CN110766117B

CN110766117B - Two-dimensional code generation method and system

Info

Publication number: CN110766117B
Application number: CN201810845760.0A
Authority: CN
Inventors: 徐明亮; 吕培; 李亚飞; 周兵; 李翔
Original assignee: Zhengzhou University
Current assignee: Zhengzhou University
Priority date: 2018-07-27
Filing date: 2018-07-27
Publication date: 2022-12-13
Anticipated expiration: 2038-07-27
Also published as: CN110766117A

Abstract

The invention discloses a method and a system for generating a two-dimensional code. The method comprises the following steps: converting the background image into a background gray image; processing the background gray image by adopting a Gaussian difference operator method based on edge tangential flow to obtain a contour image; performing fusion processing on the contour image, the background gray image and the first two-dimensional code image to obtain a second two-dimensional code image; performing color quantization processing on the second two-dimensional code image by adopting an LAB uniform color space to obtain a third two-dimensional code image; and restoring the color of the third two-dimensional code image to obtain a fourth two-dimensional code image. According to the method, the generation process of the two-dimension code and the cartoon rendering process of the background image are combined into a whole, the key area of visual feedback in the picture is identified through feature recognition and saliency area extraction, so that the generation of the black and white module in the area is adjusted through a threshold setting method based on the generation process of the two-dimension code module, and the attractiveness of the final two-dimension code generation result is improved.

Description

Two-dimensional code generation method and system

Technical Field

The invention relates to the technical field of two-dimensional codes, in particular to a method and a system for generating a two-dimensional code.

Background

A typical two-dimensional code is a pattern of black and white that is regularly distributed on a plane by a specific geometric figure. With the more and more extensive and frequent application of the two-dimensional code, the two-dimensional code has no characteristic in appearance and is not beautiful enough.

Disclosure of Invention

The embodiment of the invention provides a method and a system for generating a two-dimensional code, which aim to solve the problems that the two-dimensional code in the prior art has no characteristics and is not beautiful enough.

In a first aspect, a method for generating a two-dimensional code is provided, including:

converting the background image into a background gray image;

processing the background gray image by adopting a Gaussian difference operator method based on edge tangential flow to obtain a contour image;

fusing the contour image, the background gray image and the first two-dimensional code image to obtain a second two-dimensional code image;

performing color quantization processing on the second two-dimensional code image by adopting an LAB uniform color space to obtain a third two-dimensional code image;

and restoring the color of the third two-dimensional code image to obtain a fourth two-dimensional code image.

In a second aspect, a two-dimensional code image generation system is provided, including:

the transformation module is used for transforming the background image into a background gray image;

the Gaussian difference module is used for processing the background gray level image by adopting a Gaussian difference operator method based on edge tangential flow to obtain a contour image;

the fusion module is used for carrying out fusion processing on the contour image, the background gray image and the first two-dimensional code image to obtain a second two-dimensional code image;

the color quantization module is used for performing color quantization processing on the second two-dimensional code image by adopting an LAB uniform color space to obtain a third two-dimensional code image;

and the color restoration module is used for restoring the color of the third two-dimensional code image to obtain a fourth two-dimensional code image.

According to the embodiment of the invention, the generation process of the two-dimensional code and the cartoon rendering process of the background image are combined into a whole, and the key area of visual feedback in the picture is identified through feature identification and saliency area extraction, so that the generation of the black and white module in the area is adjusted through a threshold setting method based on the generation process of the two-dimensional code module, and the attractiveness of the final two-dimensional code generation result is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the description of the embodiments of the present invention will be briefly introduced below, and it is obvious that the drawings in the description below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive labor.

Fig. 1 is a flowchart of a method for generating a two-dimensional code according to an embodiment of the present invention;

fig. 2 is an effect diagram of each step of a two-dimensional code generation process according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a process for processing an edge tangential flow image along an edge tangential flow by using a Gaussian difference operator method according to an embodiment of the present invention;

fig. 4 is a schematic effect diagram corresponding to the number of times of processing the edge tangential flow image along the edge tangential flow by using the gaussian difference operator method according to the embodiment of the present invention;

FIG. 5 is a schematic diagram of a bilateral filtering process according to an embodiment of the present invention;

FIG. 6 is a comparison graph of the color quantization effect of the embodiment of the present invention;

fig. 7 is a block diagram of a two-dimensional code generation system according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The embodiment of the invention discloses a method for generating a two-dimensional code. As shown in fig. 1, the method comprises the steps of:

step S101: and transforming the background image into a background gray image.

