CN111583085B

CN111583085B - Color image digital watermarking method based on transform domain

Info

Publication number: CN111583085B
Application number: CN202010300067.2A
Authority: CN
Inventors: 赵满坤; 白子玉; 应翔; 王建荣; 于健; 徐天一; 丁悦
Original assignee: Tianjin University
Current assignee: Tianjin University
Priority date: 2020-04-16
Filing date: 2020-04-16
Publication date: 2023-07-04
Anticipated expiration: 2040-04-16
Also published as: CN111583085A

Abstract

The invention discloses a color image digital watermarking method based on a transformation domain, which comprises the following steps: preprocessing spline images; watermark embedding process; watermark extraction process. Another method comprises the following steps: watermarking Arnold scrambling; watermark embedding process; watermark extraction process. The invention can effectively resist noise and object modification attack on the premise of ensuring that the visual quality of carrier data is not reduced and meeting the transparency of the watermark.

Description

Color image digital watermarking method based on transform domain

Technical Field

The invention relates to the field of digital watermarking, in particular to a color image digital watermarking method based on a transform domain.

Background

Vector graphics, also known as object-oriented images or pictorial images, use straight lines and curves to describe graphics, and are represented by geometric primitives such as points, lines, arcs, polygons, etc., based on mathematical equations. Vector graphics can be printed out with the highest resolution, and the creation of many industrial designs, billboard patterns and signs are all drawn by using vector graphics. Unlike common raster images, vector graphics have the following advantages: (1) Regardless of resolution, any scaling down or up of the vector image does not cause a change in image sharpness. (2) The vector map can smooth vector lines and color edges, keeping the lines in the same scale. (3) The vector diagram stores the line and color information, which is only relevant to the complexity, so that the storage space of the vector file is small. (4) vector data accuracy is high. Therefore, the digital vector diagram is the most important data of the geographic information system (Geographic Information System, GIS), and has been widely applied to the fields of navigation, cadastral management, digital cities, intelligent transportation and the like.

The ramp vector of the Slant transformation presents a uniform stepped down discrete staggered tooth waveform. The oblique transform is computationally simple and aims to match the base vector to a region of constant luminance slope, and the slot transform has been applied to many image processes such as transform coding and image restoration. In terms of encoding, the slot transform is considered to be a suboptimal orthogonal transform for energy compression. The sign of most intermediate frequency coefficients in the Slant transform remains unchanged before and after JPEG compression and Gaussian noise addition.

In recent years, wavelet transformation is very important in image processing, and is often used in image compression, feature detection and texture analysis. The wavelet transformation can perform multi-scale study on the local characteristics of the signal in the time domain and the frequency domain at the same time, so that the frequency domain characteristics of the local time domain can be inspected, and the time domain characteristics of the local frequency domain process can be inspected. Lifting wavelets based on conventional wavelet transforms were proposed by Swelden in 1996. Lifting wavelets find wide application in the field of signal processing: in the video field, the object with any shape can be adaptively encoded by using a lifting wavelet method, the encoding efficiency is obviously improved, and the subjective evaluation effect is better; in terms of still image processing, the lifting wavelet has better compression performance at low bit rates, and can provide efficient distortion and distortion-free compression in the same coding structure. The lifting method constructs wavelet transform in three steps: splitting, predicting and updating.

For images, the main features of the human visual system are generally represented in three aspects, namely, image brightness characteristics, frequency domain characteristics and image type characteristics. The brightness characteristic refers to the sensitivity of human eyes to brightness variation, and in general, the sensitivity of human eyes to noise in a high brightness area is smaller, so that the higher the background brightness of an image is, the more additional information can be embedded. For the frequency domain characteristics, the image is transformed from the spatial domain to the spectral domain, and the human eye is less sensitive to high frequency content. From the image type characteristics, the human visual system is more sensitive to smooth areas than areas with dense textures, so the denser the image textures, the more information that can be embedded.

