CN104657737B - The method and apparatus being corrected based on cluster to QR image in 2 D code - Google Patents

The method and apparatus being corrected based on cluster to QR image in 2 D code Download PDF

Info

Publication number
CN104657737B
CN104657737B CN201310590970.7A CN201310590970A CN104657737B CN 104657737 B CN104657737 B CN 104657737B CN 201310590970 A CN201310590970 A CN 201310590970A CN 104657737 B CN104657737 B CN 104657737B
Authority
CN
China
Prior art keywords
image
code
intersection point
cluster
forecolor
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201310590970.7A
Other languages
Chinese (zh)
Other versions
CN104657737A (en
Inventor
孙科武
于志强
王启刚
唐凌
肖天柱
丁瑶
王杰斌
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Aisino Corp
Original Assignee
Aisino Corp
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 Aisino Corp filed Critical Aisino Corp
Priority to CN201310590970.7A priority Critical patent/CN104657737B/en
Publication of CN104657737A publication Critical patent/CN104657737A/en
Application granted granted Critical
Publication of CN104657737B publication Critical patent/CN104657737B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a kind of method and apparatus being corrected based on cluster to QR image in 2 D code.This method mainly includes：QR image in 2 D code after handling triple-expansion carries out Hough transform, obtains four borders of QR image in 2 D code；Fit four boundary straight line set of QR image in 2 D code respectively according to four borders, construct four intersection point set corresponding to four boundary straight line set respectively；Intersection point cluster analysis is carried out respectively to four intersection point set, obtains four final intersection point clusters, four apex coordinates of QR image in 2 D code are obtained by the central point for calculating four intersection point clusters；Conversion coefficient is calculated according to four apex coordinates, processing is corrected to QR image in 2 D code according to the conversion coefficient.The present invention can position the boundary straight line and apex coordinate of QR image in 2 D code exactly, and the correction of QR image in 2 D code is realized using geometry and perspective transform, effectively correct inclined QR image in 2 D code.

