CN106503720B - A kind of element image-recognizing method of removal suction nozzle interference - Google Patents
A kind of element image-recognizing method of removal suction nozzle interference Download PDFInfo
- Publication number
- CN106503720B CN106503720B CN201610913757.9A CN201610913757A CN106503720B CN 106503720 B CN106503720 B CN 106503720B CN 201610913757 A CN201610913757 A CN 201610913757A CN 106503720 B CN106503720 B CN 106503720B
- Authority
- CN
- China
- Prior art keywords
- element image
- suction nozzle
- point
- image
- vertex
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/06—Recognition of objects for industrial automation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of image-recognizing methods of removal suction nozzle interference, belong to image identification technical field.The image-recognizing method that the present invention removes suction nozzle interference extracts the marginal point of element image by way of edge detection, coarse positioning is carried out to edge point using minimum circumscribed rectangle, determine candidate's long side and candidate short side, best long side and best short side are determined from candidate long side and candidate short side by calculating the number being closer a little, the best long side of fitting and the straight-line intersection of best short side are the first vertex, the second vertex and third vertex are relatively obtained by derived function, peak value again, the profile of element script can be obtained.The interference of suction nozzle when this image-recognizing method can effectively remove industrial camera capturing element image, ensure that accuracy, the reliability of Placement element.
Description
Technical field
The present invention relates to image identification technical field more particularly to a kind of element image recognition sides of removal suction nozzle interference
Method.
Background technique
Chip mounter is for realizing the equipment at a high speed, being accurately fully automatically placed with component, is in entire SMT production
Most critical, most complicated equipment, as the capital equipment in the production line of SMT, chip mounter is from the low-speed machinery patch of early stage
Machine development be high speed optical centering chip mounter, and to it is multi-functional, flexible connection modular development.The mounting head of chip mounter is designed with
Suction nozzle unit and industrial camera, suction nozzle unit picks and places element using the principle of vacuum suction, and industrial camera is then used for
Position identification judgement is carried out to the element of suction nozzle unit absorption, this is because suction nozzle unit in absorptive element, does not ensure that
The position angle relationship of element and suction nozzle unit, that is to say, that element may deviate from certain distance or deflect certain angle
Degree, this just needs industrial camera to accurately identify the position of element, then passes through the Z axis rotary components of mounting head to element
Angular deflection carries out rotation compensation or carries out straight line compensation to positional shift, therefore industrial camera knows the position of element, angle
Other ability is most important for whole attachment process.In addition, suction nozzle unit is in absorptive element, by industrial camera
The element image outline of shooting is possible to can be comprising the profile of suction nozzle unit, as shown in Figure 1, Figure 2 and Figure 3, suction nozzle unit in Fig. 1
Profile expose in the long side of element image, the profile of suction nozzle element exposes in the short side of element image in Fig. 2, suction nozzle in Fig. 3
The profile of element exposes in the apex of element image, and no matter suction nozzle element exposes from which position of element image can all cause
The obstacle of element profile identification, and then influence the precision of component mounter.Just need thus one kind can effectively remove suction nozzle interference,
And the method that the profile of element itself is accurately identified from element image.
Summary of the invention
The technical problems to be solved by the invention are to provide a kind of element image-recognizing method, can effectively remove suction
Mouth interferes and accurately identifies element profile from element image.
The technical proposal adopted by the invention to solve the above technical problems is that:
The present invention provides a kind of element image-recognizing methods of removal suction nozzle interference, comprising steps of adopting to element image
Marginal point is extracted with the mode of edge detection, and coarse positioning is carried out to the marginal point using minimum circumscribed rectangle；According to most
Each edge lengths of small boundary rectangle determine two candidate long sides and two candidate short sides；Each back gauge is calculated less than m pixel
The number of point, the largest number of candidate long sides of the point met the requirements are best long side, and the largest number of candidate's short sides are best short
Side, 2≤m≤5, m ∈ N*；The point for being less than m pixel apart from best long side and best short side is fitted, the two of digital simulation are straight
The intersection point of line obtains the first vertex of element profile；All marginal points are counted to the distance of best long side, pass through calculating y1=f
(x1) first derivative at each point, x1For distance, y1For number, number peak value y is determined1max1And number minor peaks y1max2,
By number peak value y1max1And number minor peaks y1max2Corresponding distance x1max1And x1max2Compared with component size setting value
Compared with immediate distance is x1maxThe as length of the first side length of element profile obtains the second vertex of element profile；Statistics
All marginal points pass through calculating y to the distance of best short side2=f (x2) first derivative at each point, x2For distance, y2It is a
Number, determines number peak value y2max1And number minor peaks y2max2, by number peak value y2max1And number minor peaks y2max2Corresponding
Distance x2max1And x2max2It is compared with component size setting value, immediate distance is x2maxAs element profile is perpendicular to institute
The length for stating the second side length of the first side length obtains the third vertex of element profile；Element profile is determined according to three vertex.
