KR101754089B1 - Method for updating alignment mark of vision system in real time - Google Patents

Method for updating alignment mark of vision system in real time Download PDF

Info

Publication number
KR101754089B1
KR101754089B1 KR1020150150065A KR20150150065A KR101754089B1 KR 101754089 B1 KR101754089 B1 KR 101754089B1 KR 1020150150065 A KR1020150150065 A KR 1020150150065A KR 20150150065 A KR20150150065 A KR 20150150065A KR 101754089 B1 KR101754089 B1 KR 101754089B1
Authority
KR
South Korea
Prior art keywords
mark
subject
matching rate
reference mark
vision system
Prior art date
Application number
KR1020150150065A
Other languages
Korean (ko)
Other versions
KR20170049113A (en
Inventor
배익순
채승수
이상민
김장섭
배성민
배정윤
Original Assignee
주식회사 비엔에스
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 주식회사 비엔에스 filed Critical 주식회사 비엔에스
Priority to KR1020150150065A priority Critical patent/KR101754089B1/en
Publication of KR20170049113A publication Critical patent/KR20170049113A/en
Application granted granted Critical
Publication of KR101754089B1 publication Critical patent/KR101754089B1/en

Links

Images

Classifications

    • G06K9/3216—
    • G06K9/6201—
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G06K2009/3225—
    • G06K2209/19—

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Eye Examination Apparatus (AREA)

Abstract

The present invention relates to a method for real-time updating an alignment mark of a vision system, comprising a reference mark registration step (S10) of photographing and registering a reference mark in a state in which an object is aligned for correcting the position of the object, A matching rate calculation step (S30) of calculating a matching rate by comparing a mark of the object to be inspected recognized in the mark recognition step (S20) with a reference mark and recognizing the mark of the object by photographing the object to be inspected; ).

Description

[0001] The present invention relates to a method for updating an alignment mark of a vision system,

The present invention relates to a method for real-time updating of an alignment mark of a vision system, wherein a matching rate of a mark of an object to be inspected and an existing registered reference mark is calculated and the mark of the object is re- And more particularly to a method for real-time updating an alignment mark of a vision system.

Generally, a vision system used for aligning an object in an industrial field captures an object to be inspected using a camera, compares the captured image with a previously registered reference image, and aligns the position and direction of the object Respectively.

Korean Patent Registration No. 10-0801665 relates to an alignment mark recognition machine vision system and an alignment mark recognition method, which comprises the steps of converting an input image into a binary image in which a mark object area and a background are separated from each other, And labeling the first feature point pixels corresponding to the first mark image and the second feature point pixels corresponding to the second mark image in the case where the number of labeled areas in the mark object area is one, And determining a center point of the mark image and a center point of the second mark image.

However, in the case of the above-described conventional technique, since the mark can be recognized only when the mark of the object to be inspected is the same as the reference mark, the mark is not recognized when the mark is dirty or deformed.

In addition, when the shape of a part of the object to be inspected is changed, the reference mark must be registered manually.

In addition, when the position where the subject is stopped is not constant, it is difficult to recognize the mark of the subject, and it is difficult to correct the mark based on the reference mark.

Korean Patent Registration No. 10-0801665

It is an object of the present invention to solve the above problems and provide a method for real-time updating of an alignment mark of a vision system for re-registering a mark of an object to be inspected with a reference mark when a mark of an object to be inspected has a low matching rate with the reference mark .

Another object of the present invention is to provide a method for real-time updating of a reference mark and an alignment mark of a vision system which increases the recognition rate of the mark of the inspection subject by comparing the start position of the image with the unit pixel interval.

It is another object of the present invention to provide a method for real time updating of an alignment mark of a vision system capable of correcting a position to be identical to a reference mark after recognizing the position of an object to be inspected.

Another object of the present invention is to provide a method for real-time updating of an alignment mark of a vision system capable of recognizing a mark even if the stop position is changed by re-registering the mark of the subject with the reference mark when the stop position of the test subject is changed.