Description

The method and apparatus being corrected based on cluster to QR image in 2 D code
Technical field
The present invention relates to image in 2 D code technical field, more particularly to a kind of cluster that is based on to carry out school to QR image in 2 D code Positive method and apparatus.
Background technology
With the development of technology of Internet of things, Quick Response Code is applied increasingly as efficient information carrier in fields such as logistics Extensively.Wherein QR（Quick Response, fast reaction）Quick Response Code because of its abundant release type, a variety of fault-tolerant ranks and into Ripe encoding and decoding technique is got growing concern for, and quick correctly recognition QR Quick Response Codes also turn into recent years under complex environment Come the focus studied, wherein it is hot issue therein how the QR image in 2 D code of deformation to be corrected.
A kind of method being corrected to QR image in 2 D code of the prior art mainly includes：Sought by Hough transform The boundary straight line looked in QR image in 2 D code, multiple intersection point collection of QR image in 2 D code are calculated according to each boundary straight line, then The apex coordinate of QR Quick Response Codes is calculated according to multiple intersection point collection of QR image in 2 D code.Then, according to obtained apex coordinate Perspective transform and geometric transformation, the deformation pattern of the QR Quick Response Codes of correction of a final proof are carried out to QR image in 2 D code.
It is above-mentioned it is of the prior art QR image in 2 D code is corrected method the shortcomings that be：Due to QR image in 2 D code The block of pixels accumulated by some pixels, and the arrangement mode of pixel because information, version, tolerant level it is other it is different present with Machine, by the way that a plurality of noise straight line will be included in the straight line cluster of the QR image in 2 D code searched out after Hough transform, it is difficult to wrap Include four boundary straight lines that QR Quick Response Codes are determined in the straight line cluster of a plurality of noise straight line.
Due to the characteristic of Hough transform, the boundary straight line of the QR image in 2 D code searched out by Hough transform may have It is a plurality of, thus the intersection point collection calculated according to boundary straight line may have multiple and aggregation distribution is presented, in these multiple intersection points The apex coordinate how concentration accurately obtains QR Quick Response Codes is a difficult point.
The content of the invention
The embodiment provides a kind of based on the method and apparatus that are corrected to QR image in 2 D code of cluster, with Realization is effectively corrected to QR image in 2 D code.
The invention provides following scheme：
A kind of method being corrected based on cluster to QR image in 2 D code, including：
Triple-expansion processing is carried out to QR image in 2 D code, the QR image in 2 D code after expansion process is carried out Hough transform, obtain four borders of the QR image in 2 D code；
Fit four boundary straight line set of QR image in 2 D code respectively according to four borders, construct described four Four intersection point set corresponding to individual boundary straight line set respectively；
According to the initial vertax cluster and distance threshold of intersection point set set in advance, four intersection point set are entered respectively Row intersection point cluster analysis, four final intersection point clusters are obtained, the central point by calculating four intersection point clusters obtains QR two dimensions Four apex coordinates of code image；
Transformation series is calculated according to the standard coordinate of four apex coordinates of the QR image in 2 D code and Hough transform Number, processing is corrected according to the conversion coefficient to the QR image in 2 D code.
It is described to carry out triple-expansion processing to QR image in 2 D code, including：
Step 101：Obtain the gray value of each pixel in QR image in 2 D code；
Step 102：Length and width dimensions w, the h of acquisition QR image in 2 D code, initialization abscissa a=0, ordinate b=0, initially Change the foreground forecolor of QR image in 2 D code；
Step 103：Judge abscissa a whether more than QR image in 2 D code width w, if no more than if perform step 104, otherwise terminate；
Step 104：Judge ordinate b whether more than QR image in 2 D code width h, if no more than if perform step 105, otherwise perform step 108；
Step 105：Whether the gray value for judging pixel (a, b) is the foreground forecolor, is if it is performed Step 106, step 107 is otherwise performed；
Step 106：According to w, h judge pixel (a, b) whether be QR image in 2 D code boundary point, if not then holding Row 1061；If it is step 1062 is performed；
Step 1061：Gray scale replacement is carried out to 8 neighborhoods of pixel (a, b), by the pixel of 8 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a-1,b)=forecolor；
f(a-1,b-1)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b-1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 1062：Gray scale replacement is carried out to 4 neighborhoods of pixel (a, b), by the pixel of 4 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 107：Abscissa a is constant, ordinate b increases by 1, goes to step 104；
Step 108：Abscissa a increases by 1, go to step 103.
It is described to construct four intersection point set corresponding to four boundary straight line set respectively, including：
Step 301：Choose boundary straight line set L;
Step 302：Select any two straight lines l1, l2 in L;
Step 303：Judge whether l1 belongs to L, be then to perform step 304, otherwise terminate；
Step 304：Judge whether l2 belongs to L, and l1 is not equal to l2, is to perform 305, otherwise performs step 307；
Step 305：L1 and l2 intersection point are obtained according to the linear equation of two straight lines, judge whether the intersection point belongs to QR bis- Tie up on code image, if belonging to, the intersection point is had into intersection point set A-LIn, perform step 306；
Step 306：Another straight line is assigned to l2 in selection L, goes to step 304；
Step 307：Straight line is assigned to l1 in addition in selection L, goes to step 303；
Until all straight lines in the boundary straight line set L are all selected, by the intersection point set A-LAs described Intersection point set corresponding to boundary straight line set L.
The described initial vertax cluster and distance threshold according to intersection point set set in advance, to four intersection point set Intersection point cluster analysis is carried out respectively, is obtained four final intersection point clusters, is obtained by the central point for calculating four intersection point clusters Four apex coordinates of QR image in 2 D code, including：
Step 401：Choose the intersection point set A on the border of a QR image in 2 D code；
Step 402：Selected summit cluster Z at random, initialize Z central point G=(x, y), distance threshold t;
Step 403：All intersection points in intersection point set A are traveled through, calculate the distance d of any intersection point (i, j) and G points;
Step 404：Judge whether d is less than t, if then performing step 405；Otherwise, step 406 is performed；
Step 405：The intersection point (i, j) is added in Z, the central point for adjusting Z is
Step 406：Judge whether all intersection points in the intersection point set A complete by traversal, if it is, performing step 407, otherwise perform step 403；
Step 407：Judge whether all intersection points all in some summit cluster, if so, then using some summit cluster as Summit cluster corresponding to the intersection point set A, perform step 408；Otherwise, step 402 is performed；
Step 408：The center for calculating some summit cluster obtains an apex coordinate of QR image in 2 D code；
The intersection point set on the border of each QR image in 2 D code is chosen successively, according to above-mentioned processing procedure, obtains each friendship Summit cluster corresponding to point set difference, the center for calculating each summit cluster respectively obtain each summit seat of QR image in 2 D code Mark.
Described method also includes：
After all summit clusters corresponding to each intersection point set of QR image in 2 D code are obtained, according to any three summit clusters Principle point-blank does not remove the summit cluster of the redundancy on two summit cluster lines of centres at center, be finally left four Individual summit cluster, the center for calculating four summit clusters respectively obtain four apex coordinates of QR image in 2 D code.
A kind of device being corrected based on cluster to QR image in 2 D code, including：
Expansion process module, for carrying out triple-expansion processing to QR image in 2 D code, to the QR after expansion process Image in 2 D code carries out Hough transform, obtains four borders of the QR image in 2 D code；