Disclosure of Invention

The invention provides a color image digital watermarking method based on a transform domain, which can effectively resist noise and object modification attack on the premise of ensuring that the visual quality of carrier data is not reduced and meeting the transparency of watermarking, and is described in detail below:

a transform domain based color image digital watermarking method, the method comprising:

obtaining a transformation matrix of the embedded watermark and performing inverse Slant transformation to obtain LL of the embedded watermark information ^* Subband, LL ^* The sub-bands are combined with LH, HL and HH components to form a color image containing watermark information;

analyzing the color image along the inverse process of the color image generation to obtain the attributes of left_color and right_color elements, and writing the attributes into a vector diagram XML file;

performing discrete wavelet transformation on the color image to be detected and the initial color image to respectively obtain low-frequency subbands and performing Slant transformation to obtain a transformation matrix;

and ordering the zigzag of the two groups of transformation coefficients, extracting watermark information, and performing inverse DCT (discrete cosine transform) to obtain a watermark image.

The method for obtaining the transformation matrix embedded with the watermark comprises the following steps:

loading an XML file of the vector image, generating a corresponding DOM tree, and respectively obtaining B, R and G attribute values of a left_color element and a right_color element; the attribute values of B, R and G are quantized to form a two-dimensional array through taking values;

performing DCT on the watermark image once to generate a transformation matrix, and distributing different frequency parts at different positions of the transformation matrix after transformation; performing discrete wavelet transform on the color image C (x, y) to obtain a low-frequency subband LL;

performing Slant transformation on the low-frequency sub-band LL to obtain a transformation matrix LL_ST (u, v); for the above obtained transform coefficients LL_ST and W _dct Ordering according to zigzag, and adding W according to the principle of addition _dct Is embedded in the low frequency position corresponding to ll_st (u, v) to generate a watermark-embedded transform matrix.

Further, the addition principle is specifically as follows:

wherein m is _i Representing a watermark image W _dct Coefficient of x _i Coefficients representing LL _ ST (u, v) after LL subband slot transform,

representing post watermark embedding LL_ST ^* Coefficients of (u, v), and c is an embedding strength factor.

The two groups of transformation coefficients are arranged in a zigzag order, and watermark information is extracted specifically as follows:

combining the scrambled three-color components to obtain an Arnold scrambled watermark image, performing Arnold encryption and three-color base separation operation, and performing Slant transformation on the three-color base components to obtain transformation coefficients;

performing two-stage lifting wavelet transform on the original image, performing LH (liquid-liquid) on the original image ₂ Carrying out three-color base separation on the sub-bands and carrying out Slant transformation respectively to obtain respective transformation matrixes;

the watermark embedding operation is carried out by adopting different embedding intensities for different color components in combination with a human visual system;

performing inverse Slant transformation on the three-color base after the watermark is embedded, combining three-color base components, and generating an image containing watermark information through lifting wavelet inverse transformation;

respectively carrying out two-stage lifting wavelet transformation on an original image to be detected, and carrying out LH (liquid crystal display) on the obtained image ₂ The three-color base separation of the components is respectively carried out Slant transformation to obtain a transformation matrix;

and extracting the watermark, performing inverse Slant transformation on the extracted watermark information once, combining the three-color base components, and extracting a watermark image through Arnold decryption operation.

The watermark embedding operation is specifically performed by combining the human visual system to adopt different embedding intensities for different color components:

S_I ^* _R(G,B) (i,j)＝S_I _R(G,B) (i,j)+α _R(G,B) ×S_W _R(G,B) (i,j)

wherein alpha is _R(G,B) Is embedded strength, S_W _R(G,B) (I, j) is a transform coefficient, S_I _R(G,B) (I, j) is the respective transform matrix, S_I ^* _R(G,B) (i, j) is a transform matrix after embedding the watermark.

Further, the extracting watermark specifically includes:

wherein S_W _R(G,B) ^* (i, j) is the extracted watermark information.

The technical scheme provided by the invention has the beneficial effects that:

1. the invention can effectively ensure the invisibility of the watermark and has better robustness to common noise attack, JPEG compression, scaling and shearing attack;

2. the invention can utilize the visual masking phenomenon to meet the invisible characteristic of watermark perception; watermark information is widely distributed in the frequency spectrum domain of the whole image, so that the concealment is guaranteed; the main advantages of strong resistance to attack, etc.