A method for real-time updating of an alignment mark of a vision system according to the present invention includes: a reference mark registration step (S10) for photographing and registering a reference mark in a state in which an object is aligned for correcting the position of the object; (S20) for recognizing a mark of the subject by capturing an object to be inspected, a matching rate calculating unit (S20) for comparing the mark of the subject and the reference mark recognized in the mark recognizing step (S20) And a calculation step (S30).

If the matching rate calculated in the matching rate calculation step S30 of the method of real-time updating the alignment mark of the vision system of the present invention is 60 to 100%, the reference mark re-registration step of re-registering the mark of the test object with the reference mark (S50).

The mark recognizing step (S20) of the method of real-time updating the alignment mark of the vision system of the present invention compares the reference mark with the mark of the object to be inspected while moving the reference mark from the starting point of the object, And searching for an identical image.

In the matching rate calculation step S30 of the method for real-time updating the alignment marks of the vision system of the present invention, the reference marks and the marks of the subject are compared on a pixel-by-pixel basis, and the pixel brightness difference And the matching rate is calculated.

In the matching rate calculation step S30 of the method of real-time updating the alignment mark of the vision system of the present invention, the reference mark and the subject mark are converted into gray scales, and using the pixel color difference at the position corresponding to the two marks And a matching rate is calculated.

Also, when the matching rate of the reference mark and the subject mark calculated in the matching rate calculating step S30 of the method of real-time updating the alignment mark of the vision system of the present invention is less than 60%, it is determined as bad and NG Further comprising a step S60.

Also, when the matching rate calculated in the matching rate calculation step S30 of the method of real-time updating the alignment mark of the vision system of the present invention is 60 to 100%, the subject is moved or rotated in the same manner as the reference mark, (S40). ≪ IMAGE >

As described above, according to the method of real-time updating of the alignment mark of the vision system according to the present invention, when the mark of the object to be inspected has a low matching rate with the reference mark, there is an effect of re-registering the mark of the object to be inspected with the reference mark .

According to the method for real-time updating of the alignment mark of the vision system according to the present invention, the recognition rate of the reference mark and the mark of the subject can be increased by comparing the unit pixel interval from the beginning of the image.

In addition, according to the method of real-time updating the alignment mark of the vision system according to the present invention, it is possible to correct the position so as to be identical to the reference mark after recognizing the position of the object to be inspected.

According to the method for real-time updating the alignment mark of the vision system according to the present invention, if the position at which the subject is stopped is changed, the mark of the subject can be recognized again even if the stop position is changed by re-

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a flow chart illustrating in block form a method for real time updating an alignment mark of a vision system according to the present invention.
2 is a view showing a reference mark and a mark of an object to be inspected in a method of real-time updating of an alignment mark of a vision system according to the present invention.
3 is a view illustrating a method of real-time updating of an alignment mark of a vision system according to the present invention, in which pixels of an original image and a comparison image are compared.

Specific features and advantages of the present invention will be described in detail below with reference to the accompanying drawings. The detailed description of the functions and configurations of the present invention will be omitted if it is determined that the gist of the present invention may be unnecessarily blurred.

The present invention relates to a method for real-time updating of an alignment mark of a vision system, wherein a matching rate of a mark of an object to be inspected and an existing registered reference mark is calculated and the mark of the object is re- And more particularly to a method for real-time updating an alignment mark of a vision system.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of a method for real-time updating an alignment mark of a vision system according to the present invention.

As shown in FIG. 1, a method of real-time updating an alignment mark of a vision system according to the present invention includes a reference mark registration step of photographing and registering a reference mark in a state in which an object is aligned for correcting the position of the object to be examined (S20) for recognizing a mark of the object to be inspected by capturing an object to be inspected; a step of calculating a matching rate by comparing the mark of the object to be inspected recognized in the mark recognizing step (S20) with a reference mark And a matching rate calculation step (S30).

And a reference mark re-registration step (S50) of re-registering the mark of the test subject with the reference mark when the matching rate calculated in the matching rate calculating step (S30) is 60 to 100%.