The background image is a cartoon image, for example, as shown in fig. 2 (a).

Specifically, the steps include the following processes:

(1) The color of the background image is converted from the RGB color space to the LAB color space.

The LAB color space is a color space required for the gradation and binarization processing, and therefore, the background image is first converted from the RGB color space to the LAB color space, and for example, the converted image is shown in fig. 2 (b).

(2) And taking a gray level layer from the background image converted into the LAB color space to obtain a background gray level image.

For example, the background grayscale image is shown in fig. 2 (c).

Step S102: and processing the background gray image by adopting a Gaussian difference operator method based on edge tangential flow to obtain a contour image.

Specifically, in a preferred embodiment of the present invention, the step includes the following steps:

(1) And constructing an edge tangential flow on the background gray level image to obtain an edge tangential flow image.

The definition of the edge tangential flow is as follows: and for the picture I (z) (z = (x, y)) representing any pixel point of the picture, constructing a smooth edge flow field, and storing the edge flow field of the feature region. Defining t (z) as an edge tangent as a vector perpendicular to the picture gradient direction g (z) = Δ I (z). Such a vector field is defined as Edge Tangential Flow (ETF).

The edge tangential flow can be constructed on the background gray image by filtering, for example, as shown in fig. 2 (e) - (g). In particular, at each pixel-centric kernel, a non-linearly smooth vector is applied, such that the dominant edge direction is preserved and weaker edges are redirected to their dominant regions. At the same time, to maintain a sharp inflection region, smoothing is performed along directionally similar edges. Edge tangential flow filtering is defined by the formula:

wherein t (z) represents a tangent vector at the z position of the pixel point, the direction of the tangent vector points to a local edge, and 2 pi is taken as a period. Omega _μ Representing the kernel radius mu at the picture pixel point z. Pixel point z ₁ And z ₂ At omega _μ Within the defined range, the edge flow fields are jointly determined. By selecting these two points, the gradient values can be compared. k represents a vector normalization term, which can be set empirically. In the embodiment of the invention, the value of k is 3.

Wherein, ω is _s Representing a spatial weight function.

Wherein, ω is _m Representing an amplitude weighting function.

And

respectively represented at pixel point z ₁ And z ₂ The normalized gradient magnitude of (a). Omega _m Is in the range of [0, 1]]And its value follows the amplitude difference

Is increased, indicating that more weight is given to the neighboring, gradient magnitude ratio center pixel z ₁ Higher pixel z ₂ Thereby ensuring protection of the leading edge direction.

ω _d (x, y) = | t (x) · t (y) |. Wherein, ω is _d The directional weighting function is expressed and plays a role of smoothing regions with similar directions. t (x) and t (y) represent normalized tangent vectors of the horizontal and vertical coordinates x and y, respectively. When the angle between the two vectors approaches 0, ω _d The value of (a) is increased; when the two vectors are close to vertical, ω _d The value of (c) is decreased. If the angle between the two vectors is greater than 90 degrees, the direction of the vectors t (x) and t (y) is rotated by applying a function, which is defined as follows:

through the above process, an edge tangential flow field is established.

(2) And processing the edge tangential flow image along the edge tangential flow by adopting a Gaussian difference operator method to obtain a first contour image.

Preferably, the edge tangential flow image may be normalized first, and the resulting image is shown in fig. 2 (g).

To obtain a continuous smooth, noise-free picture dominant contour curve, a first contour image is obtained along the edge tangential flow using the gaussian difference operator (FDoG) method. In the edge tangential flow field, t (z) represents the tangent vector at pixel point z, whose direction points to the local edge, i.e. there is a maximum difference in its vertical direction, i.e. the gradient direction. Along the edge tangential flow, applying a Gaussian difference operator method for filtering in the gradient direction.