It as a further improvement of the above technical scheme, be to element before to element image zooming-out marginal point the step of
Image carries out binary conversion treatment.
It as a further improvement of the above technical scheme, be to element before to element image zooming-out marginal point the step of
Image carries out opening operation processing.
As a further improvement of the above technical scheme, in the fit procedure of best long side and best short side using
Least square method.
As a further improvement of the above technical scheme, the number step for being less than the point of m pixel is being calculated to each back gauge
In, the value of m is 3.
The beneficial effects of the present invention are:
Element image-recognizing method of the present invention by element image profile point carry out edge collecting, fitting, calculate derivation,
Peak value compares, and finally can accurately determine three vertex of rectangular element, and then determine the profile of element itself, effectively remove
The interference of suction nozzle unit, ensure that accuracy, the reliability of Placement element when industrial camera capturing element image.
Detailed description of the invention
Fig. 1 is the first situation schematic diagram that element image outline includes suction nozzle unit profile；
Fig. 2 is the second situation schematic diagram that element image outline includes suction nozzle unit profile；
Fig. 3 is the third situation schematic diagram that element image outline includes suction nozzle unit profile；
Fig. 4 is the edge contour schematic diagram of element image shown in Fig. 3；
Fig. 5 is that the element image of Fig. 3 uses the schematic diagram of minimum circumscribed rectangle coarse positioning；
Fig. 6 is that the element image of Fig. 3 determines the schematic diagram of best long side and best short side；
Fig. 7 is that the element image of Fig. 3 determines the length schematic diagram of the first side length；
Fig. 8 is that the element image of Fig. 3 determines the length schematic diagram of the second side length；
Fig. 9 is that the element image of Fig. 3 determines the schematic diagram of element profile.
Specific embodiment
It is carried out below with reference to technical effect of the embodiment and attached drawing to design of the invention, specific structure and generation clear
Chu is fully described by, to be completely understood by the purpose of the present invention, feature and effect.Obviously, described embodiment is this hair
Bright a part of the embodiment, rather than whole embodiments, based on the embodiment of the present invention, those skilled in the art are not being paid
Other embodiments obtained, belong to the scope of protection of the invention under the premise of creative work.
The element image-recognizing method that the present invention removes suction nozzle interference includes the following steps:
S1, pre-treatment, including image binaryzation and opening operation are carried out to element image.
Image binaryzation is exactly to set 0 or 255 for the gray value of the pixel on image, that is, be in for whole image
Reveal apparent black and white effect.The binaryzation of image is conducive to being further processed for image, becomes image simply, and data
Amount reduces, and can highlight the profile of interested target.
Opening operation, which refers to, first corrodes the operation expanded again to image using the same structural element, can eliminate member
Small object object in part image, the separating objects at very thin point, the boundary of smooth larger object, and unobvious change element image
Area.
S2, marginal point is extracted by the way of edge detection to element image, and using minimum circumscribed rectangle to described
Marginal point carries out coarse positioning.
As shown in figure 4, element image is surrounded by five lines of a, b, c, d, e, it will be apparent that, the profile of the image contains suction nozzle
At the profile of unit, that is, contour line d.The marginal point of element image can be extracted by the way of edge detection,
It is used so that subsequent step calculates.
As shown in Figure 4 and Figure 5, after extracting marginal point through edge detection, minimum circumscribed rectangle is carried out to edge point
It approaches, as can be seen that two of minimum circumscribed rectangle in a with element image and the side b is substantially straight at one from schematic diagram
On line, and the two other of minimum circumscribed rectangle is tangent in then d substantially with element image.
