CN105457908B - The sorting method for rapidly positioning and system of small size glass panel based on monocular CCD - Google Patents
The sorting method for rapidly positioning and system of small size glass panel based on monocular CCD Download PDFInfo
- Publication number
- CN105457908B CN105457908B CN201510771640.7A CN201510771640A CN105457908B CN 105457908 B CN105457908 B CN 105457908B CN 201510771640 A CN201510771640 A CN 201510771640A CN 105457908 B CN105457908 B CN 105457908B
- Authority
- CN
- China
- Prior art keywords
- row
- coordinate
- glass panel
- image
- gray level
- 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.)
- Expired - Fee Related
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/34—Sorting according to other particular properties
- B07C5/342—Sorting according to other particular properties according to optical properties, e.g. colour
- B07C5/3422—Sorting according to other particular properties according to optical properties, e.g. colour using video scanning devices, e.g. TV-cameras
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention is suitable for glass panel and positions, there is provided a kind of sorting method for rapidly positioning based on monocular CCD small size glass panels, step include：A, the card slot image of glass panel is placed with using monocular CCD collections, then the image collected progress gradation conversion is pre-processed to obtain gray level image；B, calculates the row pixel grey scale average of gray level image, determines that glass panel is expert at coordinate；C, binarization segmentation is carried out to gray level image, and area-of-interest edge coordinate is extracted according to glass panel；D, determines card slot centre coordinate according to row coordinate and the area-of-interest edge coordinate, obtains absorption position, control machinery hand reaches absorption position and adsorbed.The present invention is based on gradation conversion and edge detection, the center for every piece of glass panel for being located at card slot in the visual field can be quickly found, in combination with camera fields of view center position, the sorting fast positioning of small size panel with fast searching to the glass panel for being currently needed for crawl, can be realized.
Description
Technical field
The invention belongs to framing field, more particularly to a kind of sorting of the small size glass panel based on monocular CCD
Method for rapidly positioning and system.
Background technology
At present, localization method is broadly divided into machinery positioning and machine vision positions two major classes, and machinery positioning is fairly simple,
But adaptivity is not high, the glass panel that differs especially for size, and machine vision method positioning accuracy is high, speed
Degree is fast, adaptive high and untouchable, can meet detection in real time, thus application is more and more wider.According to CCD quantity, machine regards
Feel that localization method can be divided into monocular vision positioning and multi-vision visual localization method；According to object space dimensionality, and two can be divided into
Tie up localization method and three-dimensional fix method.Multi-vision visual localization method is usually used in more complicated Spatial Multi-Dimensional positioning.But
It is at present when carrying out sorting fast positioning using monocular CCD, the machinery positioning adaptivity of manipulator is not high, easily contacts glass
Glass panel causes panel to scratch.
The content of the invention
The technical problems to be solved by the invention are to provide a kind of sorting of the small size glass panel based on monocular CCD
Method for rapidly positioning and system, it is intended to when carrying out sorting fast positioning using monocular CCD, the machinery positioning adaptivity of manipulator
It is not high, easily contact the problem of glass panel causes panel to scratch.
The present invention is achieved in that a kind of sorting method for rapidly positioning of the small size glass panel based on monocular CCD,
Step includes：
Step A, the image of the card slot of glass panel is placed with using monocular CCD collections, and the image collected is carried out
Gradation conversion, the image for then obtaining gradation conversion are pre-processed to obtain gray level image；
Step B, calculates the row pixel grey scale average of the gray level image, then according to determining the row gray average
Glass panel is expert at coordinate；
Step C, binarization segmentation is carried out to the gray level image, and area-of-interest edge is extracted according to the glass panel
Coordinate；
Step D, determines card slot centre coordinate, with the card according to the row coordinate and the area-of-interest edge coordinate
Absorption position of the groove center coordinate as the glass panel, then control machinery hand reach the absorption position of the glass panel
Adsorbed.
Further, step A is specifically included：
Step A1, control machinery hand are moved to above rack card slot, and control the monocular CCD being fixed on the manipulator
Camera collection is placed with the image of the card slot of glass panel；
Step A2, carries out gradation conversion to the image of step A1 collections, gray level image is obtained after then being pre-processed；Institute
Stating pretreatment includes filtering, denoising.
Further, step B is specifically included：
Step B1, calculates the summation of the gray value of every one-row pixels of the gray level image；
The i-th row of gray level image jth row are represented with I (i, j), r represents the height of the gray level image, and c represents the ash
The width of image is spent, Row (i) represents the summation of the gray value of the i-th row of gray level image pixel, then:Wherein 0≤i≤r, 0≤j≤c；
Step B2, according to the summation of the gray value of every one-row pixels, calculates per a line pixel grey scale average；
I-th row grey scale pixel value average is represented with RowAve (i), then：
RowAve (i)=Row (i)/c；
Step B3, row pixel grey scale maximum is found according to every a line pixel grey scale average, very big with the row pixel grey scale
Value confirms that the glass panel is expert at coordinate.
