CN110263597A - A kind of fast and accurately bearing calibration of QR code and system - Google Patents
A kind of fast and accurately bearing calibration of QR code and system Download PDFInfo
- Publication number
- CN110263597A CN110263597A CN201910564802.8A CN201910564802A CN110263597A CN 110263597 A CN110263597 A CN 110263597A CN 201910564802 A CN201910564802 A CN 201910564802A CN 110263597 A CN110263597 A CN 110263597A
- Authority
- CN
- China
- Prior art keywords
- image
- vertex
- coordinate
- pixel coordinate
- line
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
- G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
- G06K7/14—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation using light without selection of wavelength, e.g. sensing reflected white light
- G06K7/1404—Methods for optical code recognition
- G06K7/146—Methods for optical code recognition the method including quality enhancement steps
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
- G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
- G06K7/14—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation using light without selection of wavelength, e.g. sensing reflected white light
- G06K7/1404—Methods for optical code recognition
- G06K7/146—Methods for optical code recognition the method including quality enhancement steps
- G06K7/1473—Methods for optical code recognition the method including quality enhancement steps error correction
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Electromagnetism (AREA)
- General Health & Medical Sciences (AREA)
- Toxicology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of fast and accurately bearing calibration of QR code and system, comprising: is scanned to the binary image A (x, y) of QR code, obtains the center pixel coordinate of three position sensing figures in image A (x, y)；Seek the first kind pixel coordinate on the 4th vertex in image A (x, y)；Seek the second class pixel coordinate on the 4th vertex in image A (x, y)；The first kind pixel coordinate on the 4th vertex in image A (x, y) is weighted and averaged with the second class pixel coordinate, obtains the pixel coordinate on the 4th vertex in image A (x, y)；Anti- perspective transform is carried out to four vertex in image A (x, y), obtains the standard pixel coordinate on four vertex；Perspective transformation matrix is sought according to the original pixel coordinate Yu standard pixel coordinate on four vertex in image A (x, y), and the standard picture after being corrected.Its calculating process is simple and more accurate, can effectively make up the deficiency of existing QR code bearing calibration, have very high practical value.The present invention is applied to image procossing, technical field of computer vision.
Description
Technical field
The present invention relates to image procossings, technical field of computer vision, more particularly to a kind of fast and accurately QR code to correct
Method and system.
Background technique
With the arriving of information age, the progress of computer science and the intelligence of mobile phone, barcode technology meet the tendency of and
It is raw.Barcode technology is a kind of important method that information automation obtains.Bar code stores the difference of dimension according to its information, usually has
Point of bar code and two-dimensional bar code.For bar code, not only memory capacity is big for two-dimensional bar code, secrecy ability it is strong and
And low in cost, good reliability.
China starts from the last century 90's for the exploration of 2D bar code technology, starts late.But due to section, China
The rapid development of technology, the rapid rising of logistic industry and popularizing for smart phone, 2D bar code technology is answered China
With further extensively.Currently, being QR code in the two-dimensional bar code that China occupies staple market.QR code is one of matrix two-dimensional barcode,
QR is " Quick Response ", the meaning with fast reaction.QR code is first and directly encodes to non-English character
Bar code, therefore the bar code is mainly used in China, Japan, many Asian countries such as South Korea.
The main task of QR code identification is to carry out a series of calculation process to collected QR code image, to obtain it
Included in information.Entire identifying system mainly includes 3 parts: Image Pretreatment Algorithm, location algorithm and correcting algorithm
These three aspects.Wherein correcting algorithm is essential key component.This is because QR code during Image Acquisition by
The error caused by the error and environmental factor and human factor of acquisition equipment itself, so that collected QR code image is past
Toward there are the distortions of various linearity and non-linearities.These distortions can generate tremendous influence to the precision of image.
Correcting algorithm mostly uses Hough transform to obtain the marginal information and tilt angle of image at present, and effect is general, quasi-
Exactness is not high, directly affects subsequent decoding process.
Summary of the invention
Aiming at the shortcomings in the prior art, the object of the present invention is to provide a kind of fast and accurately QR code bearing calibration and it is
System.