By C _z (s) represents a rheological curve of the edge tangential flow at the pixel point z (namely, a flow axis at the pixel point z), represents the edge direction in the middle of the contour line, s is a certain point in the longitudinal arc length at the pixel point z position in the image I, and the value of s can be positive or negative, and different values of s indicate different points on the longitudinal arc length at the pixel point z position. Suppose pixel point z is located at the center of the curve, i.e. C _z (s) =0. By means of _z,s Represents the simultaneous presence of t (C) in a straight line _z (s)) and C _z (s) the intersecting part (i.e. the gradient axis at pixel point z). t (C) _z (s)) indicates the local edge direction of the rheological curve, representing the edge direction at both ends of the contour. Will l _z,s Expressed by a determined transverse arc length parameter t, hence l _z,s (t) represents a straight line l _z,s At t. While assuming bit l _z,s In C _z (s) center, i.e. /) _z,s (0)＝C _z (s). Note that _z,s Parallel to the gradient vector g (C) _z (s)). Fig. 3 is a schematic flow chart of filtering by applying the difference of gaussian operator method. The formula of filtering by the Gaussian difference operator method is as follows:

h (z) represents the pixel point processed by adopting a Gaussian difference operator method. I (l) _z,a (a) Denotes an input image I at l _z,a (a) A gradient value of the position. l. the _z,a (a) Represents a straight line simultaneously with t (C) _z (a) ) and C _z (a) The intersecting part (i.e. the gradient axis at pixel point z). Wherein, a represents a certain point in the arc length at the z position of the pixel point in the image I, and b represents a one-dimensional filtering weight vector function on the gradient line. The meaning of this formula is as follows: when along the flow axis C _z While moving, a one-dimensional filtering f (b) is applied on the gradient lines. The response of the single filter is then taken along the flow axis C using a point a in the arc length at the z position of the pixel point in the image I _x Are accumulated and represented as

Wherein G is _σ Represents variance of G _σ (b) A univariate gaussian function of (2).

In particular, the method comprises the following steps of,

σ determines the length of the flow kernel and also the degree of line coherence. Generally, σ has a size of 3.

For filtering f (b), an edge model based on the gaussian difference operator method is applied, whose formula is as follows:

a univariate gaussian function representing the central interval region.

A univariate gaussian function representing the surrounding spaced region. Parameter sigma _c And σ _s For controlling the size of the central space and the surrounding space. Both the center spacing and the surrounding spacing act on the edges of the face. The gaussian difference operator method constructs a coordinate axis around the image edge information, as shown in fig. 3, the position represented by 0 is located at the edge center (i.e. the center of the width of the contour line), the "interval" refers to the distance from the position represented by 0, and the center interval and the surrounding interval are both a normal distribution curve based on the coordinate axis, because the distribution of the center interval and the surrounding interval is different due to different parameters. The center interval can deepen the depth of the center of the edge; the peripheral spacing deepens both the depth of the center of the edge and the depth of the pixels around the center of the edge. The central and peripheral spaces together may delineate the edges. In general, set σ _c ＝1.6，σ _s =1.6. Rho is used for controlling the noise detection degree, and the value range of rho is [0.97,1.0 ]]. In order to minimize the generated picture noise and not to affect the decoding function after being fused with the two-dimensional code, the value of ρ is 0.97.

(3) And carrying out binarization processing on the first contour image to obtain a second contour image.

After the processed pixel points H (z) are obtained, the pixel points jointly form a picture H. Converting the picture H into a black-and-white image through binarization processing, wherein a formula of the binarization processing is as follows:

tau takes a value of [0,1]. Generally, the value is 0.5. Second contour image after binarization processing

A target line graph which is a contour, for example, as shown in fig. 2 (h).

(4) And superposing the second contour image and the background image to obtain a third contour image.

To further enhance the effect of contouring, gaussian difference operator method filtering iteration operations may be applied to the input picture. The iterative operation is to superimpose the second contour image on the background image to obtain a third contour image.

(5) And processing the third contour image along the edge tangential flow by adopting a Gaussian difference operator method to obtain a fourth contour image.

And (3) processing the third contour image by adopting the method in the step (2) to obtain a fourth contour image, which is not described herein again.

(6) And carrying out binarization processing on the fourth contour image to obtain a contour image.

And (4) processing the third contour image by adopting the method in the step (3) to obtain a final contour image. As shown in fig. 2 (i), it will not be described herein.