Drawings

FIG. 1 is a flow chart of a spline image digital watermarking method based on DCT, DWT and Slant transformation;

fig. 2 is a flow chart of a dual color image digital watermarking method based on LWT and slot transforms.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in further detail below.

Example 1

1. Spline image digital watermarking method based on DCT, DWT and Slant transformation

Wherein, DCT: a discrete cosine transform (Discrete Cosine Trans form); DWT: discrete wavelet transform (Continuous Wavelet Transform); slant: and (5) oblique transformation.

A1: preprocessing spline images;

b1: watermark embedding process;

c1: watermark extraction process.

In one embodiment, step A1 performs preprocessing on the spline image, specifically as follows:

"left color" represents the color value on the left side of the spline curve, and "right color" represents the color value on the right side of the spline curve, and the attribute information such as B, R, G and the like represents the color value of the cubic bezier spline curve.

B, R and G attribute values of the left_color element and the right_color element are obtained, and the B, R and G attribute values form a two-dimensional array f (x, y) through value taking and quantization. Since the GRB color values are of integer type, (x, y) are discrete coordinates and f is a discrete magnitude, they can be regarded as pixel values of a gray-scale image, which is referred to as color image C in this method. This can be regarded as the pixel value of a gray-scale image, which is referred to as color image C in the present invention.

In one embodiment, step B1 performs watermark embedding based on step A1, and the specific steps are as follows:

the watermark image W (i, j) is subjected to DCT transformation once to generate a transformation matrix W _dct Different frequency parts are distributed at different positions of the transformation matrix after transformation. The color image C (x, y) is subjected to discrete wavelet transform to obtain a low-frequency subband LL. Then, the low-frequency subband LL is subjected to a slot transform to obtain a transform matrix ll_st (u, v). For the transformation matrix LL_ST (u, v) and transformation matrix W obtained above _dct Ordering according to zigzag, and adding W according to the principle of addition _dct Is embedded in the low frequency position corresponding to LL_ST (u, v) to generate a watermark-embedded transformation matrix LL_ST ^* (u, v). For LL_ST ^* (u, v) execution ofInverse Slant transformation to obtain LL embedded with watermark information ^* A subband. LL (light-emitting diode) ^* The subband is combined with LH, HL, HH (where LH, HL, and HH each represent detailed information such as image edge texture, as is well known to those skilled in the art) components to form a color image C containing watermark information ^* (x, y). Along the inverse of the color image generation, the color image C is parsed ^* (x, y) to obtain the attributes of the other left_color and the other right_color elements, writing into the vector diagram XML file. Accordingly, a vector image with watermark information is successfully obtained.

In one embodiment, step C1 performs watermark extraction based on step A1 and step B1, and the specific steps are as follows:

set the color image C to be detected ^* ' (x, y), initial color image C (x, y), the extracted watermark image is W ^* (i, j). Image C of color ^* ' discrete wavelet transform is performed on (x, y) and the initial color image C (x, y) to obtain low-frequency subbands LL respectively ^* ' and LL. Followed by a low frequency subband LL ^* ' and LL are respectively subjected to Slant transformation to obtain a transformation matrix LL_ST ^* ' (u, v) and ll_st (u, v). Then the two groups of transformation coefficients are arranged in a zigzag order, watermark information is extracted (the transformation coefficients are transformation matrix LL_ST ^* Coefficients in' (u, v) and ll_st (u, v). The finally extracted watermark information is subjected to inverse DCT transformation, and the watermark image W ^* (i, j) is successfully obtained.

Algorithm: pseudo code description of DDS watermark extraction algorithm

Input: color image C to be detected ^* ' (x, y), initial color image C (x, y), embedding intensity alpha

And (3) outputting: color image W containing watermark information ^* (i,j)

1.[LL ^* ′,LH ^* ′,HL ^* ′,HH ^* ′]＝dwt2(C, ^* 'haar')% wavelet decomposition

2.[LL,LH,HL,HH]＝dwt2(C,′haar′)

3. Generating Slant identity matrix S_n and identity inverse matrix S_nT

4.LL_ST ^* ’＝S_n×LL ^* ’×S_nT％Slant transformation

5.LL_ST＝S_n×LL×S_nT

6. Low frequency coefficient packing_locations zigzag arrangement

7.for location in embedding_locations:

8.W _dct (location)＝(LL_ST ^* ’(location)-LL_ST(location))/α

9.end for

10.W ^* ＝idct(W _dct )

11. Outputting a color image W containing watermark information ^*

2. Double-color image digital watermarking method based on LWT and Slant transformation

Wherein, LWT: lifting wavelet transform (Lifting Wavelet Transform)

A2: watermarking Arnold scrambling;

b2: watermark embedding process;

c2: watermark extraction process.