Intersection point set constructing module, QR is fitted respectively for four borders according to acquired in the expansion process module Four boundary straight line set of image in 2 D code, construct four intersection point collection corresponding to four boundary straight line set respectively Close；
Apex coordinate acquisition module, it is right for the initial vertax cluster and distance threshold according to intersection point set set in advance Four intersection point set that the intersection point set constructing module obtains carry out intersection point cluster analysis respectively, obtain four final intersection points Cluster, the central point by calculating four intersection point clusters obtain four apex coordinates of QR image in 2 D code；
Correction process module, four tops for the QR image in 2 D code according to acquired in the apex coordinate acquisition module Point coordinates and the standard coordinate of Hough transform calculate conversion coefficient, according to the conversion coefficient to the QR image in 2 D code It is corrected processing.
Described expansion process module, specifically for performing following step：
Step 101：Obtain the gray value of each pixel in QR image in 2 D code；
Step 102：Length and width dimensions w, the h of acquisition QR image in 2 D code, initialization abscissa a=0, ordinate b=0, initially Change the foreground forecolor of QR image in 2 D code；
Step 103：Judge abscissa a whether more than QR image in 2 D code width w, if no more than if perform step 104, otherwise terminate；
Step 104：Judge ordinate b whether more than QR image in 2 D code width h, if no more than if perform step 105, otherwise perform step 108；
Step 105：Whether the gray value for judging pixel (a, b) is the foreground forecolor, is if it is performed Step 106, step 107 is otherwise performed；
Step 106：According to w, h judge pixel (a, b) whether be QR image in 2 D code boundary point, if not then holding Row 1061；If it is step 1062 is performed；
Step 1061：Gray scale replacement is carried out to 8 neighborhoods of pixel (a, b), by the pixel of 8 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a-1,b)=forecolor；
f(a-1,b-1)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b-1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 1062：Gray scale replacement is carried out to 4 neighborhoods of pixel (a, b), by the pixel of 4 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 107：Abscissa a is constant, ordinate b increases by 1, goes to step 104；
Step 108：Abscissa a increases by 1, go to step 103.
Described intersection point set constructing module, specifically for performing following step：
Step 301：Choose boundary straight line set L;
Step 302：Select any two straight lines l1, l2 in L;
Step 303：Judge whether l1 belongs to L, be then to perform step 304, otherwise terminate；
Step 304：Judge whether l2 belongs to L, and l1 is not equal to l2, is to perform 305, otherwise performs step 307；
Step 305：L1 and l2 intersection point are obtained according to the linear equation of two straight lines, judge whether the intersection point belongs to QR bis- Tie up on code image, if belonging to, the intersection point is had into intersection point set A-LIn, perform step 306；
Step 306：Another straight line is assigned to l2 in selection L, goes to step 304；
Step 307：Straight line is assigned to l1 in addition in selection L, goes to step 303；
Until all straight lines in the boundary straight line set L are all selected, by the intersection point set A-LAs described Intersection point set corresponding to boundary straight line set L.
Described apex coordinate acquisition module, specifically for performing following step：
Step 401：Choose the intersection point set A on the border of a QR image in 2 D code；
Step 402：Selected summit cluster Z at random, initialize Z central point G=(x, y), distance threshold t;
Step 403：All intersection points in intersection point set A are traveled through, calculate the distance d of any intersection point (i, j) and G points;
Step 404：Judge whether d is less than t, if then performing step 405；Otherwise, step 406 is performed；
Step 405：The intersection point (i, j) is added in Z, the central point for adjusting Z is
Step 406：Judge whether all intersection points in the intersection point set A complete by traversal, if it is, performing step 407, otherwise perform step 403；
Step 407：Judge whether all intersection points all in some summit cluster, if so, then using some summit cluster as Summit cluster corresponding to the intersection point set A, perform step 408；Otherwise, step 402 is performed；
Step 408：The center for calculating some summit cluster obtains an apex coordinate of QR image in 2 D code；
The intersection point set on the border of each QR image in 2 D code is chosen successively, according to above-mentioned processing procedure, obtains each friendship Summit cluster corresponding to point set difference, the center for calculating each summit cluster respectively obtain each summit seat of QR image in 2 D code Mark.
Described apex coordinate acquisition module, it is additionally operable to obtaining institute corresponding to each intersection point set of QR image in 2 D code After having summit cluster, removed according to the principle of the center of any three summit clusters not point-blank and connected at two summits cluster center The summit cluster of redundancy on line, finally it is left four summit clusters, the center for calculating four summit clusters respectively obtains QR Quick Response Codes Four apex coordinates of image.
The embodiment of the present invention passes through to QR Quick Response Codes it can be seen from the technical scheme provided by embodiments of the invention described above Image synthesis application expanding processing, Hough transform method and intersection point clustering method, QR two dimensions can be positioned exactly The boundary straight line and apex coordinate of code image, finally realize the correction of QR image in 2 D code using geometry and perspective transform.Realize Rapidly QR image in 2 D code is corrected, effectively corrects inclined QR image in 2 D code.
Brief description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment Accompanying drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for this For the those of ordinary skill of field, without having to pay creative labor, other can also be obtained according to these accompanying drawings Accompanying drawing.
Fig. 1 is a kind of place for method being corrected based on cluster to QR image in 2 D code that the embodiment of the present invention one provides Manage flow chart；
Fig. 2 is a kind of place for method that triple-expansion processing is carried out to QR image in 2 D code that the embodiment of the present invention one provides Manage flow chart；
Fig. 3 is that a kind of boundary straight line set for QR image in 2 D code according to fitting that the embodiment of the present invention one provides determines The process chart of the method for intersection point set；
Fig. 4 is a kind of apex coordinate that acquisition QR image in 2 D code is clustered by intersection point that the embodiment of the present invention one provides The process chart of method；
Fig. 5 is that the embodiment of the present invention two provides a kind of knot for the device being corrected based on cluster to QR image in 2 D code Composition.
Embodiment
For ease of the understanding to the embodiment of the present invention, done further by taking several specific embodiments as an example below in conjunction with accompanying drawing Explanation, and each embodiment does not form the restriction to the embodiment of the present invention.
Embodiment one
This embodiment offers a kind of handling process such as Fig. 1 for the method being corrected based on cluster to QR image in 2 D code Shown including following processing step：
Step 1：QR image in 2 D code information is got, triple-expansion processing is carried out to QR image in 2 D code, eliminates QR bis- Tieing up influences the noise information of boundary definition in code image.The specific expansion flow of above-mentioned triple-expansion processing is as shown in Figure 2.
Step 2：Hough (Ha Fu) conversion is carried out to the QR image in 2 D code after expansion process, determines QR image in 2 D code Four borders, be fitted the boundary straight line set of QR image in 2 D code respectively according to four borders, obtain and aforementioned four side Four boundary straight line set corresponding to boundary's difference, each border correspond to a boundary straight line set respectively.
After expansion process, the detail section such as light and shade module of QR Quick Response Codes is inflated, and is had and is compared clearly border, root According to the ordinate of the arbitrfary point on straight line be multiplied by cosine and abscissa be multiplied by sine and be a definite value, it is determined that after expanding QR image in 2 D code all straight lines, finally select straight length exceed picture size straight line set, these set are shape Gather into the border of QR image in 2 D code.