And a NG step (S60) in which it is determined that the matching ratio between the reference mark and the inspection object mark calculated in the matching rate calculation step (S30) is less than 60% and is separately determined in the process.

And a correction step (S40) of correcting the position by moving or rotating the position of the subject in the same manner as the reference mark when the matching rate calculated in the matching rate calculation step (S30) is 60 to 100% do.

In the reference mark registration step (S10), the image of the subject is photographed with the subject's position at the correct position in the process to acquire the image including the mark, and the mark of the subject positioned at the correct position is the position Is registered as a reference mark so as to be a reference that can be corrected.

At this time, the reference mark may be the shape of the subject in the original image for photographing the subject, and the mark of the subject to be inspected through the reference mark is recognized and the position is corrected.

The mark recognizing step S20 is a step of capturing an image of an object to be inspected and acquiring an image, and comparing whether there is a similar mark through the reference mark registered in the reference mark registering step S10 to determine whether or not the mark exists.

At this time, the reference mark moves and rotates in units of pixels from the start point of the photographed image of the subject, and the section where the mark of the test subject similar to the reference mark exists is grasped.

Also, if the mark of the subject is not recognized in the mark recognition step S20, the process may proceed to the NG step S60 so that the alarm sound may be generated so that the user can grasp the mark.

A detailed description of the mark recognition step S20 will be described later with reference to FIG.

The matching rate calculation step S30 is a step of determining a degree of similarity of the mark of the subject recognized in the mark recognition step S20 with the reference mark and superimposing the reference mark on the recognized mark image of the subject The reference mark is moved or rotated so that the reference mark and the inspection object mark are matched with each other, and then the reference mark and the mark of the inspection object are compared on a pixel basis to calculate the matching rate.

At this time, when the matching rate of the mark of the subject and the reference mark is less than 60%, it is determined that the subject is different from the reference mark and the process proceeds to NG step S60. If the matching rate is within the 60 to 100% (S40) in which the position is corrected in the same manner as the reference mark.

When the matching rate is less than 60%, the mark recognition is remarkably deteriorated. Therefore, the position correction may not be normally performed. When the matching rate is 60% or more, the mark of the subject is recognized more than half.

A detailed method of calculating the matching rate will be described later with reference to FIG.

In the NG step S60, if the matching rate of the mark of the test subject and the reference mark is less than 60%, it is determined that the mark is defective and the subject is separately collected in the process so that the user can confirm the subject.

Since the mark matching rate differs from the reference mark due to damage, contamination, or deformation of the mark, the operator confirms the inspection object through the full inspection and then reinserts the process or judges it as defective.

In the NG step S60, when the NG occurs continuously two or three times, the alarm is outputted while the process is stopped, so that the operator can take action.

In the correction step S40, when the matching rate is between 60% and 100%, the subject is rotated in the same form as the reference mark, and the subject is aligned in the correct position in the process.

It is possible to correct the position by using an alignment capable of aligning the object to be inspected for rotating the object and a gripper for detachably attaching the object to be inspected and rotating or moving the object.

In order to determine the amount of movement and the amount of rotation for correcting the position of the object in the correction step S40, the amount of movement and rotation of the reference mark to match the mark of the object is measured, .

The reference mark re-registration step (S50) is a step of photographing the mark of the object to be inspected and re-registering it with the reference mark.

Since the subject may be aligned at a position different from the initial setting value when the subject is continuously drawn, it is necessary to update the reference mark because the matching rate may decrease if the reference mark is continuously used.

In addition, if the mark of the subject is different from the reference mark due to damage, contamination, or deformation, the matching rate is lowered and error occurs frequently. In order to increase the matching rate, the equipment should be temporarily stopped And the productivity is deteriorated.

If the matching rate calculated in the matching rate calculation step S30 is between 60% and 100% in the reference mark re-registration step S50, the mark of the current test subject is photographed and then re-registered in real time with the reference mark This is used to check the mark and matching rate of the object to be inspected next time.