S3, two candidate long sides and two candidate short sides are determined according to each edge lengths of minimum circumscribed rectangle.
Minimum circumscribed rectangle has four edges, determines two candidate long sides and two candidates according to the length on its each side
Short side.
S4, the number for being less than the point of m pixel is calculated to each back gauge, the largest number of candidate long sides of the point met the requirements are
Best long side, the largest number of candidate's short sides are best short side, 2≤m≤5, m ∈ N*, in the present embodiment, the value of m is preferably 3
It is a.
As shown in fig. 6, determining best long side 11 from two candidate long sides according to the method for S4, then short from two candidates
Best short side 12 is determined in side.
S5, the point for being less than m pixel apart from best long side and best short side is fitted, the friendship of two straight lines of digital simulation
Point obtains the first vertex of element profile.
As shown in fig. 6, preferably the intersection point of long side 11 and best short side 12 after straight line fitting is the of element profile
One vertex 13, in the present embodiment, the fitting of straight line preferably uses least square method.
The distance of S6, all marginal points of statistics to best long side, by calculating y1=f (x1) first derivative at each point,
x1For distance, y1For number, number peak value y is determined1max1And number minor peaks y1max2, by number peak value y1max1And number secondary peak
Value y1max2Corresponding distance x1max1And x1max2It is compared with component size setting value, immediate distance is x1maxAs
The length of first side length of element profile.
Obtain the accurate outer dimension of rectangular element, it is necessary to be accurately judged to the position on its at least three vertex.Such as
Shown in Fig. 7, count all marginal points to best long side 11 distance, theoretically speaking the number peak value y of same distance1maxMost
Big value should appear in x1=L1Position, can find out from schematic diagram, x1In x1maxPosition its apart from best long side 11
Equidistant point is significantly more than x1In x11And x12Position, when by number y1- distance x1Function carry out first derivation it
It is just capable of determining that several number peak values afterwards, by number peak value y1max1And number minor peaks y1max2Corresponding distance x1max1And
x1max2With component size setting value L1It is compared, distance and L1What is be more nearly is the length of the first side length.First side length
Length obtain after, so that it may determine the second vertex 14 of element profile.
The distance of S7, all marginal points of statistics to best short side, by calculating y2=f (x2) first derivative at each point,
x2For distance, y2For number, number peak value y is determined2max1And number minor peaks y2max2, by number peak value y2max1And number secondary peak
Value y2max2Corresponding distance x2max1And x2max2It is compared with component size setting value, immediate distance is x2maxAs
Length of the element profile perpendicular to the second side length of first side length.
As shown in figure 8, it is similar to S6, all marginal points are counted to the distance of best short side 12, determine number peak value
y2max1And number minor peaks y2max2Corresponding distance x2max1And x2max2, then by x2max1And x2max2With component size setting value L2
It is compared, distance and L2What is be more nearly is the length of the second side length.After the length of second side length obtains, so that it may really
Determine the third vertex 15 of element profile.
S8, element profile is determined according to three vertex.
As shown in figure 9, the first vertex 13, the second vertex 14, third vertex 15 position accurately obtain after, it will be able to will
The profile of element is accurately judged, and then realizes the purpose of removal suction nozzle interference.
It is to be illustrated to presently preferred embodiments of the present invention, but the present invention is not limited to the embodiment above,
Those skilled in the art can also make various equivalent deformation or replacement on the premise of without prejudice to spirit of the invention, this
Equivalent deformation or replacement are all included in the scope defined by the claims of the present application a bit.
