A multi-color pattern quantum image blind watermarking method based on TMQIR

Abstract

The present invention proposes a multi-color mode quantum image blind watermarking method based on a multi-mode quantum image representation (TMQIR) in a three-dimensional coordinate system, which includes a quantum watermark embedding method, a watermark extraction method and a corresponding quantum circuit design. The two-dimensional code (QR code) that plays an active role in scan code payment, third-party authentication or remote login authentication in classical applications is introduced into the quantum world. Using a binary QR code image as a watermark, an arbitrary-sized TMQIR multi-color mode quantum carrier image is embedded based on the least significant bit (LSB). In TMQIR, a variety of images such as 24-bit RGB true color images and 8-bit grayscale images can be represented. From the perspective of the quantum image representation model, it can support images of arbitrary sizes in multiple color modes and is more universal; from the perspective of the watermarking method, it has more advantages in terms of complexity and key space; and the quantum watermarking method after the introduction of the QR code will also expand the application prospects of watermarks in quantum networks.

Description

Technical Field

[0001] The invention relates to a multi-color mode quantum image blind watermark method based on TMQIR, including watermark embedding, watermark extraction and quantum circuit, and belongs to the field of quantum cryptography and information security. Background Art

[0002] Image processing technology is already quite mature and widely used. Due to different operating mechanisms, traditional image processing algorithms cannot be transplanted to quantum computing platforms. With the vigorous development of quantum computing and quantum communication technology, the storage, description, and processing of quantum images have become a research hotspot and have made certain progress. The best way to protect quantum image copyright and locate tampering is quantum image watermarking technology, which embeds quantum watermark information into quantum carrier images to achieve information hiding without affecting the visual effect of the carrier image. By extracting the watermark hidden in the quantum image, the purpose of quantum image copyright protection and tampering location can be achieved.

[0003] In quantum computers, the quantum image representation model is based on a certain number of quantum bits and their entanglement, which determines the practicality, universality and processing efficiency of quantum images. Images of any size H×W in multiple color modes can be represented in quantum computers, making TMQIR more universal. In TMQIR, the standard orthogonal quantum states |0> and |1> sequences represent the position and color of the image, respectively. In an H×W image, the Y coordinate and X coordinate need to be and Therefore, a TMQIR image can be expressed as follows:

[0004]

[0005] where |ZYX>=|Z>|Y>|X>=|z l-1 z l-2 …z 0 >|y h-1 y h-2 …y 0 >|x w-1 x w-2 …x 0 >, z k ,y i , x j ∈{0,1}, i∈{0,1,…,h-1}, j∈{0,1,…,w-1}, k∈{0,1,2}, When θ=π,φ∈{0,π / 2,π,3π / 2},|I> represents images in binary, grayscale, RGB, LAB and other color models. When |I) represents a binary image, φ=0, and |z is not required l-1z l-2 …z 0 >, that is, l = 0, the color value |C at the pixel position |YX> YX >Only 1 qubit is required to represent it, i.e. l c =1; when |I> represents a grayscale image, φ = π / 2, |z l-1 z l-2 …z 0 >Represented by 3 qubits, i.e. l ​​= 3, color value |C ZYX >Only 1 qubit is required to represent it, i.e. l c =1; when |I> represents RGB and LAB color model diagrams, l=3, That is l c =3, representing the color values ​​of the three color channels in color mode.

[0006] In the application process of quantum image watermarking technology, we often need to extract the watermark. If the detection process does not require the assistance of the original image, the watermark method is a blind watermark, which is more convenient and practical in application. In recent years, some literatures have proposed some quantum watermarking schemes based on the FRQI model or the NEQR model, but there is no quantum watermarking scheme based on the more universal TMQIR. In addition, quantum watermarking based on the least significant bit (LSB) is considered to be simple and effective.

[0007] Two-dimensional code (QR code) has a matrix of data bits and error correction codewords, a separation pattern, a positioning pattern, a position pattern, and an error correction pattern. QR code has advantages in information capacity, fault tolerance, and recognition speed, and can also encode digital information such as images, sounds, texts, signatures, fingerprints, etc. The watermark information encoding method based on QR code can effectively improve the watermark information capacity, enhance the robustness of the watermark algorithm, and facilitate watermark detection. In classic applications, QR codes play a good role as online and offline channels, while also combining the privacy and publicity of information. In the application of scanning code payment and login, QR codes play an active role in third-party authentication and remote authentication. Inspired by this, QR codes are introduced into quantum images. Combined with the quantum entanglement characteristics, it will bring new application environments and research prospects for quantum image watermarking. Summary of the invention

[0008] The purpose of the present invention is to propose a more universal quantum image blind watermarking method based on TMQIR. The quantum carrier image in the present invention supports multiple color modes and arbitrary sizes, which is more universal; the watermark image is a two-dimensional code with strong fault tolerance, high information capacity and fast recognition speed, which takes into account both information privacy and publicity, and combines the characteristics of quantum entanglement to better promote the application of quantum watermarks in quantum networks; the watermarking method used in the present invention has more advantages in key space and quantum circuit implementation complexity. The present invention further improves the performance of quantum information hiding and ensures the security of quantum cryptography.

[0009] The technical solution of the present invention is as follows:

[0010] Based on the TMQIR model, the Arnold scrambling and LSB steganography techniques are used to convert a 2 n ×2 n The QR code watermark image is embedded into a H×W(2 n ≤min(H,W)) in any size of carrier image. The watermarking scheme in the present invention mainly includes two stages: watermark embedding and watermark extraction. The specific description is as follows:

[0011] In the watermark embedding stage, the overall process is as follows: Figure 1 The detailed steps are as follows:

[0012] Emb1, carrier image and watermark image are represented by TMQIR respectively. Based on the TMQIR model, when representing RGB true color image, θ=π, φ=π, l=3, l c =3; when representing a binary grayscale image, θ = π, φ = 0, l = 0, l c = 1. Take an RGB true color carrier image of size H×W and 2 n ×2 n (2 n ≤min(H,W)) size binary grayscale QR code watermark image as an example, the TMQIR model is used to represent |I C >and|I W >.