Further, step B3 is specifically included：
Step B31, calculates the row pixel grey scale average of the gray level image；
The row pixel grey scale average of the gray level image is represented with RowAverage, then：
Step B32, calculates the gray scale difference value per one-row pixels；
The gray scale difference value of the i-th row pixel is represented with Delta (i), then the row pixel deviates the row pixel grey scale average
Size is：Delta (i)=RowAve (i)-RowAverage；
Step B33, travels through per a line pixel grey scale average, obtains maximum, and threshold value setting is carried out with the maximum；
The threshold value is represented with Delta, MaxRowAve represents the maximum, then：
Delta=(MaxRowAve-RowAverage) * 0.8；
Step B34, judges whether the gray scale difference value meets the threshold value, and gray scale maximum is determined according to judging result,
The coordinate so that it is determined that glass panel is expert at；
If Delta ＜ Delta (i), it is determined that the row where the i-th behavior gray scale maximum, i.e., described glass panel place
Row, and obtain the glass panel and be expert at coordinate.
Further, the step C is specifically included：
Step C1, carries out binarization segmentation processing to the gray level image, obtains binarization of gray value image；
Step C2, carries out BLOB analyses to the binarization of gray value image, obtains region of interest area image；
Step C3, carries out edge extracting to the region of interest area image, area-of-interest is obtained according to the edge of extraction
Edge coordinate；
The leftmost edge extreme coordinates of the i-th row of the region of interest area image are represented with ColGrayVa1 (i),
ColGrayVa2 (i) represents the right edge point coordinate of the i-th row of the region of interest area image, then：
The region of interest area image is begun stepping through from a left side to the i-th row, is jumped out when meeting ColGrayVa1 (i)=255
Circulation, records the point coordinates and is begun stepping through from the right side；When meeting ColGrayVa2 (i)=255, the point coordinates is recorded, it is then right
I+1 begins stepping through the region of interest area image from a left side.
Further, step D is specifically included：
Step D1, according to the row coordinate and the area-of-interest edge coordinate, determines the card slot center of the row；
The card slot center of the i-th row is represented with RowCenter (i), the row center of the i-th row is represented with ColCenter (i), with
ColGrayVa1 (i) represents the leftmost edge extreme coordinates of the i-th row of the region of interest area image, ColGrayVa2 (i) tables
Show the right edge point coordinate of the i-th row of the region of interest area image, then：
RowCenter (i)=i；
ColCenter (i)=(ColGrayVa1 (i)+ColGrayVa2 (i))/2；
Step D2, the adsorption potential of the glass panel is determined according to camera fields of view centre coordinate and the card slot centre coordinate
Put；
The row coordinate at camera fields of view center is represented with CameraRowCenter, CameraColCenter represents camera fields of view
The row coordinate at center, r represent the height of the gray level image, and c represents the width of the gray level image, then：
CameraRowCenter=r/2；CameraColCenter=c/2；
And if only if meet abs (CameraRowCenter (i)-RowCenter (i)) and
When abs (CameraColCenter (i)-ColCenter (i)) is minimum, the behavior glass to be adsorbed is determined
Panel is expert at, and is expert at coordinate with the row, the three-dimensional coordinate of combining camera, obtains the absorption position of the glass panel；
Step D3, control machinery hand reach the absorption position and are adsorbed.
Present invention also offers a kind of sorting quick positioning system of the small size glass panel based on monocular CCD, including：
Acquisition process unit, the image of the card slot for being placed with glass panel using monocular CCD collections, and will collect
Image carry out gradation conversion, the image for then obtaining gradation conversion pre-processed to obtain gray level image；
Computing unit, for calculating the row pixel grey scale average of the gray level image, then according to the row gray average
Determine that the glass panel is expert at coordinate；
Edge extracting unit, for carrying out binarization segmentation to the gray level image, extracts according to the glass panel and feels
Interest region edge coordinate；
Absorbing unit is positioned, for determining that card slot center is sat according to the row coordinate and the area-of-interest edge coordinate
Mark, using the card slot centre coordinate as the absorption position of the glass panel, then control machinery hand reaches the glass surface
The absorption position of plate is adsorbed.
Further, the acquisition process unit is specifically used for：
First, control machinery hand is moved to above rack card slot, and controls the monocular CCD phases being fixed on the manipulator
Machine collection is placed with the image of the card slot of glass panel；
Finally, gradation conversion is carried out to the image of collection, gray level image is obtained after then being pre-processed；The pretreatment
Including filtering, denoising.
Further, the computing unit is specifically used for：
First, the summation of the gray value of every one-row pixels of the gray level image is calculated；
The i-th row of gray level image jth row are represented with I (i, j), r represents the height of the gray level image, and c represents the ash
The width of image is spent, Row (i) represents the summation of the gray value of the i-th row of gray level image pixel, then:Wherein 0≤i≤r, 0≤j≤c；
Secondly, according to the summation of the gray value of every one-row pixels, calculate per a line pixel grey scale average；
I-th row grey scale pixel value average is represented with RowAve (i), then：
RowAve (i)=Row (i)/c；
Finally, row pixel grey scale maximum is found according to every a line pixel grey scale average, with the row pixel grey scale maximum
Confirm that the glass panel is expert at coordinate.