In one embodiment, step A2 performs watermarking Arnold scrambling as follows:

image scrambling is an effective means of secure storage and secure transmission of digital images, and is often used as an encryption method to preprocess watermark information to improve security. Scrambling algorithms for digital images can be divided into frequency domain scrambling and spatial domain scrambling, where the spatial domain includes a position space and a color space. In order for the original image to be accurately restored, a one-to-one correspondence between the host image and the transformed image must be ensured. Common scrambling methods include Arnold transformation, hillbert curve scrambling, torus self-isomorphic mapping transformation, magic square transformation, gray code scrambling methods, and the like. The Arnold transformation is used herein. The Arnold transformation is also called Cat face transformation (Cat Mapping), and is encrypted by scrambling the positions of pixels in the image. For an N-order digital image, the Arnold transform is performed using the following formula:

wherein x is _n ，y _n Representing image pixel values before Arnold transformation，x _n+1 ，y _n+1 The pixel values after Arnold transformation are represented, a and b are parameters, N represents the number of transformation times, mod is a modulo operation, and N is the image width.

The digital image was restored by an Arnold inverse transform, which was as follows:

/>

determining a scrambling key ^[1] An integer K is arbitrarily taken in the period range and is set as Arnold scrambling times. And then separating the three-color base of the watermark image, and then performing Arnold transformation operation on the three-color base matrix for K times respectively to obtain scrambled three-color components. Finally, the three scrambled color components are combined to obtain an Arnold scrambled watermark image W ^* 。

In one embodiment, step B2 performs watermark embedding on the basis of step A1, and the specific steps are as follows:

let the original image be I, the watermark image be W, the image containing the watermark be I x. Firstly Arnold encryption pretreatment is carried out on a watermark image W, and then three-color base separation operation is carried out on the encrypted image to obtain a three-color base component W _R ，W _G ，W _B Slant transformation is respectively carried out on the three-color base components to obtain transformation coefficients S_W _R(G,B) (i, j). Then, performing two-stage lifting wavelet transformation on the original image I, performing three-color base separation on an LH2 sub-band (wherein LH2 represents a frequency domain sub-band) to obtain corresponding components IR, IG and IB, and performing Slant transformation to obtain respective transformation matrixes S_I _R(G,B) (i, j) and then performing a watermark embedding operation. Finally, performing inverse Slant transformation on the three-color base after watermark embedding, combining three-color base components, and generating an image I containing watermark information through lifting wavelet inverse transformation ^* (x,y)。

In one embodiment, step C2 performs watermark extraction based on step A2 and step B2, and the specific steps are as follows:

first, an image I to be detected ^* ' and the original image I are subjected to two-stage lifting wavelet transform, respectively. Secondly, separating the three-color bases of the obtained LH2 component, and respectively performing Slant transformation to obtain a transformation matrix S_I ^* ′ _R(G,B) (I, j) and S_I _R(G,B) (i, j) (i.e., after the three components are subjected to a slot transform, two matrices are obtained). The watermark is then extracted. Finally, the extracted watermark information S_W _R(G,B) ^* (i, j) performing an inverse Slant transform, then combining the tri-color components, and finally decrypting the watermark image W by Arnold ^* Is successfully extracted.