Claims (5)
1. a kind of element image-recognizing method of removal suction nozzle interference, which is characterized in that comprising steps of
Marginal point is extracted by the way of edge detection to element image, and the edge is clicked through using minimum circumscribed rectangle
Row coarse positioning；
Two candidate long sides and two candidate short sides are determined according to each edge lengths of minimum circumscribed rectangle；
Number of each back gauge less than the point of m pixel is calculated, the largest number of candidate long sides of the point met the requirements are preferably long
Side, the largest number of candidate's short sides are best short side, 2≤m≤5, m ∈ N*；
The point for being less than m pixel apart from best long side and best short side is fitted, the intersection point of two straight lines of digital simulation obtains
First vertex of element profile；
All marginal points are counted to the distance of best long side, pass through calculating y1=f (x1) first derivative at each point, x1For away from
From y1For number, number peak value y is determined1max1And number minor peaks y1max2, by number peak value y1max1And number minor peaks
y1max2Corresponding distance x1max1And x1max2It is compared with component size setting value, immediate distance is x1maxIt is as first
The length of first side length of part profile obtains the second vertex of element profile；
All marginal points are counted to the distance of best short side, pass through calculating y2=f (x2) first derivative at each point, x2For away from
From y2For number, number peak value y is determined2max1And number minor peaks y2max2, by number peak value y2max1And number minor peaks
y2max2Corresponding distance x2max1And x2max2It is compared with component size setting value, immediate distance is x2maxIt is as first
Part profile normal obtains the third vertex of element profile in the length of the second side length of first side length；
Element profile is determined according to three vertex.
2. the element image-recognizing method of removal suction nozzle interference as described in claim 1, it is characterised in that: to element image
Before the step of extracting marginal point, binary conversion treatment is carried out to element image.
3. the element image-recognizing method of removal suction nozzle interference as described in claim 1, it is characterised in that: to element image
Before the step of extracting marginal point, opening operation processing is carried out to element image.
4. the element image-recognizing method of removal suction nozzle interference as described in claim 1, it is characterised in that: in best long side and
Using least square method in the fit procedure of best short side.
5. the element image-recognizing method of removal suction nozzle interference as described in claim 1, it is characterised in that: arrive each side calculating
Distance is less than in the number step of the point of m pixel, and the value of m is 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610913757.9A CN106503720B (en) | 2016-10-19 | 2016-10-19 | A kind of element image-recognizing method of removal suction nozzle interference |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610913757.9A CN106503720B (en) | 2016-10-19 | 2016-10-19 | A kind of element image-recognizing method of removal suction nozzle interference |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106503720A CN106503720A (en) | 2017-03-15 |
| CN106503720B true CN106503720B (en) | 2019-09-13 |
Family
ID=58319253
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610913757.9A Active CN106503720B (en) | 2016-10-19 | 2016-10-19 | A kind of element image-recognizing method of removal suction nozzle interference |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106503720B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111568197A (en) * | 2020-02-28 | 2020-08-25 | 佛山市云米电器科技有限公司 | Intelligent detection method, system and storage medium |
| CN111568199B (en) * | 2020-02-28 | 2023-11-07 | 佛山市云米电器科技有限公司 | Water receiving container identification method, system and storage medium |
| CN111568182A (en) * | 2020-02-28 | 2020-08-25 | 佛山市云米电器科技有限公司 | Intelligent water outlet method, system and storage medium |
| CN113160161B (en) * | 2021-04-14 | 2023-07-11 | 歌尔股份有限公司 | Method and device for detecting defects at edge of target |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003178299A (en) * | 2001-12-07 | 2003-06-27 | Asahi Koyo Kk | Diagram reducing compiling method and device and diagram deforming device and method |
| CN101839690A (en) * | 2010-04-13 | 2010-09-22 | 河海大学常州校区 | Visual inspection method for chip electronic component position error based on edge fitting |
| CN104809422A (en) * | 2015-04-27 | 2015-07-29 | 江苏中科贯微自动化科技有限公司 | QR code recognizing method based on image processing |