Further, positioning absorbing unit is specifically used for：
First, according to the row coordinate and the area-of-interest edge coordinate, the card slot center of the row is determined；
The card slot center of the i-th row is represented with RowCenter (i), the row center of the i-th row is represented with ColCenter (i), with
ColGrayVa1 (i) represents the leftmost edge extreme coordinates of the i-th row of the region of interest area image, ColGrayVa2 (i) tables
Show the right edge point coordinate of the i-th row of the region of interest area image, then：
RowCenter (i)=i；
ColCenter (i)=(ColGrayVa1 (i)+ColGrayVa2 (i))/2；
Secondly, the adsorption potential of the glass panel is determined according to camera fields of view centre coordinate and the card slot centre coordinate
Put；
The row coordinate at camera fields of view center is represented with CameraRowCenter, CameraColCenter represents camera fields of view
The row coordinate at center, r represent the height of the gray level image, and c represents the width of the gray level image, then：
CameraRowCenter=r/2；CameraColCenter=c/2；
And if only if meet abs (CameraRowCenter (i)-RowCenter (i)) and
When abs (CameraColCenter (i)-ColCenter (i)) is minimum, the behavior glass to be adsorbed is determined
Panel is expert at, and is expert at coordinate with the row, the three-dimensional coordinate of combining camera, obtains the absorption position of the glass panel；
Finally, control machinery hand reaches the absorption position and is adsorbed.
Compared with prior art, the present invention beneficial effect is：The present invention is established in monocular vision positioning and two-dimensional localization
On the basis of method, based on gradation conversion and edge detection, every piece of glass panel for being located at card slot in the visual field can be quickly found
Center, can be real with fast searching to the glass panel for being currently needed for crawl in combination with camera fields of view center position
The sorting fast positioning of existing small size panel.Further, the present invention utilizes machine vision, avoids machinery positioning because contacting glass
Panel and caused by secondary scuffing, while can be adjusted according to card slot error, automation is realized for glass panel detection, soon
Speed sorting, the present invention can adaptive a variety of panel models.
Brief description of the drawings
Fig. 1 is a kind of sorting fast positioning side based on monocular CCD small size glass panels provided in an embodiment of the present invention
The flow chart of method.
Fig. 2 is the gray scale schematic diagram of charging rack provided in an embodiment of the present invention.
Fig. 3 is the row gray value schematic diagram of charging rack provided in an embodiment of the present invention.
Fig. 4 is area-of-interest edge provided in an embodiment of the present invention schematic diagram.
Fig. 5 is panel absorption position schematic diagram provided in an embodiment of the present invention.
Fig. 6 is a kind of sorting fast positioning system based on monocular CCD small size glass panels provided in an embodiment of the present invention
The structure diagram of system.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to the accompanying drawings and embodiments, it is right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Monocular vision positioning is built upon based on monocular CCD small sizes glass panel sorting method for rapidly positioning and two dimension is determined
On the basis of the method for position, based on gradation conversion and edge detection, the center for every piece of glass panel for being located at card slot is found in the visual field in
Position, and combining camera visual field center position finds the glass panel currently captured.Its Integral Thought is to phase on manipulator
The gray level image of machine crawl is projected into every trade, calculates row gray average, is extracted glass panel according to maximum and is expert at, to artwork
Binaryzation is carried out, extracts area-of-interest, edge extracting is carried out to the region, and then obtains edge coordinate, with reference to where panel
Row coordinate, obtains the row and column where panel, the combining camera visual field and coordinate, obtains card slot center where currently sorting panel and sits
Mark, is sorted.
Based on above-mentioned theory, the present invention proposes a kind of point based on monocular CCD small size glass panels as shown in Figure 1
Method for rapidly positioning is picked, step includes：
S1, the image of the card slot of glass panel is placed with using monocular CCD collections, and the image collected is carried out gray scale
Conversion, the image for then obtaining gradation conversion are pre-processed to obtain gray level image；
S2, calculates the row pixel grey scale average of the gray level image, then determines the glass according to the row gray average
Glass panel is expert at coordinate.In this step, the two-dimensional image information of initial gray level image is converted into one-dimension information.Together
When, although this step obtains the row where panel, corresponding card slot is not necessarily located in picture center, it is impossible to passes through figure
The center of picture determines, therefore also needs to obtain card slot or so two endpoints further to obtain the card slot corresponding to panel row just
Center, therefore also need to carry out step S3.
S3, binarization segmentation is carried out to the gray level image, is extracted area-of-interest edge according to the glass panel and is sat
Mark；
S4, determines card slot centre coordinate, with the card slot according to the row coordinate and the area-of-interest edge coordinate
Absorption position of the centre coordinate as the glass panel, then control machinery hand reach the absorption position of the glass panel into
Row absorption.
Specifically, step S1 is specifically included：
S11, control machinery hand are moved to above rack card slot, and control the monocular CCD phases being fixed on the manipulator
Machine collection is placed with the image of the card slot of glass panel.In this step, the image of collection is as shown in Figure 2.
S12, carries out gradation conversion to the image of step S11 collections, gray level image is obtained after then being pre-processed；It is described
Pretreatment includes filtering, denoising etc..