Example 2

The invention provides two digital watermarking algorithms, which are a specific network architecture diagram of a product recommendation method of the invention, comprising the following steps:

1. spline image digital watermarking algorithm based on DCT and DWT and Slant transformation

Step S0101: loading an XML file of the vector image, generating a corresponding DOM tree (DOM is an abbreviation of Document Object Model (document object model)) and respectively obtaining B, R and G attribute values of a left_color element and a right_color element;

step S0102: the attribute values of B, R and G form a two-dimensional array f (x, y) through value taking and quantization;

since the GRB color values are all integer types, (x, y) are discrete coordinates, and f is a discrete magnitude, they can be considered as pixel values for a gray-scale image, referred to herein as color image C.

Step S0201: performing DCT on the watermark image W (i, j) once to generate a transformation matrix W _dct Different frequency parts are distributed at different positions of the transformation matrix after transformation;

step S0202: performing discrete wavelet transform on the color image C (x, y) to obtain a low-frequency subband LL;

step S0203: performing Slant transformation on the low-frequency sub-band LL to obtain a transformation matrix LL_ST (u, v);

step S0204: for the above obtained transform coefficients LL_ST and W _dct Ordered according to a "zig-zag" manner, according to an addition principle as shown in formula (3) Shown, W is _dct Is embedded in the low frequency position corresponding to LL_ST (u, v) to generate a watermark-embedded transformation matrix LL_ST ^* (u,v)。

representing post watermark embedding LL_ST ^* Coefficients of (u, v), oc is an embedded strength factor, for balancing visual quality and robustness of the algorithm.

Step S0205: for LL_ST ^* (u, v) performing an inverse Slant transform to obtain LL embedded watermark information ^* A subband. LL (light-emitting diode) ^* The sub-bands are combined with LH, HL, HH components to form a color image C containing watermark information ^* (x,y)。

Step S0206: along the inverse of the color image generation, the color image is parsed to obtain the attributes of the left_color and right_color elements, which are written into the vector image XML file. Accordingly, a vector image with watermark information is successfully obtained.

Set the color image C to be detected ^* ' (x, y), initial color image C (x, y), the extracted watermark image is W ^* (i,j)。

Step S0301: image C of color ^* ' discrete wavelet transform is carried out on (x, y) and C (x, y) to obtain low-frequency sub-band LL respectively ^* ' and LL.

Step S0302: low frequency part LL ^* ' and LL are respectively subjected to Slant transformation to obtain a transformation matrix LL_ST ^* ' (u, v) and ll_st (u, v).

Step S0303: ordering the two groups of transformation coefficients in a zigzag manner, and extracting watermark information according to a formula (4)

Wherein, the liquid crystal display device comprises a liquid crystal display device,

representing an image LL_ST to be detected ^* Coefficients of' (u, v), x _i LL_ST (u, v) coefficients, m ', representing the initial color image' _i Representing the extracted watermark information.

Step S0304: the extracted watermark information is subjected to inverse DCT transformation, and the watermark image W ^* (i, j) is successfully obtained.

2. Double-color image digital watermarking algorithm based on LWT and Slant transformation

Step S0101: and determining a scrambling key, and randomly taking an integer K in a period range to set Arnold scrambling times.

Step S0102: the watermark image is separated in three color bases. And (3) respectively separating R, G and B components of the watermark image I to obtain a trichromatic base matrix.

Step S0103: k Arnold transformation operations are respectively carried out on the three-color base matrix, K is a scrambling key, and the scrambled three-color components are obtained.

Step S0104: and combining the scrambled three-color components to obtain an Arnold scrambled watermark image W.

Step S0201: arnold encryption pretreatment is carried out on the watermark image W, and then three-color base separation operation is carried out on the encrypted image to obtain a component W _R ，W _G ，W _B Slant transformation is respectively carried out on the three-color base components to obtain transformation coefficients S_W _R(G,B) (i,j)。

Step S0202: performing two-stage lifting wavelet transform on the original image I, performing LH (left-right) on the original image I ₂ Sub-bands are subjected to three-color base separation to obtain corresponding components I _R ，I _G ，I _B And performing Slant transformation to obtain respective transformation matrixes S_I _R(G,B) (i,j)。

Step S0203: and (5) combining the human visual system to perform watermark embedding operation according to a formula (5) by adopting different embedding intensities for different color components.