At this time, if the matching rate is 60% or more, the mark of the subject can be recognized more than half, and the position of the subject can be corrected. If the matching ratio is less than 60%, the matching rate is remarkably decreased.

When the mark of the object to be inspected is photographed and re-registered with the reference mark in the reference mark re-registration step (S50), the originally registered reference mark is not deleted but stored separately.

If the matching rate between the mark of the object to be inspected and the reference mark falls to less than 60%, it is retried with the reference mark registered first, and if the matching rate falls equally to less than 60%, it is sent to the NG step (S60) The reference mark of the currently registered subject may be deleted after the process is sent to the correction step S40 and the first reference mark may be used.

In addition, when the mark of the subject is re-registered more than twice as the reference mark, the mark of the imported subject is re-registered to calculate the matching rate with the reference mark, and the matching rate with the first registered reference mark is recalculated so that the matching rate is less than 50% The process may be sent to the NG step S60.

Since the re-registered reference mark is a state in which the mark of the subject is damaged, contaminated, or deformed, the degree of actual recognition may be different when compared with the originally registered reference mark. Therefore, It is possible to judge NG by recalculating the matching rate.

At this time, if the matching rate of the first reference mark and the mark of the subject is less than 50%, the mark may not be normally recognized and the correction may not be performed properly.

Also, since the mark of the object to be inspected may not always stop at the same position, even if the mark of the inspection object does not suffer damage, contamination, or deformation, the mark of the inspection object may be photographed and registered as the reference mark .

In this case, since the position at which the object to be inspected is stopped can be grasped, the reference mark can easily recognize the mark position of the inspection object, and the recognition speed can be improved.

When the mark of the subject is re-registered with the reference mark, the position of the subject is corrected through the correction step (S40), and the mark of the subject is photographed and re-registered in a state where the subject is in the correct position in the process, The position of the object to be inspected can be corrected.

FIG. 2 is a view showing a reference mark 110 and a mark 210 of a subject in a method of real-time updating of an alignment mark of a vision system according to the present invention.

As shown in FIG. 2, the mark recognition step S20 (see FIG. 1) of the method of real-time updating the alignment mark of the vision system according to the present invention includes moving the reference mark 110 from the starting point of the subject, And searches for an image in which the reference mark 110 and the mark 210 of the inspection object coincide with the mark 210 of the inspection target object.

 The mark recognition step of FIG. 1 includes the step of superimposing the original image 100 of the reference mark 110 on the comparison image 200 of the mark 210 of the subject and comparing the original image 100 with the comparison image 200, And searches for a portion where an image similar to the reference mark 110 exists.

The reference image 110 of the original image 100 and the reference image 200 of the original image 100 are moved while the original image 100 is moved and rotated when a portion similar to the reference mark 110 formed on the original image exists in the comparison image 200. [ The mark 210 of the object of inspection is matched as much as possible.

If the reference mark 110 of the original image 100 and the mark 210 of the subject of the comparison image 200 match as much as possible, it is determined that the mark is recognized and then the process goes to the matching rate calculation step S30 The mark 210 of the subject of the comparison image 200 is searched and matched in the same manner even when the mark 210 of the subject of the comparison image 200 is damaged and the reference mark 110 is re-registered The process proceeds to the matching rate calculation step S30 in Fig.

3 is a view illustrating a method of comparing pixels of an original image 100 and a comparison image 200 in a method of real-time updating an alignment mark of a vision system according to the present invention.

As shown in FIG. 3, the matching rate calculation step S30 (see FIG. 1) of the method for real-time updating the alignment mark of the vision system according to the present invention includes a reference mark 110 (see FIG. 2) (See FIG. 2) of each pixel on the basis of a difference in pixel brightness between positions corresponding to the two marks, and calculates a matching rate.

When the reference mark 110 of the original image 100 of FIG. 2 and the mark 210 of the subject of the comparison image 200 coincide with each other, the matching rate calculation step S30 of FIG. 1 compares the mark The degree of similarity between the reference mark 110 and the mark 210 of the test subject is calculated.