Specifically, step S2 is specifically included：
S21, calculates the summation of the gray value of every one-row pixels of the gray level image；
The i-th row of gray level image jth row are represented with I (i, j), r represents the height of the gray level image, and c represents the ash
The width of image is spent, Row (i) represents the summation of the gray value of the i-th row of gray level image pixel, then:Wherein 0≤i≤r, 0≤j≤c；
S22, according to the summation of the gray value of every one-row pixels, calculates per a line pixel grey scale average；
I-th row grey scale pixel value average is represented with RowAve (i), then：
RowAve (i)=Row (i)/c；
S23, row pixel grey scale maximum is found according to every a line pixel grey scale average, true with the row pixel grey scale maximum
Recognize the glass panel to be expert at coordinate.
Specifically, above-mentioned steps S23 is specifically included：
S231, calculates the row pixel grey scale average of the gray level image；
The row pixel grey scale average of the gray level image is represented with RowAverage, then：
S232, calculates the gray scale difference value per one-row pixels；
The gray scale difference value of the i-th row pixel is represented with Delta (i), then the row pixel deviates the row pixel grey scale average
Size is：Delta (i)=RowAve (i)-RowAverage；
S233, travels through per a line pixel grey scale average, obtains maximum, and threshold value setting is carried out with the maximum；
The threshold value is represented with Delta, MaxRowAve represents the maximum, then：
Delta=(MaxRowAve-RowAverage) * 0.8.0.8 in this step is obtained according to experimental data, pin
To different card slots, it can accordingly make and change.
S234, judges whether the gray scale difference value meets the threshold value, and gray scale maximum is determined according to judging result, so that
Determine that the glass panel is expert at coordinate；
If Delta ＜ Delta (i), it is determined that the row where the i-th behavior gray scale maximum, i.e., described glass panel place
Row, and obtain the glass panel and be expert at coordinate.
Specifically, above-mentioned steps S3 is specifically included：
S31, carries out binarization segmentation processing to the gray level image, obtains binarization of gray value image；
S32, carries out BLOB analyses to the binarization of gray value image, obtains region of interest area image；
S33, carries out edge extracting to the region of interest area image, area-of-interest edge is obtained according to the edge of extraction
Coordinate；
The leftmost edge extreme coordinates of the i-th row of the region of interest area image are represented with ColGrayVa1 (i),
ColGrayVa2 (i) represents the right edge point coordinate of the i-th row of the region of interest area image, then：
The region of interest area image is begun stepping through from a left side to the i-th row, is jumped out when meeting ColGrayVa1 (i)=255
Circulation, records the point coordinates and is begun stepping through from the right side；When meeting ColGrayVa2 (i)=255, the point coordinates is recorded, it is then right
I+1 begins stepping through the region of interest area image from a left side.In this step, after the i-th row is traveled through, i+1 row is started,
Repeat above-mentioned traversal and obtain left and right edges extreme coordinates, until having traveled through region of interest area image.
Specifically, step S4 further comprises：
S41, according to the row coordinate and the area-of-interest edge coordinate, determines the card slot center of the row；
The card slot center of the i-th row is represented with RowCenter (i), the row center of the i-th row is represented with ColCenter (i), with
ColGrayVa1 (i) represents the leftmost edge extreme coordinates of the i-th row of the region of interest area image, ColGrayVa2 (i) tables
Show the right edge point coordinate of the i-th row of the region of interest area image, then：
RowCenter (i)=i；
ColCenter (i)=(ColGrayVa1 (i)+ColGrayVa2 (i))/2；
S42, the absorption position of the glass panel is determined according to camera fields of view centre coordinate and the card slot centre coordinate；
The row coordinate at camera fields of view center is represented with CameraRowCenter, CameraColCenter represents camera fields of view
The row coordinate at center, r represent the height of the gray level image, and c represents the width of the gray level image, then：
CameraRowCenter=r/2；CameraColCenter=c/2；
And if only if meet abs (CameraRowCenter (i)-RowCenter (i)) and
When abs (CameraColCenter (i)-ColCenter (i)) is minimum, the behavior glass to be adsorbed is determined
Panel is expert at, and is expert at coordinate with the row, the three-dimensional coordinate of combining camera, obtains the absorption position of the glass panel；
S43, the absorption position that control machinery hand reaches the glass panel are adsorbed.
In the present invention, extract two endpoint locations of card slot, and then obtain card slot center, coordinate where combining camera and
The visual field, obtains currently adsorbing glass panel centre coordinate, and driving manipulator reaches position absorption, completes positioning.
Present invention also offers a kind of sorting fast positioning based on monocular CCD small size glass panels as shown in Figure 6
System, including：
Acquisition process unit 1, the image of the card slot for being placed with glass panel using monocular CCD collections, and will collection
The image arrived carries out gradation conversion, and the image for then obtaining gradation conversion is pre-processed to obtain gray level image；
Computing unit 2, for calculating the row pixel grey scale average of the gray level image, then according to the row gray average
Determine that the glass panel is expert at coordinate；
Edge extracting unit 3, for carrying out binarization segmentation to the gray level image, extracts according to the glass panel and feels
Interest region edge coordinate；
Absorbing unit 4 is positioned, for determining card slot center according to the row coordinate and the area-of-interest edge coordinate
Coordinate, using the card slot centre coordinate as the absorption position of the glass panel, then control machinery hand reaches the glass
The absorption position of panel is adsorbed.