S_I ^* _R(G,B) (i,j)＝S_I _R(G,B) (i,j)+α _R(G,B) ×S_W _R(G,B) (i,j) (5)

Wherein alpha is _R(G,B) Is embedding strength for balancing visual quality and algorithm robustness, in this experiment, let alpha be _R ＝0.06，α _G ＝0.03，α _B ＝0.12。

Step S0204: performing inverse Slant transformation on the three-color base after watermark embedding, combining three-color base components, and generating an image I containing watermark information by lifting wavelet inverse transformation ^* (x,y)。

Step S0301: image I to be detected ^* ' and the original image I are subjected to two-stage lifting wavelet transform, respectively.

Step S0302: for LH obtained ₂ Component three-color base separation is carried out, and Slant transformation is carried out respectively to obtain a transformation matrix S_I ^* ′ _R(G,B) (I, j) and S_I _R(G,B) (i,j)。

Step S0303: extracting the watermark according to equation (6)

Wherein S_W _R(G,B) ^* (i, j) embedding strength alpha for the extracted watermark information _R(G,B) The process is carried out according to the ratio of R to G to B=2 to 1 to 4.

Step S0304: watermark information S_W to be extracted _R(G,B) ^* (i, j) performing an inverse Slant transform, then combining the tri-color components, and finally decrypting the watermark image W by Arnold ^* Is successfully extracted.

Example 3

The schemes in examples 1 and 2 were validated in conjunction with specific examples, as described in detail below:

with the development of internet and communication technology, the informatization of media resources is becoming more and more popular, and the problem of digital information security becomes a focus of attention when people acquire resources conveniently, so that the digital watermark technology with copyright protection, authenticity and integrity authentication functions is rapidly developed in the background. The method is based on the digital watermark theory, and the digital watermark algorithm of the transform domain is studied in an important way, and mainly works as follows:

(1) Through research and analysis of color sample bar graphs of XML files, a digital watermarking scheme (DDS algorithm) based on the combination of DCT and DWT with Slant transformation is provided. According to the scheme, color information in an XML file is captured through preprocessing a spline graph to form a color matrix, and DCT low-frequency information of a watermark image is embedded into a Slant low-frequency coefficient through carrying out Slant transformation on LL subband components decomposed by discrete wavelets. Through experimental simulation, the method can effectively resist noise and object modification attack on the premise of ensuring that the visual quality of carrier data is not reduced and the transparency of the watermark is met. By analyzing the embedded capacity, the watermark scheme is proved to fully balance the conflict requirements of fidelity and capacity while guaranteeing robustness and safety.

(2) Aiming at the characteristic that spline graphs can be stored into raster graphs in a PNG format, the method provides a double-color digital watermarking algorithm (LS algorithm) based on LWT and Slant transformation. The watermark image Arnold is encrypted through the image scrambling idea, and the embedding strength of the three-color base is adjusted by combining different perception degrees of different colors by the HVS. Simulation experiments show that the method can effectively ensure invisibility of the watermark, and has better robustness to common noise attack, JPEG compression, scaling and shearing attack.

Spline image digital watermarking algorithm based on DCT and DWT and Slant transformation designs digital watermarking algorithm aiming at the problem of copyright protection of the spline image. In the preprocessing stage, color values are obtained from the spline image XML file to form a two-dimensional matrix, so that an XML file object which is difficult to operate is converted into a gray image which is easy to operate, and the alternative watermarking method is more diversified. The watermark information is selected to be embedded in the frequency domain range in consideration of limited space domain watermark embedding capacity and poor robustness against attack. The watermark embedding and extraction process is described in detail later. In order to verify the algorithm performance, experiments are carried out on different spline color maps, and the result shows that the watermark algorithm meets the requirement of perception invisibility under the condition of no attack. Secondly, on the premise of embedding 628 low-frequency coefficients, the image added with the watermark is attacked, and experimental results show that the image is faced with noise attacks and object modification attacks with different intensities,

the watermark image can still be successfully extracted, and the robustness of the proposed watermark algorithm is proved to be better. Through evaluation analysis of embedded capacity, experiments show that the proposed scheme fully balances the conflict requirements of fidelity and capacity while ensuring robustness and safety. The robustness of the watermark algorithm proposed in this chapter is proved to be better by comparison with NC values of other algorithms. In summary, the proposed digital watermarking scheme has positive significance for solving the copyright protection problem of the color sample bar graph.