Itself the technical solution adopted is that:
A kind of fast and accurately QR code bearing calibration, the following steps are included:
Step 101, cluster is scanned to the binary image A (x, y) of QR code, obtains three vertex in image A (x, y)
Pixel coordinate, i.e., the center pixel coordinate of three position sensing figures in QR code image；
Step 102, according to the pixel coordinate and parallelogram law on three vertex in image A (x, y), image is sought
The first kind pixel coordinate on the 4th vertex in A (x, y)；
Step 103, the peripheral straight line of image A (x, y), root are sought according to the pixel coordinate on three vertex in image A (x, y)
The 4th vertex in image A (x, y) is sought according to the pixel coordinate on three vertex in image A (x, y) and the slope of peripheral straight line
The second class pixel coordinate；
Step 104, the first kind pixel coordinate to the 4th vertex in image A (x, y) and the second class pixel coordinate carry out
Weighted average obtains the pixel coordinate on the 4th vertex in image A (x, y)；
Step 105, anti-perspective transform is carried out to four vertex in image A (x, y), obtains the standard pixel on four vertex
Coordinate；
Step 106, perspective is sought according to the original pixel coordinate Yu standard pixel coordinate on four vertex in image A (x, y) to become
Matrix is changed, and the standard picture after being corrected according to perspective transformation matrix.
It is further preferred that the binary image A (x, y) to QR code is scanned cluster in step 101, obtain
The pixel coordinate on three vertex in image A (x, y), specifically:
Step 201, image A (x, y) is scanned line by line, will be connected with the identical pixel of adjacent and color in a line
Come, forms the line segment of several black and white in each row；
Step 202, filtering out continuous and length ratio in every a line is five line segments of 1:1:3:1:1 as a row
Line segment group, the starting pixels coordinate for filtering out the line segment that each line Duan Zuzhong length ratio is 3 clusters coordinate as row, right
All row cluster coordinates carry out capable cluster by condition of mutual distance, and taking includes row cluster coordinate quantity at first three
Three classifications are as row scanning result；
Step 203, image A (x, y) is scanned by column, the identical pixel of adjacent in same row and color is connected
Come, forms the line segment of several black and white in each column；
Step 204, five line segments that continuous and length ratio in each column is 1:1:3:1:1 are filtered out to arrange as one
Line segment group, the starting pixels coordinate for filtering out the line segment that each alignment Duan Zuzhong length ratio is 3 clusters coordinate as column, right
All column cluster coordinates carry out column cluster by condition of mutual distance, and taking includes column cluster coordinate quantity at first three
Three classifications are as column scan result；
Step 205, the intersection for seeking row scanning result Yu column scan result obtains the bianry image of three clusters, calculates
Out three cluster bianry image center pixel coordinate, that is, image A (x, y) in three vertex pixel coordinate.
It is further preferred that row cluster coordinate carries out row less than 4 with mutual distance for condition and gathers in step 202
Class；In step 204, column cluster coordinate carries out column cluster less than 4 with mutual distance for condition.
It is further preferred that in step 102, the first kind pixel coordinate on the 4th vertex in described image A (x, y) are as follows:
Dot4_1=(x04_1,y04_1)=(x02+x03-x01,y02+y03-y01)
In formula, (x04_1,y04_1) be the 4th vertex in image A (x, y) first kind pixel coordinate；(x01,y01) it is figure
As the pixel coordinate of first vertex dot1 in A (x, y)；(x02,y02) be second vertex dot2 in image A (x, y) pixel
Coordinate；(x03,y03) be third vertex dot3 in image A (x, y) pixel coordinate, wherein point dot2 and point dot3 are located at QR
On a diagonal line in code image.