For this, the mark pixels 220 of the subject formed at the same position as the reference mark pixel 120 are compared with each other, and the brightness of light generated in the pixels are compared with each other to determine whether the two marks are identical to each other.

In this case, the matching rate calculation step S30 of FIG. 1 expresses the brightness values of the reference mark pixel 120 and the mark pixel 220 of the subject in a range of 0 to 255, The brightness value of the mark pixel 220 of the object to be inspected is subtracted from the brightness value of the mark pixel 220, and if the value is within the range of 0 to 10, it is determined that the pixels are the same.

The pixel comparison result value 300 can be obtained by comparing the brightness values of the reference mark pixel 120 and the mark pixel 220 of the subject in FIG. 2, and the matching rate calculation step S30 of FIG. Calculates the matching rate in% by calculating how much the value in the range of 0 to 10 occupies in the total pixels in the pixel comparison result value 300. [

In addition, the matching rate calculation step S30 of FIG. 1 converts the reference mark 110 of FIG. 2 and the mark 210 of the subject of FIG. 2 into gray scale, and uses the pixel color difference of positions corresponding to the two marks And the matching rate is calculated.

The gray scale represents the range of gradual steps of gray between white and black, in which white is represented by 0, black represented by 255, and gray formed between white and black by 1 to 254, And the mark 210 of the subject of FIG.

Therefore, the pixel comparison result value 300 can be obtained through the difference between the reference mark 110 of FIG. 2 and the mark 210 of the subject of FIG. 2 represented by numerals through gray scale, and the pixel comparison result value ( 300), the rate of the portion included in the range of 0 to 10 can be calculated and the matching rate can be expressed.

As described above, according to the method of real-time updating the alignment mark of the vision system according to the present invention, when the mark of the object to be inspected is less than the matching rate with the reference mark, the mark of the object is re- The recognition rate of the reference mark and the mark of the inspection object is increased and the position of the inspected object can be corrected so as to be equal to the reference mark after the position of the inspected object is recognized, There is an effect that the mark recognition can be performed even if the position to be stopped is changed by re-registering the mark of the test subject with the reference mark.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is clearly understood that the same is by way of illustration and example only and is not to be taken as a limitation of the scope of the present invention. Or modify it. The scope of the invention should, therefore, be construed in light of the claims set forth to cover many of such variations.

Step S10: Registration of reference mark
S20: Mark recognition step
S30: matching rate calculation step
S40: correction step
S50: Re-registration step of reference mark
S60: NG step
100: Original video
110: Reference mark
120: reference mark pixel
200: Comparative video
210: Mark of subject
220: mark pixel of the subject
300: Pixel comparison result

Claims (7)