The double-color image digital watermarking algorithm (LS algorithm) based on the combination of lifting wavelet transformation and oblique transformation adopts a meaningful color image as a digital watermark, and Arnold scrambling pretreatment is carried out on the digital watermark before embedding operation so as to improve the invisibility and the safety of watermark information. When the watermark is embedded, the original carrier image is subjected to lifting wavelet transformation, then the separated three-color bases are respectively subjected to Slant frequency domain transformation, and the embedding strength is adjusted according to different perception degrees of the HVS on different color components so that the PSNR value of the processed image is higher. In order to more intuitively verify the invisibility and the robustness of the chapter algorithm, the invention compares the experimental result with the DCT-based color image scrambling digital watermarking algorithm of Wang Taiyue and Li Hongwei. The results of the comparison of the two algorithms are shown in table 1.

Table 1 comparison of the method with the results of the literature DCT algorithm experiments

As can be seen from table 1, the PSNR of the method is higher than that of the document DCT algorithm, which indicates that the similarity of the carrier image is higher in the method compared with the original image after the watermark is added, so that the idea of adjusting the embedding strength for different color components according to the HVS is proved to be scientific and effective. When no attack is added and noise with different intensities is gradually added, the NC value of the method is higher than that of the DCT algorithm of the literature when JPEG compression attack is performed, which proves that the method has stronger robustness.

Reference to the literature

[1] Digital watermarking algorithm based on discrete cosine transform [ J ]. University of Zhengzhou journal (physico-edition), 2005 (03): 63-65.

Those skilled in the art will appreciate that the drawings are schematic representations of only one preferred embodiment, and that the above-described embodiment numbers are merely for illustration purposes and do not represent advantages or disadvantages of the embodiments.

The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.

Claims

1. A transform domain based color image digital watermarking method, the method comprising:

obtaining a transformation matrix of the embedded watermark and performing inverse Slant transformation to obtain LL of the embedded watermark information ^* Subband, LL ^* The sub-bands are combined with LH, HL, HH components to form a color image C containing watermark information ^* (x,y)；

Along color image C ^* Inverse process resolution color image C generated by (x, y) ^* (x, y) to obtain the attributes of left_color and right_color elements, writing into a vector diagram XML file;

to be detected color image C ^*′ Performing discrete wavelet transformation on the (x, y) and the initial color image C (x, y) to respectively obtain low-frequency subbands and performing Slant transformation to obtain a transformation matrix;

ordering the two groups of transformation coefficients in a zigzag manner, extracting watermark information, and performing inverse DCT (discrete cosine transform) to obtain a watermark image W ^* (i,j)。

2. The method for obtaining a digital watermark of a color image based on a transform domain according to claim 1, wherein the transform matrix for obtaining the embedded watermark comprises:

performing DCT on the watermark image once to generate a transformation matrix W _dct Different frequency parts are distributed at different positions of the transformation matrix after transformation; performing discrete wavelet transform on the color image C (x, y) to obtain a low-frequency subband LL;

3. The transform domain based color image digital watermarking method according to claim 2, wherein the addition principle is specifically:

4. The transform domain based color image digital watermarking method according to claim 2, wherein the step of sorting the two sets of transform coefficients in a zigzag order is specifically to extract watermark information:

5. a transform domain based color image digital watermarking method, the method comprising:

separating R, G and B components of the watermark image I to obtain a trichromatic base matrix, and performing K Arnold transformation operations on the trichromatic base matrix respectively, wherein K is a scrambling key to obtain scrambled trichromatic components;

6. The transform domain based color image digital watermarking method according to claim 5, wherein the watermark embedding operation is specifically performed by combining the human visual system to use different embedding intensities for different color components:

S_I ^* _R(G,B) (i,j)＝S_I _R(G,B) (i,j)+α _R(G,B) ×S_W _R(G,B) (i,j)

7. The transform domain based color image digital watermarking method according to claim 5, wherein the extracting watermark is specifically:

is the extracted watermark information.