It is further preferred that the step 103 specifically includes:
Step 301, the column where point dot1 are scanned from bottom to up, the starting pixels for recording second of black line segment are sat
Mark (x01 1,y01), the column where point dot2 are scanned from bottom to up, record the starting pixels coordinate of second of black line segment
(x02 2,y02), and seek the slope angle, θ of line between point dot1 and point dot212；
Step 302, with θ12Centered on, 0.1 degree of step-length is fluctuated, whole to be no more than 1 degree, acquisition includes 21 angle values
Slope angle set；
Step 303, with point (x01 1,y01) be on straight line a bit, respectively using slope angle set 21 angle values as slope
Angle obtains 21 straight lines；
Step 304, with point (x02 2,y02) be on straight line a bit, respectively using slope angle set 21 angle values as slope
Angle obtains 21 straight lines, in conjunction with the straight line obtained in step 303 totally 42 straight lines；
Step 305, straight line most with image A (x, y) intersection in 42 straight lines is filtered out as point dot1 and point dot2
Between peripheral straight line, and record its slope n θ12；
Step 306, the periphery between point dot1 and point dot3 is obtained directly using the same method of step 301- step 305
Line, and record its slope n θ13；
Step 307, remembered according to the slope of the pixel coordinate of three vertex dot1, dot2 and dot3 and two peripheral straight lines
For n θ12With n θ13, it can be in the hope of the last one apex coordinate of a quadrangle, i.e. the of the 4th vertex in image A (x, y)
Two class pixel coordinate dot4_2 (x04_2,y04_2)。
It is further preferred that in step 104, the pixel coordinate on the 4th vertex in image A (x, y) are as follows:
Ndot4=t × dot4_1+ (1-t) × dot4_2
In formula, t is weighted factor.
It is further preferred that in step 105, the standard pixel coordinate on four vertex be dot1 '=(4,4), dot2 '=
(4, LTH-3), dot3 '=(LTH-3,4), dot4 '=(LTH-3, LTH-3), wherein LTH is the side length of standard picture.
It is further preferred that the finding process of the side length LTH of the standard picture are as follows:
Step 401, the side length lth of image A (x, y) is sought:
In formula,Indicate the overall length of all line segments in same row or column, v (i) indicates line section group in the row or column
Or the length of first line segment of alignment Duan Zuzhong；
Step 402, immediate integer is found as LTH in 4 incremental arrays to lth.
A kind of fast and accurately QR code correction system, including memory and processor, the memory are stored with computer
The step of program, the processor realizes the above method when executing the computer program.
Advantageous effects of the invention:
The present invention passes through the method based on parallelogram law and the side based on QR code image periphery straight line information respectively
Method seeks the pixel coordinate on the 4th vertex in QR code image, and two kinds of pixel coordinates on the 4th vertex are then weighted place
Reason obtains final coordinate, and obtains perspective transformation matrix by the pixel coordinate on four vertex, finally according to perspective transform square
Battle array can be corrected QR code original image, and calculating process is simple and more accurate, can effectively make up existing QR code school
The deficiency of correction method has very high practical value.
Detailed description of the invention
Fig. 1 is QR code sign structural schematic diagram in the present embodiment；
Fig. 2 is the structural schematic diagram of QR code position detection figure in the present embodiment；
Fig. 3 is the flow diagram of fast and accurately QR code bearing calibration；
Fig. 4 is the finding process flow diagram of the pixel coordinate on three vertex；
Fig. 5 is the relation schematic diagram of three kinds of cluster centres in the present embodiment；
Fig. 6 is the finding process flow diagram of the second class pixel coordinate on the 4th vertex.
Specific embodiment
In order to which the purposes, technical schemes and advantages of the disclosure are more clearly understood, under in conjunction with specific embodiments, and according to
Attached drawing, the present invention is described in more detail.It should be noted that in attached drawing or specification description, the content that does not describe with
And part English is abbreviated as content known to those of ordinary skill in technical field.The some spies given in the present embodiment
Parameter is determined only as demonstration, and the value can change accordingly to suitably be worth in various embodiments.
The binary image A (x, y) comprising QR code Jing Guo Primary Location is read in, in the collection process of QR code image,
Generally there are a degree of burr and a small amount of noise in image vertex after positioning.Therefore, if directly passing through scanning
Entire image, and the vertex for the outermost that will acquire can be generated directly as the outermost point in the identification process in QR code region
Great error.
As shown in Figs. 1-2, QR code sign is a two-dimensional bar code array being made of a series of square modules, wherein
Position sensing figure is successively with 1:1:3:1:1 depth checker, and this property is not too serious and any in anamorphose
It is stabilized in the case where rotation.