A method for real-time updating of an alignment mark of a vision system,
A reference mark registration step (S10) of photographing and registering the reference mark in a state in which the inspection object is aligned to correct the position of the inspection object;
A mark recognizing step (S20) of recognizing a mark of the subject by photographing the subject to be inspected;
A matching rate calculation step (S30) of comparing the mark of the subject and the reference mark recognized in the mark recognition step (S20) and calculating a matching rate;
A reference mark re-registration step (S50) of re-registering the mark of the test subject with the reference mark when the matching rate calculated in the matching rate calculating step (S30) is 60 to 100%;
And an NG step (S60) in which it is determined that the match rate of the reference mark calculated in the matching rate calculation step (S30) is less than 60%
When the mark of the object to be inspected is registered in the reference mark re-registration step (S50) and re-registered to the reference mark, the first registered reference mark is not deleted but stored separately,
The matching rate calculation step S30 recalculates the matching rate of the subject using the reference mark registered first when the matching rate of the mark of the subject to be inspected and the re-registered reference mark is less than 60%
If the matching rate of the subject and the reference mark registered first is less than 60%, the subject is sent to the NG stage. If the matching rate is 60% or more, the reference mark re-registration step (S50) And re-registering the reference mark registered first.
How to update the alignment mark of the vision system in real time.
delete The method according to claim 1,
Wherein the mark recognition step (S20) compares the reference mark with a mark of the subject while moving and rotating the reference mark from the starting point of the subject, and searches for an image in which the mark of the subject matches the mark of the subject
How to update the alignment mark of the vision system in real time.
The method according to claim 1,
The matching rate calculation step (S30) compares the reference mark and the mark of the subject with each other on a pixel basis, and calculates the matching rate using the pixel brightness difference at the position corresponding to the two marks
How to update the alignment mark of the vision system in real time.
The method according to claim 1,
The matching rate calculation step S30 is characterized in that the reference mark and the subject mark are converted into gray scales and the matching rate is calculated using the pixel color difference at the position corresponding to the two marks
How to update the alignment mark of the vision system in real time.
delete The method according to claim 1,
And a correction step (S40) of correcting the position by moving or rotating the position of the subject in the same manner as the reference mark when the matching rate calculated in the matching rate calculation step (S30) is 60 to 100% To
How to update the alignment mark of the vision system in real time.
KR1020150150065A 2015-10-28 2015-10-28 Method for updating alignment mark of vision system in real time KR101754089B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020150150065A KR101754089B1 (en) 2015-10-28 2015-10-28 Method for updating alignment mark of vision system in real time

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020150150065A KR101754089B1 (en) 2015-10-28 2015-10-28 Method for updating alignment mark of vision system in real time

Publications (2)

Publication Number Publication Date
KR20170049113A KR20170049113A (en) 2017-05-10
KR101754089B1 true KR101754089B1 (en) 2017-07-19

Family

ID=58744205

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020150150065A KR101754089B1 (en) 2015-10-28 2015-10-28 Method for updating alignment mark of vision system in real time

Country Status (1)

Country Link
KR (1) KR101754089B1 (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011095226A (en) * 2009-11-02 2011-05-12 Saki Corp:Kk Inspection device of material to be inspected, and position correction device of electronic substrate

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011095226A (en) * 2009-11-02 2011-05-12 Saki Corp:Kk Inspection device of material to be inspected, and position correction device of electronic substrate

Also Published As

Publication number Publication date
KR20170049113A (en) 2017-05-10

Similar Documents

Publication Publication Date Title
US9292915B2 (en) Digital optical comparator
CN112304954B (en) Part surface defect detection method based on line laser scanning and machine vision
JP2003254748A (en) Stereo image characteristic inspection system
US20090046922A1 (en) Surface Inspecting Apparatus
CN112589401B (en) Assembling method and system based on machine vision
US20140293035A1 (en) Image processing apparatus, imaging apparatus, microscope system, image processing method, and computer-readable recording medium
US6490369B1 (en) Method of viewing and identifying a part for a robot manipulator
CN109916914B (en) Product defect detection method and device
US20190033728A1 (en) Measurement device, method and display device
CN108871185B (en) Method, device and equipment for detecting parts and computer readable storage medium
CN109227352B (en) Substrate defect repairing method and system
CN114494462A (en) Binocular camera ranging method based on Yolov5 and improved tracking algorithm
JP2005265661A (en) Image processing method and image processing device
CN109425327B (en) Inspection system and method for correcting inspection image
KR101754089B1 (en) Method for updating alignment mark of vision system in real time
CN108520499B (en) Image offset correction method based on white-band microscopic imaging
CN103905762B (en) Method for automatically detecting projection picture for projection module
US20110032352A1 (en) Imaging position determining method and imaging position determining device
CN107860933B (en) Digital image-based automatic detection method and device for fiber content in textile
KR20090114565A (en) Wafer location error indicating system and method using vision sensor
JP2009250777A (en) Surface inspection device and surface inspection method
US11102920B2 (en) Component mounting device and position recognition method
KR20160137020A (en) Method for sensing of fiducial Mark
CN110689491B (en) Method for correcting side image of cheese under machine vision
CN111380869B (en) Optical detection system and method with height information

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E701 Decision to grant or registration of patent right