Further, acquisition process unit 1 is specifically used for：
First, control machinery hand is moved to above rack card slot, and controls the monocular CCD phases being fixed on the manipulator
Machine collection is placed with the image of the card slot of glass panel；
Finally, gradation conversion is carried out to the image of collection, gray level image is obtained after then being pre-processed；The pretreatment
Including filtering, denoising etc..
Further, computing unit 2 is specifically used for：
First, the summation of the gray value of every one-row pixels of the gray level image is calculated；
The i-th row of gray level image jth row are represented with I (i, j), r represents the height of the gray level image, and c represents the ash
The width of image is spent, Row (i) represents the summation of the gray value of the i-th row of gray level image pixel, then:Wherein 0≤i≤r, 0≤j≤c；
Secondly, according to the summation of the gray value of every one-row pixels, calculate per a line pixel grey scale average；
I-th row grey scale pixel value average is represented with RowAve (i), then：
RowAve (i)=Row (i)/c；
Finally, row pixel grey scale maximum is found according to every a line pixel grey scale average, with the row pixel grey scale maximum
Confirm that the glass panel is expert at coordinate.Wherein, also include in this implementation process into one,
Calculate the row pixel grey scale average of the gray level image；
The row pixel grey scale average of the gray level image is represented with RowAverage, then：
Calculate the gray scale difference value per one-row pixels；
The gray scale difference value of the i-th row pixel is represented with Delta (i), then the row pixel deviates the row pixel grey scale average
Size is：Delta (i)=RowAve (i)-RowAverage；
Traversal obtains maximum per a line pixel grey scale average, and threshold value setting is carried out with the maximum；
The threshold value is represented with Delta, MaxRowAve represents the maximum, then：
Delta=(MaxRowAve-RowAverage) * 0.8；
Judge whether the gray scale difference value meets the threshold value, gray scale maximum is determined according to judging result, so that it is determined that
Glass panel is expert at coordinate；
If Delta ＜ Delta (i), it is determined that the row where the i-th behavior gray scale maximum, i.e. row where glass panel,
And obtain glass panel and be expert at coordinate.
Further, edge extracting unit 3 is specifically used for：
First, binarization segmentation processing is carried out to the gray level image, obtains binarization of gray value image；
Then, BLOB analyses are carried out to the binarization of gray value image, obtains region of interest area image；
Finally, edge extracting is carried out to the region of interest area image, area-of-interest side is obtained according to the edge of extraction
Edge coordinate；
The leftmost edge extreme coordinates of the i-th row of the region of interest area image are represented with ColGrayVa1 (i),
ColGrayVa2 (i) represents the right edge point coordinate of the i-th row of the region of interest area image, then：
The region of interest area image is begun stepping through from a left side to the i-th row, is jumped out when meeting ColGrayVa1 (i)=255
Circulation, records the point coordinates and is begun stepping through from the right side；When meeting ColGrayVa2 (i)=255, the point coordinates is recorded, it is then right
I+1 begins stepping through the region of interest area image from a left side.
Further, positioning absorbing unit is specifically used for：
First, according to the row coordinate and the area-of-interest edge coordinate, determine the card slot center of the row, and obtain
Its coordinate；
The card slot center of the i-th row is represented with RowCenter (i), the row center of the i-th row is represented with ColCenter (i), with
ColGrayVa1 (i) represents the leftmost edge extreme coordinates of the i-th row of the region of interest area image, ColGrayVa2 (i) tables
Show the right edge point coordinate of the i-th row of the region of interest area image, then：
RowCenter (i)=i；
ColCenter (i)=(ColGrayVa1 (i)+ColGrayVa2 (i))/2；
Secondly, the adsorption potential of the glass panel is determined according to camera fields of view centre coordinate and the card slot centre coordinate
Put；
The row coordinate at camera fields of view center is represented with CameraRowCenter, CameraColCenter represents camera fields of view
The row coordinate at center, r represent the height of the gray level image, and c represents the width of the gray level image, then：
CameraRowCenter=r/2；CameraColCenter=c/2；
And if only if meet abs (CameraRowCenter (i)-RowCenter (i)) and
When abs (CameraColCenter (i)-ColCenter (i)) is minimum, the behavior glass to be adsorbed is determined
Panel is expert at, and is expert at coordinate with the row, the three-dimensional coordinate of combining camera, obtains the absorption position of the glass panel；
Finally, control machinery hand reaches the absorption position and is adsorbed.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
All any modification, equivalent and improvement made within refreshing and principle etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of sorting method for rapidly positioning of the small size glass panel based on monocular CCD, it is characterised in that the sorting is fast
Fast positioning method step includes：
Step A, the image of the card slot of glass panel is placed with using monocular CCD collections, and the image collected is carried out gray scale
Conversion, the image for then obtaining gradation conversion are pre-processed to obtain gray level image；
Step B, calculates the row pixel grey scale average of the gray level image, then determines the glass according to the row gray average
Panel is expert at coordinate；
Step C, binarization segmentation is carried out to the gray level image, is extracted area-of-interest edge according to the glass panel and is sat
Mark；
Step D, determines card slot centre coordinate, with the card slot according to the row coordinate and the area-of-interest edge coordinate
Absorption position of the heart coordinate as the glass panel, then control machinery hand reach the glass panel absorption position carry out
Absorption.