The present embodiment finds 3 according to the characteristics of QR code image by carrying out scanning line by line to entire image
Then vertex is obtained the coordinate on the 4th vertex by two different methods, is weighted and averaged to two times result, meter is improved
The precision for calculating result calculates transformation matrix finally by the coordinate on this 4 vertex, and is completed according to transformation matrix to QR code image
Correction, with reference to Fig. 3, specifically includes the following steps:
Step 101, cluster is scanned to the binary image A (x, y) of QR code, obtains three vertex in image A (x, y)
Pixel coordinate, i.e., the center pixel coordinate of three position sensing figures in QR code image；
Step 102, according to the pixel coordinate and parallelogram law on three vertex in image A (x, y), image is sought
The first kind pixel coordinate on the 4th vertex in A (x, y)；
Step 103, the peripheral straight line of image A (x, y), root are sought according to the pixel coordinate on three vertex in image A (x, y)
The 4th vertex in image A (x, y) is sought according to the pixel coordinate on three vertex in image A (x, y) and the slope of peripheral straight line
The second class pixel coordinate, wherein the image A (x, y) that the peripheral straight line of image A (x, y) refers to is after removing surrounding margins region
Contour line；
Step 104, the first kind pixel coordinate to the 4th vertex in image A (x, y) and the second class pixel coordinate carry out
Weighted average obtains the pixel coordinate on the 4th vertex in image A (x, y)；
Step 105, anti-perspective transform is carried out to four vertex in image A (x, y), obtains the standard pixel on four vertex
Coordinate；
Step 106, perspective is sought according to the original pixel coordinate Yu standard pixel coordinate on four vertex in image A (x, y) to become
Matrix is changed, and the standard picture after being corrected according to perspective transformation matrix.
With reference to Fig. 4, in a step 101, cluster is scanned to the binary image A (x, y) of QR code, obtain image A (x,
Y) pixel coordinate on three vertex in, specifically:
Step 201, image A (x, y) is scanned line by line, will be connected with the identical pixel of adjacent and color in a line
Come, since the pixel of binary image A (x, y) is non-black i.e. white, forms the line of several black and white in each row
Section；
Step 202, filtering out continuous and length ratio in every a line is five line segments of 1:1:3:1:1 as a row
Line segment group, the starting pixels coordinate for filtering out the line segment that each line Duan Zuzhong length ratio is 3 clusters coordinate as row, right
All row cluster coordinates carry out capable cluster by condition of mutual distance, and taking includes row cluster coordinate quantity at first three
Three classifications are as row scanning result；
By taking the first row as an example, it is assumed that the first row is divided into m line segment, and the number of pixels of each line segment is indicated with one-dimensional vector
V=(v1,v2,...,vm), if it find that the ratio between the number of pixels of some line segment and 4 line segments thereafter meets 1:1:3:1:1
When, namely there are some serial number i, so that
V (i): v (i+1): v (i+2): v (i+3): v (i+4)=1:1:3:1:1
The starting pixels coordinate of the line segment of serial number i+2 is filtered out as row cluster coordinate, the row in every a line is gathered
Class coordinate is summarized and is clustered, and is taken comprising row cluster coordinate quantity in first three three classifications as row scanning result, wherein
Row cluster coordinate carries out capable cluster less than 4 with mutual distance for condition, and therefore, three obtained row scanning result exists
The cluster of the pixel coordinate of three obtained position sensing figure in the case where progressive scan.
Step 203, image A (x, y) is scanned by column, the identical pixel of adjacent in same row and color is connected
Come, forms the line segment of several black and white in each column；
Step 204, five line segments that continuous and length ratio in each column is 1:1:3:1:1 are filtered out to arrange as one
Line segment group, the starting pixels coordinate for filtering out the line segment that each alignment Duan Zuzhong length ratio is 3 clusters coordinate as column, right
All column cluster coordinates carry out column cluster less than 4 with mutual distance for condition, take and exist comprising column cluster coordinate quantity
Three classifications of first three are as column scan result；Principle identical with step 201-202, the result of three obtained column scan here
I.e. in the case where scanning by column the pixel coordinate of three obtained position sensing figure cluster.