It should be understood that the number of iterations in the embodiment of the present invention is two, but the present invention is not limited thereto, and the above steps (4) to (6) may be repeated until a satisfactory contour image is obtained.

Generally, the more the iteration times, the more obvious the line is, and the more abundant the details are; however, if the number of iterations is too large, the lines are more densely distributed, and the picture will appear more "disorderly", thereby reducing the visual effect thereof. Fig. 4 shows the background image (a), the image after the first gaussian difference operator method processing (b) and the second gaussian difference operator method processing (c). It is obvious from the figure that the line effect is clearer when the iteration times are 2, the details are richer, and the texture contour is more obvious.

In addition, before each time of processing by using the gaussian difference operator method, the input image may be subjected to a gaussian filtering operation, so as to obtain a smoother and softer line effect, as shown in fig. 2 (d).

Further, after each processing by the gaussian difference operator method, the obtained image may be subjected to region smoothing processing. The goal of region smoothing is to remove unnecessary detail inside the region while preserving important parts, and therefore this is a feature preserving picture smoothing method, usually implemented with bilateral filtering. However, conventional bilateral filtering has some limitations that ignore the direction of color contrast when a circular region smoothes out the differences in secondary colors, thereby removing some tiny but meaningful shape boundaries, making the image appear edge-roughened. The bilateral filtering method based on the edge tangential flow can better overcome the problems. Embodiments of the present invention may utilize two separate linear bilateral smoothing operations, one along the edge direction and the other along the gradient direction.

Specifically, the following formula is used to describe the bilateral filtering process based on the edge tangential flow, and the linear bilateral filtering along the edge direction is defined as follows:

wherein, C _z Shows the rheological curve of the edge tangential flow, V _e A weight normalization parameter is represented by a weight normalization parameter,

the spatial weight function in the bilateral filtering has the same expression as the spatial weight function, and is not described herein again. Along the rheological curve C _z Direction, σ _e The kernel size of pixel point z is determined. h represents a similarity weight function for following the rheological curve C _z And when the direction is in, comparing the color difference of the pixel points of the on-axis point and the width center of the contour line. h (z, C) _z (a),σ)＝G _σ (||I(z)-I(C _z (a))||)。

Similarly, the bilateral filtering operation along the gradient direction is defined as follows:

l _z (b) Representing the gradient axis at the z-position of the pixel point. V _g A weight normalization parameter is represented by a weight normalization parameter,

representing a univariate gaussian function.

Step S103: and carrying out fusion processing on the contour image, the background gray image and the first two-dimensional code image to obtain a second two-dimensional code image.

Specifically, the steps include the following processes:

(1) And carrying out visual saliency area extraction on the background gray level image to obtain a visual saliency image.

The visually significant image is shown in fig. 2 (j).

Specifically, the process of extracting the visually significant region is as follows:

(1) the background grayscale image is divided into a plurality of sub-images.

If the side length of the background gray-scale image is n pixels and the side length of the sub-image is m pixels, the background gray-scale image is divided into n sub-images

Sub-images, and each sub-image has a size of m x m pixels.

(2) And carrying out binarization processing on each sub-image to obtain a visual saliency image.

Specifically, the calculation formula of the binarization processing is as follows:

wherein, mod G _r RepresentThe binarization effect of the r-th sub-image of the background gray level image determines whether the sub-image is filled with black or white. Each subimage is symbolized as subG _r . r denotes the subscript of the subimage, which ranges from 1 to

G _w (x, y) represents the pixel weight at point (x, y) of each sub-image, and

G _w representing a gaussian function.

(2) And carrying out binarization processing on the background gray level image to obtain a background binarization image.

The background binary image is shown in fig. 2 (k).

(3) And superposing the outline image, the visual saliency image and the first two-dimensional code image to obtain a fifth two-dimensional code image.

Wherein, the first two-dimensional code image is an original standard two-dimensional code image composed of black and white squares, as shown in fig. 2 (l); the fifth two-dimensional code image is shown in fig. 2 (m).

(4) And obtaining a sixth two-dimensional code image according to the binarization color value of the fifth two-dimensional code image and the binarization color value of the first two-dimensional code image.

Specifically, the steps include the following processes:

(1) and determining the binarization color value of each module of the fifth two-dimensional code image.

