科技有限公司 A kind of screening cover chip mounter image identification system and method
CN106814307B (en) * 2017-01-10 2020-05-12 深圳鼎缘电子科技有限公司 Automatic debugging method and system for cavity filter
CN107578442B (en) * 2017-09-05 2019-11-26 西北工业大学 A kind of centralized positioning aligning method and system
CN110854723A (en) * 2019-10-17 2020-02-28 湖南长高高压开关集团股份公司 Method for correcting porcelain insulator of open-type isolating switch

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102103336A (en) * 2011-03-14 2011-06-22 张雯 High-accuracy alignment mark structure based on machine vision alignment
CN105323455A (en) * 2014-07-31 2016-02-10 宁波舜宇光电信息有限公司 Positioning compensation method based on machine vision

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1349114A3 (en) * 2002-03-19 2011-06-15 Canon Kabushiki Kaisha Sensor calibration apparatus, sensor calibration method, program, storage medium, information processing method, and information processing apparatus
JP5440461B2 (en) * 2010-09-13 2014-03-12 株式会社リコー Calibration apparatus, distance measurement system, calibration method, and calibration program

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102103336A (en) * 2011-03-14 2011-06-22 张雯 High-accuracy alignment mark structure based on machine vision alignment
CN105323455A (en) * 2014-07-31 2016-02-10 宁波舜宇光电信息有限公司 Positioning compensation method based on machine vision

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于机器视觉的贴片机元件定位的图像处理算法研究;王力等;《电工电气》;20090131(第1期);第57-59页 *

Also Published As

Publication number Publication date
CN104729406A (en) 2015-06-24

Similar Documents

Publication Publication Date Title
CN105678785B (en) A kind of laser and the scaling method of camera relative pose relation
CN107301654B (en) Multi-sensor high-precision instant positioning and mapping method
CN103292701B (en) The online dimension measurement method of accurate device based on machine vision
TWI708216B (en) Method and system for calibrating vision system in environment
CN102107374B (en) On-line detection method for diameter size of shaft disc type part
CN105345194B (en) A kind of PCB pad localization methods based on machine vision
CN104331896B (en) A kind of system calibrating method based on depth information
CN103499297B (en) A kind of high-precision measuring method based on CCD
CN101630409B (en) Hand-eye vision calibration method for robot hole boring system
CN105551039B (en) The scaling method and device of structural light three-dimensional scanning system
CN100458359C (en) Small-displacement measuring system in long-distance plane
WO2017128865A1 (en) Multiple lens-based smart mechanical arm and positioning and assembly method
CN103747617A (en) PCB expansion compensation method
CN101726246B (en) Correcting method
CN102930266B (en) Method for locating Mark points on PCB (printed circuit board) by utilizing outline gravity center method
CN103438826B (en) The three-dimension measuring system of the steel plate that laser combines with vision and method
CN103743352B (en) A kind of 3 D deformation measuring method based on polyphaser coupling
CN103714535A (en) Binocular vision measurement system camera parameter online adjustment method
CN106493042A (en) Dispensing method and dispenser system
US20160291571A1 (en) System and method for aligning a coordinated movement machine reference frame with a measurement system reference frame
CN101033972A (en) Method for obtaining three-dimensional information of space non-cooperative object
US8553971B2 (en) Method and system for measuring object
CN103292695B (en) A kind of single eye stereo vision measuring method
CN101788805B (en) High-accuracy machine vision two-dimensional positioning method based on motion servo correction
CN106204574B (en) Camera pose self-calibrating method based on objective plane motion feature

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
CB02 Change of applicant information
CB03 Change of inventor or designer information
CB02 Change of applicant information

Address after: 201821 Shanghai city Jiading District Zhaoxian Road No. 655

Applicant after: Shanghai ruibode intelligent system Limited by Share Ltd

Address before: 200335 Shanghai city Changning District road of 8 Building No. 3

Applicant before: Shanghai Ruibode Intelligent System Sci. & Tech. Co., Ltd.

Address after: 201821 Shanghai city Jiading District Zhaoxian Road No. 655

Applicant after: Shanghai ruibode intelligent system Limited by Share Ltd

Address before: 200335 Shanghai city Changning District road of 8 Building No. 3

Applicant before: Shanghai Ruibode Intelligent System Sci. & Tech. Co., Ltd.

CB03 Change of inventor or designer information

Inventor after: Wang Demin

Inventor after: Cao Xin

Inventor after: Yang Xi

Inventor after: Li Jizhou

Inventor after: Pan Xiaodi

Inventor before: Lv Deyun

Inventor before: Wang Demin

Inventor after: Wang Demin

Inventor after: Cao Xin

Inventor after: Yang Xi

Inventor after: Li Jizhou

Inventor after: Pan Xiaodi

Inventor before: Lv Deyun

Inventor before: Wang Demin

GR01 Patent grant
GR01 Patent grant
CP01 Change in the name or title of a patent holder
CP01 Change in the name or title of a patent holder

Address after: 201821 Shanghai city Jiading District Zhaoxian Road No. 655

Patentee after: SHANGHAI CHINESE CAR RIBERD INTELLIGENT SYSTEM Co.,Ltd.

Address before: 201821 Shanghai city Jiading District Zhaoxian Road No. 655

Patentee before: SHANGHAI RUIBODE INTELLIGENT SYSTEMS Co.,Ltd.