Step 3：Any two straight lines are extracted from above-mentioned boundary straight line set, and obtain the friendship of above-mentioned any two straight lines Point, by intersection point set corresponding to the above-mentioned boundary straight line set of obtained all intersection points composition, construct four boundary straight lines Four intersection point set corresponding to set difference.Intersection point set is determined according to the boundary straight line set of the QR image in 2 D code of fitting Specific handling process is as shown in Figure 3.
So as to obtain four boundary straight line set, intersection point set corresponding to the difference of aforementioned four border.
Then, the illegal intersection point in above-mentioned intersection point set is eliminated according to the boundary condition of the QR image in 2 D code of setting.It is main Eliminate the intersection point between same border difference straight line and the intersection point on QR image in 2 D code opposite side border.According to QR image in 2 D code Border feature, two intersection points that any three adjacent boundaries are formed are located at the both ends of medial side, and identical border difference straight line Intersection point is located at the middle part on the border, so as to be eliminated.The intersection point of any opposite side is then according to QR image in 2 D code data area Size judged, if the intersection point calculated outside image, for illegal intersection point.
Step 4：For each intersection point set for eliminating illegal intersection point, according to the thought of maximum neighborhood, intersection point collection is set The initial vertax cluster and distance threshold of conjunction, antinode set carry out intersection point cluster analysis, obtain final intersection point cluster, calculate final Intersection point cluster central point, that is, obtain the apex coordinate of a QR image in 2 D code, the apex coordinate and above-mentioned intersection point set pair Should.
According to above-mentioned processing procedure, four summits that QR image in 2 D code can be respectively obtained based on four intersection point set are sat Mark.
Step 5：Four apex coordinates of the QR image in 2 D code obtained according to step 4 and the standard coordinate of Hough transform Calculation of transform coefficients, according to the above-mentioned QR image in 2 D code of conversion coefficient correction, at the correction for completing above-mentioned QR image in 2 D code Reason.
The fixed point coordinate obtained according to step 4 is（x,y）, calibration coordinate（That is the standard coordinate of Hough transform）For（x′, y′）, pass through transformation for mula x '=c1x+c2y+c3xy+c4,y′=c5x+c6y+c7xy+c8, calculate 8 conversion coefficients（c1,c2,c3, c4,c5,c6,c7,c8）Afterwards, processing is corrected to the Point Coordinates of original image using above-mentioned 8 conversion coefficients.
A kind of handling process such as Fig. 2 for method that triple-expansion processing is carried out to QR image in 2 D code that the embodiment provides Shown including following processing step：
Step 101：Obtain the gray value of each pixel in QR image in 2 D code.
Step 102：Length and width dimensions w, the h of acquisition QR image in 2 D code, initialization abscissa a=0, ordinate b=0, initially Change the foreground forecolor of QR image in 2 D code, foreground is arbitrarily set, and general foreground is 0（Black）, background colour 1 （White）.
Step 103：Judge abscissa a whether more than QR image in 2 D code width w, if no more than if perform step 104, otherwise terminate；
Step 104：Judge ordinate b whether more than QR image in 2 D code width h, if no more than if perform step 105, otherwise perform step 108；
Step 105：Whether the gray value for judging pixel (a, b) is foreground, that is, is judged, f (a, b)=forecolor, Whether set up, if it is perform step 106, otherwise perform step 107；
Step 106：According to w, h judges whether pixel (a, b) is boundary point, if not then performing 1061；If Then perform step 1062；
Step 1061：Gray scale replacement is carried out to 8 neighborhoods of pixel (a, b), by the pixel of 8 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a-1,b)=forecolor；
f(a-1,b-1)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b-1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 1062：Gray scale replacement is carried out to 4 neighborhoods of pixel (a, b), by the pixel of 4 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 107：Abscissa a is constant, ordinate b increases by 1, goes to step 104；
Step 108：Abscissa a is reset, and increase by 1, goes to step 103.
A kind of boundary straight line set for QR image in 2 D code according to fitting that the embodiment provides determines intersection point set The handling process of method is as shown in figure 3, including following processing step：
Step 301：Obtain the boundary straight line set L that step 2 obtains;
Step 302：Select any two straight lines l1, l2 in L;
Step 303：Judge whether l1 belongs to L, be then to perform step 304, otherwise terminate；
Step 304：Judge whether l2 belongs to L, and l1 is not equal to l2, is to perform 305, otherwise performs step 307；
Step 305：L1 and l2 intersection point are obtained according to the linear equation of two straight lines, judge whether the intersection point belongs to QR bis- Tie up on code image, if belonging to, the intersection point is had into intersection point set A-LIn；
Step 306：Another straight line is assigned to l2 in selection L, goes to step 304；
Step 307：Straight line is assigned to l1 in addition in selection L, goes to step 303.
Until all straight lines in the boundary straight line set L are all selected, by the intersection point set A-LAs described Intersection point set corresponding to boundary straight line set L.
For the intersection point for eliminating an illegal intersection point set, accurate QR Quick Response Codes figure how is obtained in intersection point set The apex coordinate of picture is the key of QR two dimension code conversions.Due to the point coordinates in intersection point set all in certain threshold range ripple It is dynamic, by calculate intersecting point coordinate apart from automatic cluster into an intersection point cluster, the center for finally calculating intersection point cluster just obtains QR bis- Tie up the apex coordinate of code conversion.
A kind of processing for method that the apex coordinate for obtaining QR image in 2 D code is clustered by intersection point that the embodiment provides Flow is as shown in figure 4, including following processing step：
Step 401：Choose the intersection point set A on the border of a QR image in 2 D code for eliminating illegal intersection point.
Step 402：Selected summit cluster Z at randomi, initialize ZiCentral point G=(x, y), distance threshold t;
Step 403：All intersection points in intersection point set A are traveled through, calculate the distance d of any intersection point (i, j) and G points;
Step 404：Judge whether d is less than t, if then performing step 405；Otherwise, step 406 is performed；
Step 405：Intersection point (i, j) is added in Z, the central point for recalculating the Coordinate Adjusting Z of Z central point is
Step 406：Whether judge whether step 405 to being carried out a little one time, if then performing step 407, otherwise Perform step 403；
Step 407：Judge whether all intersection points all in some summit cluster, if so, then making some summit cluster For summit cluster corresponding to the intersection point set A, step 408 is performed：Otherwise step 402 is performed.
Step 408：The intersection point set on the border of each QR image in 2 D code is chosen successively, according to above-mentioned processing procedure, is divided Summit cluster corresponding to each intersection point set is not obtained.
After all summit clusters corresponding to each intersection point set of QR image in 2 D code are obtained, according to any three summit clusters Principle point-blank does not remove the summit cluster of the redundancy on two summit cluster lines of centres at center, be finally left four Individual summit cluster, the center for calculating four summit clusters respectively obtain four apex coordinates of QR image in 2 D code：(x0,y0),(x1, y1),(x2,y2),(x3,y3)。
Embodiment two
This embodiment offers a kind of device being corrected based on cluster to QR image in 2 D code, and it implements structure As shown in figure 5, it can specifically include following module：
Expansion process module 51, for carrying out triple-expansion processing to QR image in 2 D code, described in after expansion process QR image in 2 D code carries out Hough transform, obtains four borders of the QR image in 2 D code；
Intersection point set constructing module 52, fitted respectively for four borders according to acquired in the expansion process module Four boundary straight line set of QR image in 2 D code, construct four intersection point collection corresponding to four boundary straight line set respectively Close；