The value of each block of the fifth two-dimensional code image is calculated according to the following formula:

wherein, the set M represents the two-dimension code function part, the format information, the version information and the relevant module of the input data code.

And

respectively representing a first two-dimensional code image Q ^s And a fifth two-dimensional code image Q ⁱ The binarized color value of the r-th module. From the calculated values of the above formula, it can be determined whether the module is black or white, where 1 represents black and 0 represents white.

(2) And according to the binarization color value of each module of the fifth two-dimensional code image and the binarization color value of each module of the first two-dimensional code image, carrying out information resetting of bit streams on parts of the completion codes and the error correcting codes in each module of the fifth two-dimensional code image to obtain a sixth two-dimensional code image.

In particular, it can be seen from the above formula

And

may be different, lacking parts of the completion code and error correction code. Therefore, it is necessary to reset information of the bit stream B to maintain the bit streams B and B

And (5) the consistency is achieved. Since the change to the module is the same as the change to one byte (bit) in bitstream B, the values in bitstream B are modified for all bits of the padding and error correction codes according to the following rules: (1) If it is used

And

are the same, let B _L(r) Keeping the original shape; (2) If it is used

And

is calculated according to the following formula, L (r) represents the bit of the index bit stream B corresponding to the r-th module:

by this step, the values in the bit stream B are reset in preparation for the following steps.

It should be understood that the sixth two-dimensional code image after the fusion processing through the above steps is a beautified binarized two-dimensional code without RS (Reed-Solomon) coding added.

(5) And coding the sixth two-dimensional code image according to the Reed-Solomon coding rule to obtain a second two-dimensional code image.

Specifically, the steps include the following processes:

and selecting k RS code words from the c information codes of the sixth two-dimensional code image, calculating the values of the remaining (c-k) code words according to an RS coding rule, and determining the word number of the complementary codes of the sixth two-dimensional code image to obtain the second two-dimensional code.

c denotes the total number of data codewords and error correction codewords. k denotes the number of data codewords. S. the _c Representing the selected set of k RS codes. S. the _c Can be based on minimizing the second two-dimensional code image Q ^b And a fifth two-dimensional code image Q ⁱ The visual distortion of (a) results in:

is a measure of

And

a function of coherence.

Second two-dimensional code image Q ^b Representing a two-dimensional code that can be successfully decoded and requires beautified binarization to be generated, as shown in fig. 2 (n).

η (r) represents the visual importance value of the r-th module. The value of visual importance η (r) is used to assemble S _c The visual importance value η (r) is defined by introducing edge features and salient features. The specific visual importance value η (r) is defined as follows:

η(r)＝∑ _x,y λ ₁ ·subE _r (x,y)+λ ₂ ·subS _r (x,y)

λ ₁ and λ ₂ Represents a weighting coefficient, the value of which can be set empirically. In the embodiment of the invention, λ ₁ ＝0.65，λ ₂ ＝0.35。E _r And S _r Respectively representing an edge image and a saliency image of a picture. According to the method and the device, an opencv tool is utilized, and a sobel operator or a Laplacian operator in an opencv tool package is adopted to process to obtain the edge image. The Salient images were acquired with reference to the article effective medical Region Detection with Soft Image extraction by Cheng, M.M. et al (Cheng M, warrell J, lin W Y, et al. Effective medical Region Detection with Soft Image extraction [ C)]I/IEEE International Conference on Computer Vision. IEEE Computer Society,2013, 1529-1536).

The number of the filled codes can affect the aesthetic degree of the following two-dimensional code, the larger the number of the filled codes is, the larger the filled code field is, the larger the range of the beautifying operation for changing the picture is, so that the number of the k-d filled codes needs to be determined to determine the range which can be beautified by the subsequent steps. Since the data code word of the input d length has to be included in the set S _c In practice, only (k-d) codewords need to be selected from the remaining (n-d) codewords. Such a selection method is too computationally expensive for exhaustive enumeration. For example, when d =20,for a two-dimensional code of error correction level (9, L), the search space is

Wherein d is determined by the amount of data written in the two-dimensional code. Because of S _c Is a set of k code words, so there are k-d code word positions in addition to d data code words. Wherein n-d refers to code words at other positions of the picture, and k-d code words are selected from the code words and placed in the set S _c And (5) as a completion code. In addition, the embodiment of the invention can apply a greedy algorithm in the prior art and determine a global optimal solution through a local optimal solution, thereby improving the efficiency of problem solution and solving the problem of overlarge time complexity so as to select k visual saliency codewords. Code words each containing 8 bytes are recycled and each byte corresponds to a black and white module. The visual importance value for a codeword is calculated by accumulating the visual importance values of all bits. The codewords are then ordered according to the visual importance value. The first k codewords with the highest visual importance value are selected as S _c Among them.