2. sorting method for rapidly positioning as claimed in claim 1, it is characterised in that step A is specifically included：
Step A1, control machinery hand are moved to above rack card slot, and control the monocular CCD camera being fixed on the manipulator
Collection is placed with the image of the card slot of glass panel；
Step A2, carries out gradation conversion to the image of step A1 collections, gray level image is obtained after then being pre-processed；It is described pre-
Processing includes filtering, denoising.
3. sorting method for rapidly positioning as claimed in claim 1, it is characterised in that step B is specifically included：
Step B1, calculates the summation of the gray value of every one-row pixels of the gray level image；
The i-th row of gray level image jth row are represented with I (i, j), r represents the height of the gray level image, and c represents the gray-scale map
The width of picture, Row (i) represent the summation of the gray value of the i-th row of gray level image pixel, then:
Wherein 0≤i≤r, 0≤j≤c；
Step B2, according to the summation of the gray value of every one-row pixels, calculates per a line pixel grey scale average；
I-th row grey scale pixel value average is represented with RowAve (i), then：
RowAve (i)=Row (i)/c；
Step B3, row pixel grey scale maximum is found according to every a line pixel grey scale average, true with the row pixel grey scale maximum
Recognize the glass panel to be expert at coordinate.
4. sorting method for rapidly positioning as claimed in claim 3, it is characterised in that step B3 is specifically included：
Step B31, calculates the row pixel grey scale average of the gray level image；
The row pixel grey scale average of the gray level image is represented with RowAverage, then：
<mrow>
<mi>R</mi>
<mi>o</mi>
<mi>w</mi>
<mi>A</mi>
<mi>v</mi>
<mi>e</mi>
<mi>r</mi>
<mi>a</mi>
<mi>g</mi>
<mi>e</mi>
<mo>=</mo>
<msubsup>
<mi>&Sigma;</mi>
<mn>0</mn>
<mi>r</mi>
</msubsup>
<mi>R</mi>
<mi>o</mi>
<mi>w</mi>
<mi>A</mi>
<mi>v</mi>
<mi>e</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mi>r</mi>
<mo>;</mo>
</mrow>
Step B32, calculates the gray scale difference value per one-row pixels；
The gray scale difference value of the i-th row pixel is represented with Delta (i), then the row pixel deviates the size of the row pixel grey scale average
For：Delta (i)=RowAve (i)-RowAverage；
Step B33, travels through per a line pixel grey scale average, obtains maximum, and threshold value setting is carried out with the maximum；
The threshold value is represented with Delta, MaxRowAve represents the maximum, then：
Delta=(MaxRowAve-RowAverage) * 0.8；
Step B34, judges whether the gray scale difference value meets the threshold value, and gray scale maximum is determined according to judging result, so that
Determine that the glass panel is expert at coordinate；
If Delta ＜ Delta (i), it is determined that the row where the i-th behavior gray scale maximum, i.e., the row where described glass panel,
And obtain the glass panel and be expert at coordinate.
5. sorting method for rapidly positioning as claimed in claim 1, it is characterised in that the step C is specifically included：
Step C1, carries out binarization segmentation processing to the gray level image, obtains binarization of gray value image；
Step C2, carries out BLOB analyses to the binarization of gray value image, obtains region of interest area image；
Step C3, carries out edge extracting to the region of interest area image, area-of-interest edge is obtained according to the edge of extraction
Coordinate；
The leftmost edge extreme coordinates of the i-th row of the region of interest area image, ColGrayVa2 are represented with ColGrayVa1 (i)
(i) the right edge point coordinate of the i-th row of the region of interest area image is represented, then：
The region of interest area image is begun stepping through from a left side to the i-th row, circulation is jumped out when meeting ColGrayVa1 (i)=255,
Record the point coordinates and begun stepping through from the right side；When meeting ColGrayVa2 (i)=255, the point coordinates is recorded, then to i+1
The region of interest area image is begun stepping through from a left side.
6. sorting method for rapidly positioning as claimed in claim 1, it is characterised in that step D is specifically included：
Step D1, according to the row coordinate and the area-of-interest edge coordinate, determines the card slot center of the row, and obtains it
Coordinate；
The card slot center of the i-th row is represented with RowCenter (i), the row center of the i-th row is represented with ColCenter (i), with
ColGrayVa1 (i) represents the leftmost edge extreme coordinates of the i-th row of the region of interest area image, ColGrayVa2 (i) tables
Show the right edge point coordinate of the i-th row of the region of interest area image, RowCenter (i)=i is represented in the i-th row card slot
The row coordinate of the heart, then：
RowCenter (i)=i；
ColCenter (i)=(ColGrayVa1 (i)+ColGrayVa2 (i))/2；
Step D2, the absorption position of the glass panel is determined according to camera fields of view centre coordinate and the card slot centre coordinate；
The row coordinate at camera fields of view center is represented with CameraRowCenter, CameraColCenter represents camera fields of view center
Row coordinate, r represents the height of the gray level image, and c represents the width of the gray level image, then：
CameraRowCenter=r/2；CameraColCenter=c/2；
And if only if meet abs (CameraRowCenter (i)-RowCenter (i)) and
When abs (CameraColCenter (i)-ColCenter (i)) is minimum, the behavior glass panel to be adsorbed is determined
Be expert at, be expert at coordinate with the row, the three-dimensional coordinate of combining camera, obtains the absorption position of the glass panel；
Step D3, control machinery hand reach the absorption position and are adsorbed.