Apex coordinate acquisition module 53, for the initial vertax cluster and distance threshold according to intersection point set set in advance, The four intersection point set obtained to the intersection point set constructing module carry out intersection point cluster analysis respectively, obtain four final friendships Point cluster, the central point by calculating four intersection point clusters obtain four apex coordinates of QR image in 2 D code；
Correction process module 54, four for the QR image in 2 D code according to acquired in the apex coordinate acquisition module Apex coordinate and the standard coordinate of Hough transform calculate conversion coefficient, according to the conversion coefficient to the QR Quick Response Codes figure As being corrected processing.
Further, described expansion process module 51, specifically for performing following step：
Step 101：Obtain the gray value of each pixel in QR image in 2 D code；
Step 102：Length and width dimensions w, the h of acquisition QR image in 2 D code, initialization abscissa a=0, ordinate b=0, initially Change the foreground forecolor of QR image in 2 D code；
Step 103：Judge abscissa a whether more than QR image in 2 D code width w, if no more than if perform step 104, otherwise terminate；
Step 104：Judge ordinate b whether more than QR image in 2 D code width h, if no more than if perform step 105, otherwise perform step 108；
Step 105：Whether the gray value for judging pixel (a, b) is the foreground forecolor, is if it is performed Step 106, step 107 is otherwise performed；
Step 106：According to w, h judge pixel (a, b) whether be QR image in 2 D code boundary point, if not then holding Row 1061；If it is step 1062 is performed；
Step 1061：Gray scale replacement is carried out to 8 neighborhoods of pixel (a, b), by the pixel of 8 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a-1,b)=forecolor；
f(a-1,b-1)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b-1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 1062：Gray scale replacement is carried out to 4 neighborhoods of pixel (a, b), by the pixel of 4 neighborhoods of pixel (a, b) The gray value of point is set to foreground forecolor, i.e.,：
f(a,b)=forecolor；
f(a+1,b)=forecolor；
f(a+1,b+1)=forecolor；
f(a,b+1)=forecolor；
Then step 107 is performed；
Step 107：Abscissa a is constant, ordinate b increases by 1, goes to step 104；
Step 108：Abscissa a increases by 1, go to step 103.
Further, described intersection point set constructing module 52, specifically for performing following step：
Step 301：Choose boundary straight line set L;
Step 302：Select any two straight lines l1, l2 in L;
Step 303：Judge whether l1 belongs to L, be then to perform step 304, otherwise terminate；
Step 304：Judge whether l2 belongs to L, and l1 is not equal to l2, is to perform 305, otherwise performs step 307；
Step 305：L1 and l2 intersection point are obtained according to the linear equation of two straight lines, judge whether the intersection point belongs to QR bis- Tie up on code image, if belonging to, the intersection point is had into intersection point set A-LIn, perform step 306；
Step 306：Another straight line is assigned to l2 in selection L, goes to step 304；
Step 307：Straight line is assigned to l1 in addition in selection L, goes to step 303；
Until all straight lines in the boundary straight line set L are all selected, by the intersection point set A-LAs described Intersection point set corresponding to boundary straight line set L.
Further, described apex coordinate acquisition module 53, specifically for performing following step：
Step 401：Choose the intersection point set A on the border of a QR image in 2 D code；
Step 402：Selected summit cluster Z at random, initialize Z central point G=(x, y), distance threshold t;
Step 403：All intersection points in intersection point set A are traveled through, calculate the distance d of any intersection point (i, j) and G points;
Step 404：Judge whether d is less than t, if then performing step 405；Otherwise, step 406 is performed；
Step 405：The intersection point (i, j) is added in Z, the central point for adjusting Z is
Step 406：Judge whether all intersection points in the intersection point set A complete by traversal, if it is, performing step 407, otherwise perform step 403；
Step 407：Judge whether all intersection points all in some summit cluster, if so, then using some summit cluster as Summit cluster corresponding to the intersection point set A, perform step 408；Otherwise, step 402 is performed；
Step 408：The center for calculating some summit cluster obtains an apex coordinate of QR image in 2 D code；
The intersection point set on the border of each QR image in 2 D code is chosen successively, according to above-mentioned processing procedure, obtains each friendship Summit cluster corresponding to point set difference, the center for calculating each summit cluster respectively obtain each summit seat of QR image in 2 D code Mark.
Further, described apex coordinate acquisition module 53, it is additionally operable to obtaining each intersection point of QR image in 2 D code After all summit clusters corresponding to set, removed according to the principle of the center of any three summit clusters not point-blank at two The summit cluster of redundancy on the cluster line of centres of summit, finally it is left four summit clusters, calculates the center of four summit clusters respectively Obtain four apex coordinates of QR image in 2 D code.
With the device of the embodiment of the present invention QR image in 2 D code is corrected based on cluster the detailed process of processing Similar with preceding method embodiment, here is omitted.
In summary, the embodiment of the present invention passes through to QR image in 2 D code integrated application expanding processings, Hough transform Method and intersection point clustering method, the boundary straight line and apex coordinate of QR image in 2 D code can be positioned exactly, finally should The correction of QR image in 2 D code is realized with geometry and perspective transform.Realize and rapidly QR image in 2 D code is corrected, have Effect ground corrects inclined QR image in 2 D code.
The embodiment of the present invention to QR image in 2 D code by carrying out expansion process, after effectively can eliminating Hough transform Noise straight line in the straight line cluster of obtained QR image in 2 D code, and then the four of QR Quick Response Codes can be determined by above-mentioned straight line cluster Individual boundary straight line.
The embodiment of the present invention carries out intersection point cluster analysis by searching out multiple intersection point collection to Hough transform, and combines superfluous Remaining summit cluster rejecting processing, it is achieved thereby that concentrating the apex coordinate for how being accurately positioned QR Quick Response Codes by multiple intersection points.
One of ordinary skill in the art will appreciate that：Accompanying drawing is the schematic diagram of one embodiment, module in accompanying drawing or Flow is not necessarily implemented necessary to the present invention.
As seen through the above description of the embodiments, those skilled in the art can be understood that the present invention can Realized by the mode of software plus required general hardware platform.Based on such understanding, technical scheme essence On the part that is contributed in other words to prior art can be embodied in the form of software product, the computer software product It can be stored in storage medium, such as ROM/RAM, magnetic disc, CD, including some instructions are causing a computer equipment （Can be personal computer, server, or network equipment etc.）Perform some of each embodiment or embodiment of the invention Method described in part.
Each embodiment in this specification is described by the way of progressive, identical similar portion between each embodiment Divide mutually referring to what each embodiment stressed is the difference with other embodiment.Especially for device or For system embodiment, because it is substantially similar to embodiment of the method, so describing fairly simple, related part is referring to method The part explanation of embodiment.Apparatus and system embodiment described above is only schematical, wherein the conduct The unit that separating component illustrates can be or may not be it is physically separate, can be as the part that unit is shown or Person may not be physical location, you can with positioned at a place, or can also be distributed on multiple NEs.Can root Factually border needs to select some or all of module therein realize the purpose of this embodiment scheme.Ordinary skill Personnel are without creative efforts, you can to understand and implement.
The foregoing is only a preferred embodiment of the present invention, but protection scope of the present invention be not limited thereto, Any one skilled in the art the invention discloses technical scope in, the change or replacement that can readily occur in, It should all be included within the scope of the present invention.Therefore, protection scope of the present invention should be with scope of the claims It is defined.