Step 205, the intersection for seeking row scanning result Yu column scan result obtains the bianry image of three clusters, calculates
Out in center pixel coordinate, that is, image A (x, y) of the bianry image of three clusters three vertex pixel coordinate, i.e. QR code figure
In three position sensing figures center pixel coordinate；With reference to Fig. 5, in the present embodiment, three are indicated with dot1, dot2 and dot3
The central pixel point of a position sensing figure, wherein point dot2 and dot3 are located on a diagonal line in QR code image, specifically
Finding process are as follows:
In formula, N1It is the total number of all pixels point in first cluster intersection；x′1With y '1It is that first cluster is handed over respectively
The transverse and longitudinal coordinate of each pixel is concentrated, remembers dot1=(x01,y01) be first cluster intersection center pixel, i.e., first
A vertex；Same method can cluster the center pixel dot2=(x of intersection in the hope of second02,y02), i.e. second vertex；
With the center pixel dot3=(x of third cluster intersection03,y03), i.e. third vertex.
In a step 102, in image A (x, y) the 4th vertex first kind pixel coordinate are as follows:
Dot4_1=(x04_1,y04_1)=(x02+x03-x01,y02+y03-y01)
In formula, (x04_1,y04_1) be the 4th vertex in image A (x, y) first kind pixel coordinate；(x01,y01) it is figure
As the pixel coordinate of first vertex dot1 in A (x, y)；(x02,y02) be second vertex dot2 in image A (x, y) pixel
Coordinate；(x03,y03) be third vertex dot3 in image A (x, y) pixel coordinate, wherein point dot2 and point dot3 are located at QR
On a diagonal line in code image.
With reference to Fig. 6, in step 103, according to the peripheral information of QR bar code, peripheral slope is calculated, is then pushed up according to 3
The coordinate information of dot1, dot2 and dot3, peripheral slope information, calculate the second class pixel coordinate on the 4th vertex, process
Include:
Step 301, two o'clock dot1 and dot2 group is in alignment, the column where point dot1 are swept from bottom to up
It retouches, can successively pass through black line segment, white line section, black line segment, record the starting pixels coordinate (x of second of black line segment01 1,y01)；Similarly
Column where point dot2 are scanned from bottom to up, also can successively pass through black line segment, white line section, black line segment, record second
Starting pixels coordinate (the x of secondary black line segment02 2,y02), while seeking the slope angle, θ of line between point dot1 and point dot212
Step 302, with θ12Centered on, 0.1 degree of step-length is fluctuated, whole to be no more than 1 degree, acquisition includes 21 angle values
Slope angle set, specially { θ12-1,θ12-0.9,θ12-0.8,...,θ12-0.1,θ12,θ12+0.1,...,θ12+0.8,θ12+
0.9,θ12+1}；
Step 303, with point (x01 1,y01) be on straight line a bit, respectively using slope angle set 21 angle values as slope
Angle obtains 21 straight lines；
Step 304, with point (x02 2,y02) be on straight line a bit, respectively using slope angle set 21 angle values as slope
Angle obtains 21 straight lines, in conjunction with the straight line obtained in step 303 totally 42 straight lines；
Step 305, straight line most with image A (x, y) intersection in 42 straight lines is filtered out as point dot1 and point dot2
Between peripheral straight line, and record its slope n θ12；
Step 306, the periphery between point dot1 and point dot3 is obtained directly using the same method of step 301- step 305
Line, and record its slope n θ13；
Step 307, remembered according to the slope of the pixel coordinate of three vertex dot1, dot2 and dot3 and two peripheral straight lines
For n θ12With n θ13, it can be in the hope of the last one apex coordinate of a parallelogram, i.e. the 4th vertex in image A (x, y)
The second class pixel coordinate dot4_2 (x04_2,y04_2)。
At step 104, in image A (x, y) the 4th vertex pixel coordinate are as follows:
Ndot4=t × dot4_1+ (1-t) × dot4_2.