7. a kind of sorting quick positioning system of the small size glass panel based on monocular CCD, it is characterised in that the sorting is fast
Fast alignment system includes：
Acquisition process unit, the image of the card slot for being placed with glass panel using monocular CCD collections, and the figure that will be collected
As carrying out gradation conversion, the image for then obtaining gradation conversion is pre-processed to obtain gray level image；
Computing unit, for calculating the row pixel grey scale average of the gray level image, then determines according to the row gray average
The glass panel is expert at coordinate；
Edge extracting unit, for carrying out binarization segmentation to the gray level image, is extracted interested according to the glass panel
Edges of regions coordinate；
Absorbing unit is positioned, for determining card slot centre coordinate according to the row coordinate and the area-of-interest edge coordinate,
Using the card slot centre coordinate as the absorption position of the glass panel, then control machinery hand reaches the glass panel
Absorption position is adsorbed.
8. sorting quick positioning system as claimed in claim 7, it is characterised in that the acquisition process unit is specifically used for：
First, control machinery hand is moved to above rack card slot, and controls the monocular CCD camera being fixed on the manipulator to adopt
Collection is placed with the image of the card slot of glass panel；
Finally, gradation conversion is carried out to the image of collection, gray level image is obtained after then being pre-processed；The pretreatment includes
Filtering, denoising.
9. sorting quick positioning system as claimed in claim 7, it is characterised in that the computing unit is specifically used for：
First, the summation of the gray value of every one-row pixels of the gray level image is calculated；
The i-th row of gray level image jth row are represented with I (i, j), r represents the height of the gray level image, and c represents the gray-scale map
The width of picture, Row (i) represent the summation of the gray value of the i-th row of gray level image pixel, then:
Wherein 0≤i≤r, 0≤j≤c；
Secondly, according to the summation of the gray value of every one-row pixels, calculate per a line pixel grey scale average；
I-th row grey scale pixel value average is represented with RowAve (i), then：
RowAve (i)=Row (i)/c；
Finally, row pixel grey scale maximum is found according to every a line pixel grey scale average, is confirmed with the row pixel grey scale maximum
The glass panel is expert at coordinate.
10. sorting quick positioning system as claimed in claim 7, it is characterised in that positioning absorbing unit is specifically used for：
First, according to the row coordinate and the area-of-interest edge coordinate, determine the card slot center of the row, and obtain its seat
Mark；
The card slot center of the i-th row is represented with RowCenter (i), the row center of the i-th row is represented with ColCenter (i), with
ColGrayVa1 (i) represents the leftmost edge extreme coordinates of the i-th row of the region of interest area image, ColGrayVa2 (i) tables
Show the right edge point coordinate of the i-th row of the region of interest area image, RowCenter (i)=i is represented in the i-th row card slot
The row coordinate of the heart, then：
RowCenter (i)=i；
ColCenter (i)=(ColGrayVa1 (i)+ColGrayVa2 (i))/2；
Secondly, the absorption position of the glass panel is determined according to camera fields of view centre coordinate and the card slot centre coordinate；
The row coordinate at camera fields of view center is represented with CameraRowCenter, CameraColCenter represents camera fields of view center
Row coordinate, r represents the height of the gray level image, and c represents the width of the gray level image, then：
CameraRowCenter=r/2；CameraColCenter=c/2；
And if only if meet abs (CameraRowCenter (i)-RowCenter (i)) and
When abs (CameraColCenter-ColCenter (i)) is minimum, the behavior glass panel to be adsorbed is determined
It is expert at, is expert at coordinate with the row, the three-dimensional coordinate of combining camera, obtains the absorption position of the glass panel；
Finally, control machinery hand reaches the absorption position and is adsorbed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510771640.7A CN105457908B (en) | 2015-11-12 | 2015-11-12 | The sorting method for rapidly positioning and system of small size glass panel based on monocular CCD |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510771640.7A CN105457908B (en) | 2015-11-12 | 2015-11-12 | The sorting method for rapidly positioning and system of small size glass panel based on monocular CCD |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105457908A CN105457908A (en) | 2016-04-06 |
| CN105457908B true CN105457908B (en) | 2018-04-13 |
Family
ID=55596316
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510771640.7A Expired - Fee Related CN105457908B (en) | 2015-11-12 | 2015-11-12 | The sorting method for rapidly positioning and system of small size glass panel based on monocular CCD |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105457908B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106493086B (en) * | 2016-10-21 | 2019-03-05 | 北京源著智能科技有限公司 | Sheet sorting method and system |