Claims (10)

  1. A kind of 1. method being corrected based on cluster to QR image in 2 D code, it is characterised in that methods described includes：
    Gray value based on foreground as pixel carries out triple-expansion processing to QR image in 2 D code, after expansion process The QR image in 2 D code carries out Hough transform, obtains four borders of the QR image in 2 D code；
    Fit four boundary straight line set of QR image in 2 D code respectively according to four borders, construct four sides Four intersection point set corresponding to boundary's straight line set respectively；
    According to the initial vertax cluster and distance threshold of intersection point set set in advance, four intersection point set are handed over respectively Point cluster analysis, obtains four final intersection point clusters, the central point by calculating four intersection point clusters obtains QR Quick Response Code figures Four apex coordinates of picture；
    Conversion coefficient, root are calculated according to the standard coordinate of four apex coordinates of the QR image in 2 D code and Hough transform Processing is corrected to the QR image in 2 D code according to the conversion coefficient.
  2. 2. the method according to claim 1 being corrected based on cluster to QR image in 2 D code, it is characterised in that described To QR image in 2 D code carry out triple-expansion processing, including：
    Step 101：Obtain the gray value of each pixel in QR image in 2 D code；
    Step 102：Length and width dimensions w, the h of QR image in 2 D code are obtained, initializes abscissa a=0, ordinate b=0, initialization The foreground forecolor of QR image in 2 D code；
    Step 103：Judge abscissa a whether more than QR image in 2 D code width w, if no more than if perform step 104, it is no Then terminate；
    Step 104：Judge ordinate b whether more than QR image in 2 D code width h, if no more than if perform step 105, it is no Then perform step 108；
    Step 105：Whether the gray value for judging pixel (a, b) is the foreground forecolor, if it is performs step 106, otherwise perform step 107；
    Step 106：According to w, h judge pixel (a, b) whether be QR image in 2 D code boundary point, if not then performing 1061；If it is step 1062 is performed；
    Step 1061：Gray scale replacement is carried out to 8 neighborhoods of pixel (a, b), by the pixel of 8 neighborhoods of pixel (a, b) Gray value is set to foreground forecolor, i.e.,：
    F (a, b)=forecolor；
    F (a-1, b)=forecolor；
    F (a-1, b-1)=forecolor；
    F (a+1, b)=forecolor；
    F (a+1, b+1)=forecolor；
    F (a, b-1)=forecolor；
    F (a, b+1)=forecolor；
    Then step 107 is performed；
    Step 1062：Gray scale replacement is carried out to 4 neighborhoods of pixel (a, b), by the pixel of 4 neighborhoods of pixel (a, b) Gray value is set to foreground forecolor, i.e.,：
    F (a, b)=forecolor；
    F (a+1, b)=forecolor；
    F (a+1, b+1)=forecolor；
    F (a, b+1)=forecolor；
    Then step 107 is performed；
    Step 107：Abscissa a is constant, ordinate b increases by 1, goes to step 104；
    Step 108：Abscissa a increases by 1, go to step 103.
  3. 3. the method according to claim 1 being corrected based on cluster to QR image in 2 D code, it is characterised in that described Construct four intersection point set corresponding to four boundary straight line set respectively, including：
    Step 301：Choose boundary straight line set L；
    Step 302：Select any two straight lines l1, l2 in L；
    Step 303：Judge whether l1 belongs to L, be then to perform step 304, otherwise terminate；
    Step 304：Judge whether l2 belongs to L, and l1 is not equal to l2, is to perform 305, otherwise performs step 307；
    Step 305：L1 and l2 intersection point are obtained according to the linear equation of two straight lines, judge whether the intersection point belongs to QR Quick Response Codes On image, if belonging to, intersection point set A is present into the intersection point-LIn, perform step 306；
    Step 306：Another straight line is assigned to l2 in selection L, goes to step 304；
    Step 307：Straight line is assigned to l1 in addition in selection L, goes to step 303；
    Until all straight lines in the boundary straight line set L are all selected, by the intersection point set A-LAs the border Intersection point set corresponding to straight line set L.
  4. 4. the method being corrected based on cluster to QR image in 2 D code according to claim 1 or 2 or 3, its feature are existed In the initial vertax cluster and distance threshold according to intersection point set set in advance, to four intersection point set difference Intersection point cluster analysis is carried out, four final intersection point clusters is obtained, QR bis- is obtained by the central point for calculating four intersection point clusters Four apex coordinates of code image are tieed up, including：
    Step 401：Choose the intersection point set A on the border of a QR image in 2 D code；
    Step 402：Selected summit cluster Z at random, initialize Z central point G=(x, y), distance threshold t；
    Step 403：All intersection points in intersection point set A are traveled through, calculate the distance d of any intersection point (i, j) and G points；
    <mrow> <mi>d</mi> <mo>=</mo> <msqrt> <mrow> <msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>-</mo> <mi>i</mi> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>+</mo> <msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>-</mo> <mi>j</mi> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </msqrt> <mo>;</mo> </mrow>
    Step 404：Judge whether d is less than t, if then performing step 405；Otherwise, step 406 is performed；
    Step 405：The intersection point (i, j) is added in Z, the central point for adjusting Z is
    Step 406：Judge whether all intersection points in the intersection point set A complete by traversal, if it is, step 407 is performed, Otherwise step 403 is performed；
    Step 407：Judge whether all intersection points all in some summit cluster, if so, then using some summit cluster as described in Summit cluster corresponding to intersection point set A, perform step 408；Otherwise, step 402 is performed；
    Step 408：The center for calculating some summit cluster obtains an apex coordinate of QR image in 2 D code；
    The intersection point set on the border of each QR image in 2 D code is chosen successively, according to above-mentioned processing procedure, obtains each intersection point collection Summit cluster corresponding to difference is closed, the center for calculating each summit cluster respectively obtains each apex coordinate of QR image in 2 D code.
  5. 5. the method according to claim 4 being corrected based on cluster to QR image in 2 D code, it is characterised in that described Method also include：
    After all summit clusters corresponding to each intersection point set of QR image in 2 D code are obtained, according in any three summit clusters The principle of the heart not point-blank removes the summit cluster of the redundancy on two summit cluster lines of centres, final remaining four tops Point cluster, the center for calculating four summit clusters respectively obtains four apex coordinates of QR image in 2 D code.
  6. 6. a kind of device being corrected based on cluster to QR image in 2 D code, it is characterised in that described device includes：