Step S104: and carrying out color quantization processing on the second two-dimensional code image by adopting an LAB uniform color space to obtain a third two-dimensional code image.

Specifically, the steps include the following processes:

(1) And converting the second two-dimensional code image into an LAB color space.

(2) And performing region smoothing processing on the second two-dimensional code image converted into the LAB color space.

The region smoothing method is the region smoothing method described above, and is not repeated here, and the obtained image is shown in fig. 2 (o).

(3) And carrying out color quantization processing on the part corresponding to the filling codes of the second two-dimensional code image after the area smoothing processing to obtain a third two-dimensional code image.

In order to make the smoothed image have a cartoon effect, the image after the area smoothing is further processed by adopting a color quantization method. Color quantization is to integrate a plurality of visual colors in an image into a smaller number of colors and to regenerate an original image using the integrated colors. The integration process is based on the threshold setting criteria in the quantization formula. If the standard control is reasonable, the state of minimum quantization error is achieved, that is, the contrast between the generated image and the original image after quantization is small. Therefore, color quantization is a method used to reduce the number of colors in the original image while it is desirable to keep the visual effect of the original image unchanged. Processing in this way can make the image appear to have a painted effect, as shown in fig. 2 (p). Fig. 6 shows differences after color quantization of a picture, where (a) is an original image and (b) is a color quantized image.

The specific algorithm for color quantization may use the color quantization method proposed in the paper Real-time video interaction ([ 1] olsen S C, gooch B.real-time video interaction [ C ]// ACM SIGGRAPH.ACM, 2006. The calculation formula is as follows:

wherein, the first and the second end of the pipe are connected with each other,

representing the quantized image. Δ q represents the interval width between the pixel points around the target pixel point (i.e., the pixel point to be quantized) and the target pixel point.

And representing the position of the target pixel point. q. q.s _nt And the maximum value of the interval width closest to the target pixel point is represented.

Representing a parameter that controls sharpness.

For color quantized images, the transition sharpness is independent of the underlying image, possibly producing many visible transitions in smooth shaded areas. To minimize disharmony transitions, parameters are defined that control sharpness

As a function of the luminance gradient in the cartoonized image. The embodiment of the invention only allows the high interval boundary to appear in the place with high brightness gradient value, and when the brightness gradient value is higher than 1, the high interval boundary appears; when the brightness gradient value is lower than 1, the interval boundaries are dispersed over a large area. Thus, embodiments of the present invention provide a way to trade off color reduction variation against increased quantization by defining a target sharpness range and a gradient range. The scope of definition of the target of the embodiment of the invention is defined by

Gradient range is Λ _δ ,Ω _δ ]. Embodiments of the present invention limit the calculated gradient to the gradient range Λ _δ ,Ω _δ ]Then by mapping it linearly to

In the step (1), the first step,

the values of (b) are taken as parameters in linear mapping and can be obtained by computer induction according to the values of the gradient range, thereby ensuring better control of the image sharpness. In the present example, q ∈ [8,10 ]]，Λ _δ ＝0，Ω _δ ＝2，

A second advantage is the consistency for the color quantization of pictures. For standard quantization, a small brightness change can cause a large change in output, especially for the case where the input contains noise, which may enlarge the effective range of noise and affect the quality of the picture. By the color quantization mode based on the picture gradient, the color change generated in the output process can be softened in a relatively low contrast area, so that the visual effect generated by noise is not objectionable.

Step S105: and restoring the color of the third two-dimensional code image to obtain a fourth two-dimensional code image.

Specifically, the process is as follows:

(1) And carrying out color space reconstruction processing on the third two-dimensional code image.

The third two-dimensional code image is changed from black and white to a color defined by the LAB color space. The reconstructed image is shown in fig. 2 (q).