In step 105, under the premise of known four vertex dot1, dot2, dot3 and dot4, available four
The standard pixel coordinate on vertex be dot1 '=(4,4), dot2 '=(4, LTH-3), dot3 '=(LTH-3,4), dot4 '=
(LTH-3, LTH-3), wherein LTH is the side length of standard picture, and standard picture is the square of LTH × LTH, wherein standard drawing
The finding process of the side length LTH of picture are as follows:
Step 401, the side length lth of image A (x, y) is sought:
In formula,Indicate the overall length of all line segments in same row or column, v (i) indicates line section group in the row or column
Or the length of first line segment of alignment Duan Zuzhong；
Step 402, immediate integer is found as LTH in 4 incremental arrays to lth.Wherein, it is incremented by with 4
Array is { 1,5,9,13,17,21,25,29 ..., 177 }, such as when lth is 24.1, then LTH is immediate integer
It is exactly 25.
In step 106, the relationship of the standard picture after original image, perspective transformation matrix and correction are as follows:
In formula, the coordinate of original image is (x, y), and the coordinate of standard picture is (p, q), and H is perspective transformation matrix, and ω is
Scale factor；
Wherein, perspective transformation matrix can be calculated according to the original pixel coordinate on four vertex and standard pixel coordinate.
Contain the explanation of the preferred embodiment of the present invention above, this be for the technical characteristic that the present invention will be described in detail, and
Be not intended to for summary of the invention being limited in concrete form described in embodiment, according to the present invention content purport carry out other
Modifications and variations are also protected by this patent.The purport of the content of present invention is to be defined by the claims, rather than by embodiment
Specific descriptions are defined.
Claims (9)
1. a kind of fast and accurately QR code bearing calibration, which comprises the following steps:
Step 101, cluster is scanned to the binary image A (x, y) of QR code, obtains the picture on three vertex in image A (x, y)
Plain coordinate, i.e., the center pixel coordinate of three position sensing figures in QR code image；
Step 102, according to the pixel coordinate and parallelogram law on three vertex in image A (x, y), seek image A (x,
Y) first kind pixel coordinate on the 4th vertex in；
Step 103, the peripheral straight line that image A (x, y) is sought according to the pixel coordinate on three vertex in image A (x, y), according to figure
As the pixel coordinate on three vertex in A (x, y) and the slope of peripheral straight line seek in image A (x, y) the of the 4th vertex
Two class pixel coordinates；
Step 104, the first kind pixel coordinate on the 4th vertex in image A (x, y) is weighted with the second class pixel coordinate
It is average, obtain the pixel coordinate on the 4th vertex in image A (x, y)；
Step 105, anti-perspective transform is carried out to four vertex in image A (x, y), the standard pixel for obtaining four vertex is sat
Mark；
Step 106, perspective transform square is sought according to the original pixel coordinate Yu standard pixel coordinate on four vertex in image A (x, y)
Battle array, and the standard picture after being corrected according to perspective transformation matrix.
2. fast and accurately QR code bearing calibration according to claim 1, which is characterized in that described to QR code in step 101
Binary image A (x, y) be scanned cluster, obtain the pixel coordinate on three vertex in image A (x, y), specifically:
Step 201, image A (x, y) is scanned line by line, will be connected with the identical pixel of adjacent and color in a line,
The line segment of several black and white is formed in each row；
Step 202, filtering out continuous and length ratio in every a line is five line segments of 1:1:3:1:1 as a line section
Group filters out the starting pixels coordinate for the line segment that each line Duan Zuzhong length ratio is 3 as row cluster coordinate, to all
Row cluster coordinate by condition of mutual distance carry out capable cluster, take and cluster coordinate quantity in first three three comprising row
Classification is as row scanning result；
Step 203, image A (x, y) is scanned by column, the identical pixel of adjacent in same row and color is connected,
The line segment of several black and white is formed in each column；
Step 204, filtering out continuous and length ratio in each column is five line segments of 1:1:3:1:1 as an alignment section
Group filters out the starting pixels coordinate for the line segment that each alignment Duan Zuzhong length ratio is 3 as column cluster coordinate, to all
Column cluster coordinate by condition of mutual distance carry out column cluster, take and cluster coordinate quantity in first three three comprising column
Classification is as column scan result；
Step 205, the intersection for seeking row scanning result Yu column scan result obtains the bianry image of three clusters, calculates three
The pixel coordinate on three vertex in center pixel coordinate, that is, image A (x, y) of the bianry image of a cluster.