| CN104981105A (en) * | 2015-07-09 | 2015-10-14 | 广东工业大学 | Detecting and error-correcting method capable of rapidly and accurately obtaining element center and deflection angle |
| CN105046271A (en) * | 2015-06-25 | 2015-11-11 | 哈尔滨工业大学 | MELF (Metal Electrode Leadless Face) component positioning and detecting method based on match template |
| CN105809705A (en) * | 2016-03-30 | 2016-07-27 | 广东工业大学 | Patch element positioning identification method based on minimum enclosing rectangle |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4525787B2 (en) * | 2008-04-09 | 2010-08-18 | 富士ゼロックス株式会社 | Image extraction apparatus and image extraction program |
-
2016
- 2016-10-19 CN CN201610913757.9A patent/CN106503720B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003178299A (en) * | 2001-12-07 | 2003-06-27 | Asahi Koyo Kk | Diagram reducing compiling method and device and diagram deforming device and method |
| CN101839690A (en) * | 2010-04-13 | 2010-09-22 | 河海大学常州校区 | Visual inspection method for chip electronic component position error based on edge fitting |
| CN104809422A (en) * | 2015-04-27 | 2015-07-29 | 江苏中科贯微自动化科技有限公司 | QR code recognizing method based on image processing |
| CN105046271A (en) * | 2015-06-25 | 2015-11-11 | 哈尔滨工业大学 | MELF (Metal Electrode Leadless Face) component positioning and detecting method based on match template |
| CN104981105A (en) * | 2015-07-09 | 2015-10-14 | 广东工业大学 | Detecting and error-correcting method capable of rapidly and accurately obtaining element center and deflection angle |
| CN105809705A (en) * | 2016-03-30 | 2016-07-27 | 广东工业大学 | Patch element positioning identification method based on minimum enclosing rectangle |
Non-Patent Citations (3)
| Title |
|---|
| Evaluating Edge Detection through Boundary Detection;Song Wang 等;《EURASIP Journal on Applied Signal Processing》;20061231;第2006卷;第1-15页 * |
| 视觉系统在贴片头定位于片状元件检测纠偏中的应用研究;余大伟;《中国优秀硕士学位论文全文数据库 信息科技辑》;20110115;第2011年卷(第01期);第I138-1159页 * |
| 贴片元件几何尺寸精密测量系统的关键技术研究;李小伟;《中国优秀硕士学位论文全文数据库 信息科技辑》;20150115;第2015年卷(第01期);第I138-1235页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106503720A (en) | 2017-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106503720B (en) | A kind of element image-recognizing method of removal suction nozzle interference | |
| CN111044522B (en) | Defect detection method and device and terminal equipment | |
| CN108364311B (en) | Automatic positioning method for metal part and terminal equipment | |
| CN103499297B (en) | A kind of high-precision measuring method based on CCD | |
| CN106824806B (en) | The detection method of low module plastic gear based on machine vision | |
| US20180122083A1 (en) | Method and device for straight line detection and image processing | |
| CN108510476B (en) | Mobile phone screen circuit detection method based on machine vision | |
| CN110189375B (en) | Image target identification method based on monocular vision measurement | |
| CN114494045B (en) | Large spur gear geometric parameter measurement system and method based on machine vision | |
| CN110480127A (en) | A kind of seam tracking system and method based on structured light visual sensing | |
| CN110910359B (en) | FPC circuit detection method based on feature tracking and skeleton feature matching | |
| CN104315978A (en) | Method and device for measuring pipeline end face central points | |
| CN108537772B (en) | Visual detection method for printing defects of chip resistor positive conductor | |
| CN113077437B (en) | Workpiece quality detection method and system | |
| CN114279357A (en) | Die casting burr size measurement method and system based on machine vision | |
| CN112505055B (en) | Method and device for detecting copper leakage of circuit board | |
| CN105678737A (en) | Digital image corner point detection method based on Radon transform | |
| CN113822810A (en) | Method for positioning workpiece in three-dimensional space based on machine vision | |
| CN113705564B (en) | Pointer type instrument identification reading method | |
| CN112734779A (en) | Dot calibration plate edge sub-pixel detection method | |
| CN111476792A (en) | Method for extracting plate strip steel image outline | |
| CN111709954A (en) | Calibration method of go robot vision system | |
| CN113781508B (en) | She Xingkong image contour line extraction and segmentation method, system, equipment and storage medium | |
| CN111325802A (en) | Circular mark point identification matching method in helicopter wind tunnel test | |
| JPH06288732A (en) | Position detection of electronic part |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200911 Address after: 518000 Fourth Floor, First Floor and Third Floor of Block G, 52 Huangpu Road, Shangliao Community, Xinqiao Street, Baoan District, Shenzhen City, Guangdong Province Patentee after: SHENZHEN FAROAD INTELLIGENT EQUIPMENT Co.,Ltd. Address before: 518000,, Guangdong, Baoan District province Shenzhen manhole street, community 52, Whampoa Road, two floor Patentee before: SHENZHEN LUYUAN AUTOMATION EQUIPMENT Co.,Ltd. |