(2) And converting the color of the reconstructed third two-dimensional code image from an LAB color space to an RGB color space to obtain a fourth two-dimensional code image.

The fourth two-dimensional code image is shown in fig. 2 (r).

In summary, the two-dimensional code generation method provided by the embodiment of the invention combines the two-dimensional code generation process and the cartoon rendering process of the background image, and identifies the key area of visual feedback in the picture through feature recognition and saliency area extraction, so that the generation of the black and white module in the area is adjusted through a threshold setting method based on the two-dimensional code module generation process, and the attractiveness of the final two-dimensional code generation result is improved.

The embodiment of the invention also discloses a system for generating the two-dimensional code. As shown in fig. 7, the system includes the following modules:

and a transformation module 801, configured to transform the background image into a background grayscale image.

Preferably, the background image is a cartoon image.

And the gaussian difference module 802 is configured to process the background grayscale image by using a gaussian difference operator method based on edge tangential flow to obtain a contour image.

And the fusion module 803 is configured to perform fusion processing on the contour image, the background grayscale image, and the first two-dimensional code image to obtain a second two-dimensional code image.

And the color quantization module 804 is configured to perform color quantization processing on the second two-dimensional code image by using an LAB uniform color space to obtain a third two-dimensional code image.

And the color restoration module 805 is configured to restore the color of the third two-dimensional code image to obtain a fourth two-dimensional code image.

Preferably, the transformation module 801 comprises:

and the first conversion submodule is used for converting the color of the background image from an RGB color space to an LAB color space.

And the gray level taking submodule is used for taking a gray level from the background image converted into the LAB color space to obtain a background gray level image.

Preferably, the system further comprises:

and the Gaussian blur module is used for carrying out Gaussian blur processing on the background gray level image before the step of obtaining the contour image.

Preferably, the gaussian difference module 802 includes:

and the construction submodule is used for constructing the edge tangential flow on the background gray level image to obtain an edge tangential flow image.

And the first edge processing submodule is used for processing the edge tangential flow image along the edge tangential flow by adopting a Gaussian difference operator method to obtain a first contour image.

And the first binarization submodule is used for carrying out binarization processing on the first contour image to obtain a second contour image.

And the first superposition submodule is used for superposing the second contour image and the background image to obtain a third contour image.

And the second edge processing submodule is used for processing the third contour image along the edge tangential flow by adopting a Gaussian difference operator method to obtain a fourth contour image.

And the second binarization submodule is used for carrying out binarization processing on the fourth contour image to obtain a contour image.

Preferably, the system further comprises:

and the first region smoothing module is used for performing region smoothing processing on the contour image after the step of obtaining the contour image.

Preferably, the fusion module 803 includes:

and the extraction submodule is used for carrying out visual saliency region extraction on the background gray level image to obtain a visual saliency image.

And the third binarization submodule is used for carrying out binarization processing on the background gray level image to obtain a background binarization image.

And the second superposition submodule is used for superposing the outline image, the visual saliency image and the first two-dimensional code image to obtain a fifth two-dimensional code image.

And the fusion submodule is used for obtaining a sixth two-dimensional code image according to the binarization color value of the fifth two-dimensional code image and the binarization color value of the first two-dimensional code image.

And the coding submodule is used for coding the sixth two-dimensional code image according to the Reed-Solomon coding rule to obtain a second two-dimensional code image.

Preferably, the color quantization module 804 includes:

and the second conversion submodule is used for converting the second two-dimensional code image into an LAB color space.

And the second region smoothing module is used for performing region smoothing processing on the second two-dimensional code image converted into the LAB color space.

And the color quantization submodule is used for performing color quantization processing on the part corresponding to the filling code of the second two-dimensional code image after the region smoothing processing to obtain a third two-dimensional code image.

Preferably, the color restoration module 805 includes:

and the reconstruction submodule is used for carrying out color space reconstruction processing on the third two-dimensional code image.

And the third conversion submodule is used for converting the color of the reconstructed third two-dimensional code image from an LAB color space to an RGB color space to obtain a fourth two-dimensional code image.