    Expansion process module, for being carried out based on gray value of the foreground as pixel to QR image in 2 D code at triple-expansion Reason, Hough transform is carried out to the QR image in 2 D code after expansion process, obtains four sides of the QR image in 2 D code Boundary；
    Intersection point set constructing module, QR two dimensions are fitted respectively for four borders according to acquired in the expansion process module Four boundary straight line set of code image, construct four intersection point set corresponding to four boundary straight line set respectively；
    Apex coordinate acquisition module, for the initial vertax cluster and distance threshold according to intersection point set set in advance, to described Four intersection point set that intersection point set constructing module obtains carry out intersection point cluster analysis respectively, obtain four final intersection point clusters, Central point by calculating four intersection point clusters obtains four apex coordinates of QR image in 2 D code；
    Correction process module, four summits for the QR image in 2 D code according to acquired in the apex coordinate acquisition module are sat Mark and the standard coordinate of Hough transform calculate conversion coefficient, and the QR image in 2 D code is carried out according to the conversion coefficient Correction process.
  7. 7. the device according to claim 6 being corrected based on cluster to QR image in 2 D code, it is characterised in that：
    Described expansion process module, specifically for performing following step：
    Step 101：Obtain the gray value of each pixel in QR image in 2 D code；
    Step 102：Length and width dimensions w, the h of QR image in 2 D code are obtained, initializes abscissa a=0, ordinate b=0, initialization The foreground forecolor of QR image in 2 D code；
    Step 103：Judge abscissa a whether more than QR image in 2 D code width w, if no more than if perform step 104, it is no Then terminate；
    Step 104：Judge ordinate b whether more than QR image in 2 D code width h, if no more than if perform step 105, it is no Then perform step 108；
    Step 105：Whether the gray value for judging pixel (a, b) is the foreground forecolor, if it is performs step 106, otherwise perform step 107；
    Step 106：According to w, h judge pixel (a, b) whether be QR image in 2 D code boundary point, if not then performing 1061；If it is step 1062 is performed；
    Step 1061：Gray scale replacement is carried out to 8 neighborhoods of pixel (a, b), by the pixel of 8 neighborhoods of pixel (a, b) Gray value is set to foreground forecolor, i.e.,：
    F (a, b)=forecolor；
    F (a-1, b)=forecolor；
    F (a-1, b-1)=forecolor；
    F (a+1, b)=forecolor；
    F (a+1, b+1)=forecolor；
    F (a, b-1)=forecolor；
    F (a, b+1)=forecolor；
    Then step 107 is performed；
    Step 1062：Gray scale replacement is carried out to 4 neighborhoods of pixel (a, b), by the pixel of 4 neighborhoods of pixel (a, b) Gray value is set to foreground forecolor, i.e.,：
    F (a, b)=forecolor；
    F (a+1, b)=forecolor；
    F (a+1, b+1)=forecolor；
    F (a, b+1)=forecolor；
    Then step 107 is performed；
    Step 107：Abscissa a is constant, ordinate b increases by 1, goes to step 104；
    Step 108：Abscissa a increases by 1, go to step 103.
  8. 8. the device according to claim 6 being corrected based on cluster to QR image in 2 D code, it is characterised in that：
    Described intersection point set constructing module, specifically for performing following step：
    Step 301：Choose boundary straight line set L；
    Step 302：Select any two straight lines l1, l2 in L；
    Step 303：Judge whether l1 belongs to L, be then to perform step 304, otherwise terminate；
    Step 304：Judge whether l2 belongs to L, and l1 is not equal to l2, is to perform 305, otherwise performs step 307；
    Step 305：L1 and l2 intersection point are obtained according to the linear equation of two straight lines, judge whether the intersection point belongs to QR Quick Response Codes On image, if belonging to, intersection point set A is present into the intersection point-LIn, perform step 306；
    Step 306：Another straight line is assigned to l2 in selection L, goes to step 304；
    Step 307：Straight line is assigned to l1 in addition in selection L, goes to step 303；
    Until all straight lines in the boundary straight line set L are all selected, by the intersection point set A-LAs the border Intersection point set corresponding to straight line set L.
  9. 9. the device being corrected based on cluster to QR image in 2 D code according to claim 6 or 7 or 8, its feature are existed In：
    Described apex coordinate acquisition module, specifically for performing following step：
    Step 401：Choose the intersection point set A on the border of a QR image in 2 D code；
    Step 402：Selected summit cluster Z at random, initialize Z central point G=(x, y), distance threshold t；
    Step 403：All intersection points in intersection point set A are traveled through, calculate the distance d of any intersection point (i, j) and G points；
    <mrow> <mi>d</mi> <mo>=</mo> <msqrt> <mrow> <msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>-</mo> <mi>i</mi> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>+</mo> <msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>-</mo> <mi>j</mi> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </msqrt> <mo>;</mo> </mrow>
    Step 404：Judge whether d is less than t, if then performing step 405；Otherwise, step 406 is performed；
    Step 405：The intersection point (i, j) is added in Z, the central point for adjusting Z is
    Step 406：Judge whether all intersection points in the intersection point set A complete by traversal, if it is, step 407 is performed, Otherwise step 403 is performed；
    Step 407：Judge whether all intersection points all in some summit cluster, if so, then using some summit cluster as described in Summit cluster corresponding to intersection point set A, perform step 408；Otherwise, step 402 is performed；
    Step 408：The center for calculating some summit cluster obtains an apex coordinate of QR image in 2 D code；
    The intersection point set on the border of each QR image in 2 D code is chosen successively, according to above-mentioned processing procedure, obtains each intersection point collection Summit cluster corresponding to difference is closed, the center for calculating each summit cluster respectively obtains each apex coordinate of QR image in 2 D code.
  10. 10. the device according to claim 9 being corrected based on cluster to QR image in 2 D code, it is characterised in that：
    Described apex coordinate acquisition module, it is additionally operable to obtaining all tops corresponding to each intersection point set of QR image in 2 D code After point cluster, removed according to the principle of the center of any three summit clusters not point-blank on two summit cluster lines of centres Redundancy summit cluster, finally be left four summit clusters, the center for calculating four summit clusters respectively obtains QR image in 2 D code Four apex coordinates.
CN201310590970.7A 2013-11-20 2013-11-20 The method and apparatus being corrected based on cluster to QR image in 2 D code Active CN104657737B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310590970.7A CN104657737B (en) 2013-11-20 2013-11-20 The method and apparatus being corrected based on cluster to QR image in 2 D code

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310590970.7A CN104657737B (en) 2013-11-20 2013-11-20 The method and apparatus being corrected based on cluster to QR image in 2 D code

Publications (2)

Publication Number Publication Date
CN104657737A CN104657737A (en) 2015-05-27
CN104657737B true CN104657737B (en) 2018-01-16

Family

ID=53248837

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310590970.7A Active CN104657737B (en) 2013-11-20 2013-11-20 The method and apparatus being corrected based on cluster to QR image in 2 D code

Country Status (1)

Country Link
CN (1) CN104657737B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105069394B (en) * 2015-07-23 2017-10-10 福建联迪商用设备有限公司 Quick Response Code weighted average gray level method coding/decoding method and system
CN106022333A (en) * 2016-04-26 2016-10-12 安凯 Vehicle license plate tilt image correcting method
CN107025455B (en) * 2017-04-01 2019-12-24 浙江华睿科技有限公司 Positioning method and device for quickly reflecting QR (quick response) code area
CN107895138B (en) * 2017-10-13 2020-06-23 西安艾润物联网技术服务有限责任公司 Method and device for detecting space obstacle and computer readable storage medium
CN109815762B (en) * 2018-12-29 2022-02-11 福建天泉教育科技有限公司 Method and storage medium for remotely identifying two-dimensional code
CN109815763A (en) * 2019-01-04 2019-05-28 广州广电研究院有限公司 Detection method, device and the storage medium of two dimensional code
CN113392669B (en) * 2021-05-31 2022-09-23 苏州中科华影健康科技有限公司 Image information detection method, detection device and storage medium
CN113536822B (en) * 2021-07-28 2024-05-03 中移（杭州）信息技术有限公司 Two-dimensional code correction method and device and computer readable storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7079702B2 (en) * 2001-01-30 2006-07-18 Fujitsu Limited Image processing method and apparatus for judging image condition and correcting image
CN101059834A (en) * 2007-05-31 2007-10-24 中国农业大学 QR two-dimensional bar code recognition method based on pickup head for chatting
CN101964045A (en) * 2010-09-30 2011-02-02 西北工业大学 Distortion correction method of data matrix code marked on cylindrical surface product
CN103177416A (en) * 2013-04-10 2013-06-26 浙江大学 QR (quick response) code image locating method based on least square method

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7079702B2 (en) * 2001-01-30 2006-07-18 Fujitsu Limited Image processing method and apparatus for judging image condition and correcting image
CN101059834A (en) * 2007-05-31 2007-10-24 中国农业大学 QR two-dimensional bar code recognition method based on pickup head for chatting
CN101964045A (en) * 2010-09-30 2011-02-02 西北工业大学 Distortion correction method of data matrix code marked on cylindrical surface product
CN103177416A (en) * 2013-04-10 2013-06-26 浙江大学 QR (quick response) code image locating method based on least square method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
"基于Hough变换的二维条码图像倾斜校正研究";郑河荣 等;《浙江工业大学学报》;20130428(第02期);第169-172页 *

Also Published As

Publication number Publication date
CN104657737A (en) 2015-05-27

Similar Documents

Publication Publication Date Title
CN104657737B (en) The method and apparatus being corrected based on cluster to QR image in 2 D code
CN109816012B (en) Multi-scale target detection method fusing context information
CN110245655B (en) Single-stage object detection method based on lightweight image pyramid network
CN102999913B (en) A kind of sectional perspective matching process based on credible propagation
CN106462724B (en) Method and system based on normalized images verification face-image
CN111507901B (en) Aerial image splicing and positioning method based on aerial GPS and scale invariant constraint
Zhou et al. Octr: Octree-based transformer for 3d object detection
CN111723947A (en) Method and device for training federated learning model
CN106991355A (en) The face identification method of the analytical type dictionary learning model kept based on topology
CN111275192B (en) Auxiliary training method for improving accuracy and robustness of neural network simultaneously
CN106557579A (en) A kind of vehicle model searching system and method based on convolutional neural networks
CN104616247B (en) A kind of method for map splicing of being taken photo by plane based on super-pixel SIFT
CN103325130B (en) Geometric iteration image approximating method based on T batten
CN104143186B (en) A kind of SLIC super-pixel piecemeal optimization method
CN107346550A (en) It is a kind of to be directed to the three dimensional point cloud rapid registering method with colouring information
CN110084743B (en) Image splicing and positioning method based on multi-flight-zone initial flight path constraint
CN109657081A (en) Distributed approach, system and the medium of EO-1 hyperion satellite remote sensing date
CN108122280A (en) The method for reconstructing and device of a kind of three-dimensional point cloud
CN113378721B (en) Symmetrical and local discrimination-based face correction method and system for generating countermeasure
CN105513014A (en) Method and system for reconstruction of multiframe image super resolution
CN111553296B (en) Two-value neural network stereo vision matching method based on FPGA
CN110503113A (en) A kind of saliency object detection method restored based on low-rank matrix
CN108961385A (en) A kind of SLAM patterning process and device
CN110826534B (en) Face key point detection method and system based on local principal component analysis
CN103679639A (en) Image denoising method and device based on non-local mean value

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