| CN106378663A (en) * | 2016-11-28 | 2017-02-08 | 重庆大学 | Machine tool auxiliary tool setting system based on machine vision |
| CN107798683A (en) * | 2017-11-10 | 2018-03-13 | 珠海格力智能装备有限公司 | Method and device for detecting edge of specific area of product and terminal |
| CN109035135B (en) * | 2018-07-13 | 2023-03-24 | 常州宏大智能装备产业发展研究院有限公司 | Machine vision-based online automatic fabric pattern finishing method |
| CN109143624B (en) | 2018-08-28 | 2020-06-16 | 武汉华星光电技术有限公司 | Panel adsorption device and automatic adsorption method adopting same |
| CN110517318B (en) * | 2019-08-28 | 2022-05-17 | 昆山国显光电有限公司 | Positioning method and device, and storage medium |
| CN111797695B (en) * | 2020-06-10 | 2023-09-29 | 盐城工业职业技术学院 | Automatic identification method and system for twist of folded yarn |
| CN114603715A (en) * | 2022-03-10 | 2022-06-10 | 郴州旗滨光伏光电玻璃有限公司 | Glass punching method, device and computer readable storage medium |
| CN115546316A (en) * | 2022-10-13 | 2022-12-30 | 中国大恒(集团)有限公司北京图像视觉技术分公司 | A method for automatic ROI setting of industrial cameras |
| CN117314875A (en) * | 2023-10-16 | 2023-12-29 | 东北大学秦皇岛分校 | A float glass defect detection method based on particle swarm optimization algorithm |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101334263A (en) * | 2008-07-22 | 2008-12-31 | 东南大学 | Method for locating the center of a circular target |
| US8224078B2 (en) * | 2000-11-06 | 2012-07-17 | Nant Holdings Ip, Llc | Image capture and identification system and process |
| EP2639745A1 (en) * | 2012-03-16 | 2013-09-18 | Thomson Licensing | Object identification in images or image sequences |
| CN104751147A (en) * | 2015-04-16 | 2015-07-01 | 成都汇智远景科技有限公司 | Image recognition method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4493679B2 (en) * | 2007-03-29 | 2010-06-30 | 富士フイルム株式会社 | Target region extraction method, apparatus, and program |
| GB2499799B (en) * | 2012-02-28 | 2019-06-26 | Snell Advanced Media Ltd | Identifying points of interest in an image |
-
2015
- 2015-11-12 CN CN201510771640.7A patent/CN105457908B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8224078B2 (en) * | 2000-11-06 | 2012-07-17 | Nant Holdings Ip, Llc | Image capture and identification system and process |
| CN101334263A (en) * | 2008-07-22 | 2008-12-31 | 东南大学 | Method for locating the center of a circular target |
| EP2639745A1 (en) * | 2012-03-16 | 2013-09-18 | Thomson Licensing | Object identification in images or image sequences |
| CN104751147A (en) * | 2015-04-16 | 2015-07-01 | 成都汇智远景科技有限公司 | Image recognition method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105457908A (en) | 2016-04-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105457908B (en) | The sorting method for rapidly positioning and system of small size glass panel based on monocular CCD | |
| CN103077521B (en) | A kind of area-of-interest exacting method for video monitoring | |
| CN108549864B (en) | Vehicle-mounted thermal imaging pedestrian detection-oriented region-of-interest filtering method and device | |
| CN108090435B (en) | Parking available area identification method, system and medium | |
| US20160180188A1 (en) | Method for detecting salient region of stereoscopic image | |
| CN103955949B (en) | Moving target detecting method based on Mean-shift algorithm | |
| CN104794479B (en) | This Chinese detection method of natural scene picture based on the transformation of local stroke width | |
| CN103996198A (en) | Method for detecting region of interest in complicated natural environment | |
| CN110910421B (en) | Weak and small moving object detection method based on block characterization and variable neighborhood clustering | |
| CN103839038A (en) | People counting method and device | |
| US20150186723A1 (en) | Method and apparatus for detecting generalized passerby | |
| CN110827312B (en) | Learning method based on cooperative visual attention neural network | |
| CN104299009A (en) | Plate number character recognition method based on multi-feature fusion | |
| CN105787930A (en) | Sharpness-based significance detection method and system for virtual images | |
| CN105404868A (en) | Interaction platform based method for rapidly detecting text in complex background | |
| CN105740751A (en) | Object detection and identification method and system | |
| CN107097256B (en) | A Model-Free Target Tracking Method Based on Vision Incomplete Robot in Polar Coordinates | |
| CN106503683A (en) | A kind of video well-marked target detection method based on dynamic focal point | |
| CN106156691A (en) | The processing method of complex background image and device thereof | |
| CN103136517B (en) | A kind of real-time joining method of rolling fingerprint image sequence selected based on key column | |
| Yanagisawa et al. | Face detection for comic images with deformable part model | |
| CN107506777A (en) | A real-time multi-license plate recognition method and device based on wavelet variation and support vector machine | |
| Zhang et al. | Pedestrian detection with EDGE features of color image and HOG on depth images | |
| CN106778504A (en) | A kind of pedestrian detection method | |
| CN103390162B (en) | A kind of station caption detection method |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180413 |