3. fast and accurately QR code bearing calibration according to claim 2, which is characterized in that in step 202, row cluster coordinate
It is clustered with mutual distance less than 4 for condition；In step 204, column cluster coordinate is with mutual distance less than 4
Column cluster is carried out for condition.
4. fast and accurately QR code bearing calibration according to claim 1, which is characterized in that in step 102, described image A
The first kind pixel coordinate on the 4th vertex in (x, y) are as follows:
Dot4_1=(x04_1,y04_1)=(x02+x03-x01,y02+y03-y01)
In formula, (x04_1,y04_1) be the 4th vertex in image A (x, y) first kind pixel coordinate；(x01,y01) it is image A
The pixel coordinate of first vertex dot1 in (x, y)；(x02,y02) it is that the pixel of second vertex dot2 in image A (x, y) is sat
Mark；(x03,y03) be third vertex dot3 in image A (x, y) pixel coordinate, wherein point dot2 and point dot3 are located at QR code
On a diagonal line in image.
5. fast and accurately QR code bearing calibration according to claim 4, which is characterized in that the step 103 specifically includes:
Step 301, the column where point dot1 are scanned from bottom to up, record the starting pixels coordinate of second of black line segment
(x01 1,y01), the column where point dot2 are scanned from bottom to up, record the starting pixels coordinate (x of second of black line segment02 2,
y02), and seek the slope angle, θ of line between point dot1 and point dot212；
Step 302, with θ12Centered on, 0.1 degree of step-length is fluctuated, whole to be no more than 1 degree, acquisition includes the slope of 21 angle values
Angle set；
Step 303, with point (x01 1,y01) be on straight line a bit, respectively using slope angle gather 21 angle values as slope angle
Degree, obtains 21 straight lines；
Step 304, with point (x02 2,y02) be on straight line a bit, respectively using slope angle gather 21 angle values as slope angle
Degree, obtains 21 straight lines, in conjunction with the straight line obtained in step 303 totally 42 straight lines；
Step 305, straight line most with image A (x, y) intersection in 42 straight lines is filtered out as between point dot1 and point dot2
Peripheral straight line, and record its slope n θ12；
Step 306, the peripheral straight line between point dot1 and point dot3 is obtained using the same method of step 301- step 305, and
Record its slope n θ13；
Step 307, n θ is denoted as according to the slope of the pixel coordinate of three vertex dot1, dot2 and dot3 and two peripheral straight lines12
With n θ13, it can be in the hope of the last one apex coordinate of a quadrangle, i.e. the second class picture on the 4th vertex in image A (x, y)
Plain coordinate dot4_2 (x04_2,y04_2)。
6. fast and accurately QR code bearing calibration according to claim 5, which is characterized in that in step 104, image A (x, y)
In the 4th vertex pixel coordinate are as follows:
Ndot4=t × dot4_1+ (1-t) × dot4_2
In formula, t is weighted factor.
7. fast and accurately QR code bearing calibration according to claim 2, which is characterized in that in step 105, four vertex
Standard pixel coordinate be dot1 '=(4,4), dot2 '=(4, LTH-3), dot3 '=(LTH-3,4), dot4 '=(LTH-3,
LTH-3), wherein LTH is the side length of standard picture.
8. fast and accurately QR code bearing calibration according to claim 7, which is characterized in that the side length of the standard picture
The finding process of LTH are as follows:
Step 401, the side length lth of image A (x, y) is sought:
In formula,Indicate the overall length of all line segments in same row or column, v (i) indicates line section group or column in the row or column
The length of first line segment in line segment group；
Step 402, immediate integer is found as LTH in 4 incremental arrays to lth.