In summary, the two-dimensional code generation system of the embodiment of the invention integrates the two-dimensional code generation process and the cartoon rendering process of the background image, and identifies the key area of visual feedback in the picture through feature recognition and saliency area extraction, so that the generation of the black and white module in the area is adjusted through a threshold setting method based on the two-dimensional code module generation process, and the attractiveness of the final two-dimensional code generation result is improved.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the technical solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one type of logical functional division, and other divisions may be realized in practice, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a U disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.

The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for generating a two-dimensional code is characterized by comprising the following steps:

converting the background image into a background gray image;

processing the background gray level image by adopting a Gaussian difference operator method based on edge tangential flow to obtain a contour image;

restoring the color of the third two-dimensional code image to obtain a fourth two-dimensional code image;

the step of obtaining the second two-dimensional code image includes:

performing visual saliency region extraction on the background gray level image to obtain a visual saliency image;

carrying out binarization processing on the background gray level image to obtain a background binarization image;

superposing the outline image, the visual saliency image and the first two-dimensional code image to obtain a fifth two-dimensional code image;

obtaining a sixth two-dimensional code image according to the binarization color value of the fifth two-dimensional code image and the binarization color value of the first two-dimensional code image;

encoding the sixth two-dimensional code image according to a Reed-Solomon encoding rule to obtain a second two-dimensional code image;

wherein, the step of obtaining the sixth two-dimensional code image includes:

determining a binarization color value of each module of the fifth two-dimensional code image;

and according to the binarization color value of each module of the fifth two-dimensional code image and the binarization color value of each module of the first two-dimensional code image, carrying out information resetting of a bit stream on parts of a complete code and an error correcting code in each module of the fifth two-dimensional code image to obtain a sixth two-dimensional code image.

2. The method of claim 1, wherein the step of transforming the background image into a background grayscale image comprises:

converting the color of the background image from an RGB color space to an LAB color space;

and taking a gray level layer from the background image converted into the LAB color space to obtain the background gray level image.

3. The method of claim 1, wherein the step of obtaining a contour image is preceded by the method further comprising:

and performing Gaussian blur processing on the background gray level image.

4. The method of claim 1, wherein the step of obtaining a contour image comprises:

constructing the edge tangential flow on the background gray level image to obtain an edge tangential flow image;

processing the edge tangential flow image along the edge tangential flow by adopting the Gaussian difference operator method to obtain a first contour image;

carrying out binarization processing on the first contour image to obtain a second contour image;

superposing the second contour image and the background image to obtain a third contour image;

processing the third contour image along the edge tangential flow by adopting the Gaussian difference operator method to obtain a fourth contour image;

and carrying out binarization processing on the fourth contour image to obtain the contour image.

5. The method of claim 1, wherein after the step of obtaining a contour image, the method further comprises:

and performing region smoothing processing on the contour image.

6. The method of claim 1, wherein the step of obtaining the third two-dimensional code image comprises:

converting the second two-dimensional code image into an LAB color space;

performing region smoothing processing on the second two-dimensional code image converted into the LAB color space;

and carrying out color quantization processing on the part corresponding to the filling code of the second two-dimensional code image after the area smoothing processing to obtain a third two-dimensional code image.

7. The method according to claim 1, wherein the step of obtaining the fourth two-dimensional code image comprises:

performing color space reconstruction processing on the third two-dimensional code image;

and converting the color of the reconstructed third two-dimensional code image from an LAB color space to an RGB color space to obtain a fourth two-dimensional code image.

8. The method of claim 1, wherein: the background image is a cartoon image.

9. A system for generating a two-dimensional code image, comprising:

the color restoration module is used for restoring the color of the third two-dimensional code image to obtain a fourth two-dimensional code image;

the fusion module includes:

the extraction submodule is used for carrying out visual saliency region extraction on the background gray level image to obtain a visual saliency image;

the third binarization submodule is used for carrying out binarization processing on the background gray level image to obtain a background binarization image;

the second superposition submodule is used for superposing the outline image, the visual saliency image and the first two-dimensional code image to obtain a fifth two-dimensional code image;

the fusion submodule is used for obtaining a sixth two-dimensional code image according to the binarization color value of the fifth two-dimensional code image and the binarization color value of the first two-dimensional code image;

the coding submodule is used for coding the sixth two-dimensional code image according to a Reed-Solomon coding rule to obtain the second two-dimensional code image;

wherein, the obtaining of the sixth two-dimensional code image includes: