CN101093544A - Method for correcting pattern in 2D code of perspective quick response matrix - Google Patents

Abstract

A method for calibrating 2-D code pattern of perspective quick response matrix includes obtaining 2-D code pattern of perspective quick response matrix, obtaining convex besetment box of said 2-D code pattern according to SKlansky algorithm, seeking out four top points under image coordinate at top of besetment box, applying singular value decomposing means to estimate perspective transform parameter and carrying out re-sampling on original image according to obtained transform parameter for obtaining calibrated 2-D code pattern of quick response matrix.

Description

A kind of method and system of having an X-rayed the 2 D code of quick response matrix correcting pattern

Technical field

The present invention relates to network communications technology field, relate in particular to a kind of method and system of perspective QR (2 D code of quick response matrix) correcting pattern.

Background technology

At present, the reference decoding algorithm that " quick response matrix code QR Code " (State Standard of the People's Republic of China GB/T18284-2000) provides is at first determined view finding figure (Finder Pattern according to horizontal and vertical scanning, FP), according to the view finding figure, determine correction graph (AlignmentPattern, AP), set up sampling grid according to FP and AP then, obtain the preceding information of formats.In practice, general except horizontal and vertical scanning for the image of QR sign indicating number pattern in order accurately to obtain the position of view finding figure through obtaining after the rotational transform, also increase the crisscross scanning of 45 degree.

Yet because the restriction of image-forming condition, QR sign indicating number pattern tends to produce perspective transform, and above method promptly allows to successfully find FP, also can't obtain correct AP position and sampling grid according to the reference decoding algorithm.Therefore, solving the decoding difficulty that perspective transform brings is the problem that many application can't be avoided.

Summary of the invention

Technical matters to be solved by this invention provides a kind of method and system of the QR of perspective sign indicating number correcting pattern.

For solving the problems of the technologies described above, the present invention is achieved by the following technical solutions:

A kind of method of having an X-rayed the quick response matrix code correcting pattern may further comprise the steps:

A: obtain perspective QR sign indicating number pattern;

B: ask for the protruding bounding box of QR sign indicating number pattern according to Sklansky algorithm (a kind of algorithm known), in the summit of bounding box, seek four summits under the image coordinate, adopt SVD (svd) to estimate the perspective transform parameter;

C: according to the perspective transform parameter that obtains, original image is carried out resampling, the QR sign indicating number pattern that obtains proofreading and correct.

Wherein, also comprise after the step c,, decode again according to the reference decoding algorithm of GB to the QR sign indicating number pattern after proofreading and correct.

Wherein, four summits seeking under the image coordinate in the summit of bounding box described in the step b further may further comprise the steps:

B1: obtain the view finding figure with reference to decoding algorithm according to GB;

B2: the distributed points that obtains four periphery sides by the scanning of DDA (digital differential analysis) algorithm;

B3: estimate four summits under the described image coordinate according to the protruding bounding box that obtains by the Sklansky algorithm.

Wherein, adopt SVD to estimate that the perspective transform parameter specifically comprises described in the step b, use image coordinate system down four summits and four summits under the corresponding world coordinate system, according to having an X-rayed parameter, form the linear equations group, adopt SVD to find the solution the value of perspective transform parameter.

A kind of system of having an X-rayed QR sign indicating number correcting pattern comprises: image collection module, image correction module;

Image collection module is used to obtain perspective QR sign indicating number pattern;

Image correction module is used for the protruding bounding box that basis is obtained by the Sklansky algorithm, estimate four summits under its image coordinate, and set up the corresponding relation of the point between image coordinate and the world coordinate system, and according to the perspective transform parameter that obtains, original image is carried out resampling, the QR sign indicating number pattern that obtains proofreading and correct;

Wherein, also comprise picture recognition module; Picture recognition module is used for the pattern after proofreading and correct is decoded the image after obtaining proofreading and correct again according to the reference decoding algorithm of GB.

The present invention has following beneficial effect:

The present invention is by adopting the Sklansky algorithm, accurately obtain the image coordinate on four summits of QR sign indicating number pattern, the parameter that adopts the method find the solution linear equation to estimate to have an X-rayed matrix then reaches correcting image, the purpose that is correctly decoded, thus the difficult problem of perspective QR sign indicating number pattern decoding solved.

Description of drawings

Fig. 1 is a system construction drawing of the present invention;

Fig. 2 A/2B/2C/2D obtains synoptic diagram for summit of the present invention;

Fig. 3 is the synoptic diagram after perspective pattern process is proofreaied and correct among Fig. 2.

Embodiment

The present invention is described in further detail below in conjunction with the drawings and specific embodiments:

See also the system shown in Figure 1 structural drawing, comprising: image collection module 101, image correction module 102, picture recognition module 103; Image collection module 101 can directly be obtained image by camera, also can obtain image by other means such as image files; Image correction module 102 is utilized the Sklansk algorithm to obtain protruding bounding box and is estimated four summits under its image coordinate, set up the corresponding relation of the point between image coordinate and the world coordinate system, and according to the perspective transform parameter that obtains, original image is carried out resampling, the QR sign indicating number pattern that obtains proofreading and correct; Picture recognition module 103 is used for the pattern after proofreading and correct is decoded the image after obtaining proofreading and correct again according to the reference algorithm of GB.

QR sign indicating number image after can obtaining proofreading and correct by image correction module among the present invention.Key step is as follows:

(1) at first, image correction module is utilized horizontal and vertical scanning according to the reference decoding algorithm that national standard proposes, and adds 45 degree cross scans, obtain three FP points (P1, P2, P3); Shown in Fig. 2 A;

(2) image correction module is starting point with P1, is direction with P3-＞P1 and P1-＞P3 respectively, scans one group of starting point set S1; From the S1 mid point, be direction with P2-＞P1, scan the component L1 that layouts; In like manner, obtain distributed points L2, L3 and L4; Scanning direction adopts the DDA method; Shown in Fig. 2 B;

(3) image correction module is according to point set L1, L2, and L3 and L4 ask for peripheral mask, the place to go background images; Shown in Fig. 2 C;

(4) image correction module adopts the Sklansky algorithm to ask for protruding bounding box, shown in Fig. 2 D; Obtain n order summit of protruding bounding box, n1, n2 ... np;

(5), find two the some nk and the nl of minimum distance if the distance between adjacent 2 is then asked in p＞4, line segment l1 according to consecutive point ni, nk determine according to the line segment l2 that consecutive point nl, nm determine, uses the intersection point of l1 and l2 to replace nk and nl point, up to only remaining four summit (x0, y0), (x1, y1), (x2, y2), (x3, y3) till;

(6) (u0, v0), (u1, v1), (u2, v2), (u3 is known v3), therefore, can use Matrix C owing to four rectangle summits in the world coordinates

$\left[\begin{array}{ccc}c00& c01& c02\\ c10& c11& c12\\ c20& c21& c22\end{array}\right]$

Set up the corresponding relation of the point between image coordinate system and the world coordinate system:

ui＝(c00*xi+c01*yi+c02)/(c20*xi+c21*yi+c22)

vi＝(c10*xi+c11*yi+c12)/(c20*xi+c21*yi+c22)

And obtain the linear equations group:

$\left[\begin{array}{cccccccc}x0& y0& 1& 0& 0& 0& -x0*u0& -y0*u0\\ x1& y1& 1& 0& 0& 0& -x1*u1& -y1*u1\\ x2& y2& 1& 0& 0& 0& -x2*u2& -y2*u2\\ x3& y3& 1& 0& 0& 0& -x3*u3& -y3*u2\\ 0& 0& 0& x0& y0& 1& -x0*v0& -y0*v0\\ 0& 0& 0& x1& y1& 1& -x1*v1& -y1*v1\\ 0& 0& 0& x2& y2& 1& -x2*v2& -y2*v2\\ 0& 0& 0& x3& y3& 1& -x3*v3& -y3*v3\end{array}\right]\·\left[\begin{array}{c}c00\\ c01\\ c02\\ c10\\ c11\\ c12\\ c20\\ c21\end{array}\right]=\begin{array}{}\end{array}\left[\begin{array}{c}u0\\ u1\\ u2\\ u3\\ v0\\ v1\\ v2\\ v3\end{array}\right]$

Wherein, cij is a matrix coefficient, c22=1;

(7) image correction module adopts SVD to separate above-mentioned linear equations, obtains separating of Matrix C;

(8) image correction module is carried out resampling according to the perspective transform parameters C that obtains to original image, the QR sign indicating number pattern that obtains proofreading and correct, as shown in Figure 3;

(9) picture recognition module is decoded according to the reference algorithm of GB, i.e. decodable code again to the pattern after proofreading and correct.

The above only is preferred embodiment of the present invention, not in order to restriction the present invention, all any modifications of being done within the spirit and principles in the present invention, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.

Claims (6)

1, a kind of method of having an X-rayed the 2 D code of quick response matrix correcting pattern is characterized in that, may further comprise the steps:

A: obtain perspective 2 D code of quick response matrix pattern;

B: ask for the protruding bounding box of 2 D code of quick response matrix pattern according to the Sklansky algorithm, in the summit of bounding box, seek four summits under the image coordinate; Adopt singular value decomposition method to estimate the perspective transform parameter;

C: according to the perspective transform parameter that obtains, original image is carried out resampling, the 2 D code of quick response matrix pattern that obtains proofreading and correct.

2, the method for perspective 2 D code of quick response matrix correcting pattern as claimed in claim 1 is characterized in that, also comprises after the step c, to the 2 D code of quick response matrix pattern after proofreading and correct, decodes again according to the reference decoding algorithm of GB.

3, the method for perspective 2 D code of quick response matrix correcting pattern as claimed in claim 1 or 2 is characterized in that, four summits seeking under the image coordinate in the summit of bounding box described in the step b further may further comprise the steps:

B1: obtain the view finding figure with reference to decoding algorithm according to GB;

B2: the distributed points that obtains four periphery sides by the scanning of digital differential analysis algorithm;

B3: estimate four summits under the described image coordinate according to the protruding bounding box that obtains by the Sklansky algorithm.

4, the method for perspective 2 D code of quick response matrix correcting pattern as claimed in claim 1 or 2, it is characterized in that, adopt singular value decomposition method to estimate that the perspective transform parameter specifically comprises described in the step b, use image coordinate system four summits and four summits of corresponding world coordinate under being down, according to the perspective parameter, form the linear equations group, adopt singular value decomposition method to find the solution the value of perspective transform parameter.

5, a kind of system of having an X-rayed the 2 D code of quick response matrix correcting pattern is characterized in that, comprising:

Image collection module, image correction module;

Image collection module is used to obtain perspective 2 D code of quick response matrix pattern;

Image correction module is used for the protruding bounding box that basis is obtained by the Sklansky algorithm, estimate four summits under its image coordinate, and set up the corresponding relation of the point between image coordinate and the world coordinate system, and according to the perspective transform parameter that obtains, original image is carried out resampling, the 2 D code of quick response matrix pattern that obtains proofreading and correct;

6, as the system of perspective 2 D code of quick response matrix correcting pattern as described in the claim 5, it is characterized in that, also comprise picture recognition module; Picture recognition module is used for the pattern after proofreading and correct is decoded the image after obtaining proofreading and correct again according to the reference decoding algorithm of GB.

Images