9. a kind of fast and accurately QR code corrects system, including memory and processor, the memory are stored with computer journey
Sequence, which is characterized in that the processor realizes any one of claims 1 to 8 the method when executing the computer program
The step of.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910564802.8A CN110263597B (en) | 2019-06-27 | 2019-06-27 | Quick and accurate QR (quick response) code correction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910564802.8A CN110263597B (en) | 2019-06-27 | 2019-06-27 | Quick and accurate QR (quick response) code correction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110263597A true CN110263597A (en) | 2019-09-20 |
| CN110263597B CN110263597B (en) | 2022-08-19 |
Family
ID=67922064
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910564802.8A Active CN110263597B (en) | 2019-06-27 | 2019-06-27 | Quick and accurate QR (quick response) code correction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110263597B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113378595A (en) * | 2020-03-10 | 2021-09-10 | 顺丰科技有限公司 | Two-dimensional code positioning method, device, equipment and storage medium |
| CN114140330A (en) * | 2021-11-26 | 2022-03-04 | 武汉理工大学 | Bar code correction method and device based on perspective coordinate transformation and electronic equipment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070071320A1 (en) * | 2005-09-20 | 2007-03-29 | Fuji Xerox Co., Ltd. | Detection method of two-dimensional code, detection device for the same, and storage medium storing detection program for the same |
| CN101398895A (en) * | 2007-09-26 | 2009-04-01 | 杨高波 | Image preprocess method based on data matrix two-dimension bar code identification |
-
2019
- 2019-06-27 CN CN201910564802.8A patent/CN110263597B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070071320A1 (en) * | 2005-09-20 | 2007-03-29 | Fuji Xerox Co., Ltd. | Detection method of two-dimensional code, detection device for the same, and storage medium storing detection program for the same |
| CN101398895A (en) * | 2007-09-26 | 2009-04-01 | 杨高波 | Image preprocess method based on data matrix two-dimension bar code identification |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113378595A (en) * | 2020-03-10 | 2021-09-10 | 顺丰科技有限公司 | Two-dimensional code positioning method, device, equipment and storage medium |
| CN113378595B (en) * | 2020-03-10 | 2023-09-22 | 顺丰科技有限公司 | Two-dimensional code positioning method, device, equipment and storage medium |
| CN114140330A (en) * | 2021-11-26 | 2022-03-04 | 武汉理工大学 | Bar code correction method and device based on perspective coordinate transformation and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110263597B (en) | 2022-08-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107633192B (en) | Bar code segmentation and reading method based on machine vision under complex background | |
| US10679175B2 (en) | Two-dimensional code, system for creation of two-dimensional code, and analysis program | |
| CN100390807C (en) | Trilateral poly-dimensional bar code easy for omnibearing recognition and reading method thereof | |
| CN107609451A (en) | A kind of high-precision vision localization method and system based on Quick Response Code | |
| CN105989317B (en) | Two-dimensional code identification method and device | |
| US20150339508A1 (en) | Two-dimensional code | |
| CN104346597B (en) | A kind of QR codes detection and correction extracting method and IP kernel | |
| US9898637B2 (en) | Two-dimensional code | |
| CN105095822B (en) | A kind of Chinese letter co pattern image detection method and system | |
| KR20150044848A (en) | Method for decoding matrix-type two-dimensional code | |
| CN112801094B (en) | Pointer instrument image inclination correction method | |
| CN101794404A (en) | Method for coding and reading invisible two-dimensional codes | |
| CN103593826A (en) | Image ring artifact correcting method | |
| CN113506276B (en) | Marker and method for measuring structural displacement | |
| CN101882220B (en) | Bar code image correction method based on dynamic template and method for acquiring correction point | |
| CN105095937B (en) | A kind of visual identity method of the circular array graphic code based on straight line cluster | |
| CN102763121B (en) | Method for decoding a linear bar code | |
| CN110263597A (en) | A kind of fast and accurately bearing calibration of QR code and system | |
| CN101833644A (en) | Correction graph searching method based on dynamic template | |
| CN114399677A (en) | Pointer instrument identification method based on text region reading | |
| CN112184825B (en) | Calibration plate and calibration method | |
| CN103413319A (en) | Industrial camera parameter on-site calibration method | |
| CN111583191B (en) | Light field EPI Fourier transform-based refraction characteristic detection method | |
| CN101882213B (en) | Method for sampling barcode images | |
| CN101908144B (en) | Bar code image correction method and correction point acquisition method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | 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 |