[0013] Emb2, convert the TMQIR color carrier image to TMQIR grayscale image. For example, each pixel position in the true color image has three color channels: red, green and blue. R ,I G and I B Indicates that the value of each color channel is an unsigned integer ranging from 0 to 255. In a grayscale image, grayscale refers to the color depth in a black and white image, represented by I grayIndicates. The values ​​of the matrix elements of a grayscale image are unsigned integers from 0 to 255, so its data type is usually an 8-bit unsigned integer. When a color image is converted to a grayscale image, the effective grayscale value of each pixel must be calculated, as shown in the following formula:

[0014] I gray =W R ×I R +W G ×I G +W B ×I B

[0015] Where W R ,W G and W B are the weights of the red, green and blue channels respectively, and W R +W G +W B = 1. Therefore, if any two of these three weights are determined, the corresponding grayscale image can be obtained.

[0016] In this invention, we introduce quantum color mapping |M Cg >, the information is converted from the RGB true-color TMQIR image stored on the quantum computer to the TMQIR grayscale image.

[0017] We designed and implemented a quantum circuit for converting RGB true-color TMQIR images to grayscale TMQIR images (see Figure 3 ), based on the quantum color map |M Cg >, realizing the transformation from color image to |I C >Transform to grayscale image |I Cg >, that is, in grayscale |I Cg >In, θ=π, φ=π / 2, l=3, l c =1.

[0018] Note: In the present invention, there is no Cg >Scrambling also reduces the complexity of the solution to a certain extent. Mainly considering |I Cg >There are only 8 bit planes, and the scrambling period is small. In addition, we later chose LSB-based steganography, so the scrambling has little significance for the system.

[0019] Emb3, for binary TMQIR watermark image |I W >Execute the Arnold scrambling algorithm. W > is a binary image. We use the improved quantum Arnold image scrambling algorithm to scramble the watermark image by changing the coordinates of the watermark image pixels. W >Change to unreadable watermark image|IWs >.

[0020] Emb4, quantum watermark image embedding based on LSB steganography algorithm. n ×2 n Binary TMQIR watermark image |I Ws > will be embedded in H×W(2 n ≤min(H,W)) grayscale TMQIR carrier image |I C > in the LSB. You need to change |I Ws > Zoom in, get Watermark image in Then, each pixel bit of the watermark image is embedded into the lowest bit of the color of the corresponding position of the carrier image in turn, and the TMQIR grayscale image embedded with the watermark is obtained. CWg >. The specific watermark is embedded in the quantum circuit as follows Figure 4 shown.

[0021] Emb5, restore the TMQIR grayscale image embedded with watermark to color mode. According to the two RGB three-color channel weight values ​​selected in the Emb2 step, the grayscale image can be restored to color, realizing the quantum circuit such as Figure 5 shown.

[0022] With the help of quantum color map |M Cg >, we can find the original TMQIR color map |I C >With the help of CWg >Convert to TMQIR color image|I CW >. The quantum circuit that realizes the conversion is as follows Figure 5 shown.

[0023] TMQIR color image with embedded watermark |I CW >, and can also be converted into other required representation models as needed.

[0024] In the watermark extraction phase, the overall process is as follows: Figure 2 The detailed steps are as follows:

[0025] Ext1, the color image embedded with watermark is represented by TMQIR model. Referring to step Emb1 of the watermark embedding stage, take the RGB true color watermarked image with size H×W as an example, and represent it by TMQIR, denoted as |I CW >.

[0026] Ext2, put the TMQIR image with watermark |I CW >Convert to TMQIR grayscale image|I CWg>, refer to step Emb2 of the watermark embedding stage, convert the quantum circuit used and Figure 3 resemblance.

[0027] Ext3, extract quantum watermark image based on LSB. In the TMQIR grayscale image of size H×W |I CWg >, according to step Emb4 of the watermark embedding stage, The lowest bit of the color in the row and column is extracted The binary TMQIR watermark image, if Then we need to further reduce the watermark image to 2 n ×2 n Watermark image |I Ws >. The quantum circuits required for the extraction process are as follows Figure 6 shown.

[0028] Ext4, reverse scramble the watermark image to restore the original TMQIR watermark image. Ws >2 n ×2 n The watermark image is the scrambled watermark image. It is necessary to perform the inverse process of the scrambling algorithm in step Emb3 of the embedding stage to restore the watermark image, thereby obtaining the original TMQIR watermark image |I W >. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is the quantum image watermark embedding flow chart of the present invention

[0030] Figure 2 This is the quantum image watermark extraction flow chart of the present invention

[0031] Figure 3 The quantum circuit diagram for converting the RGB true color TMQIR image to the grayscale TMQIR image of the present invention

[0032] Figure 4 The quantum circuit diagram of the watermark embedding process based on LSB of the present invention

[0033] Figure 5 The quantum circuit diagram for converting the grayscale TMQIR image to the RGB true color TMQIR image of the present invention

[0034] Figure 6 This is the quantum circuit diagram of the watermark extraction process based on LSB of the present invention DETAILED DESCRIPTION

[0035] In order to understand the present invention more clearly and in detail, the technical solution of the present invention will be specifically described in conjunction with the accompanying drawings. The description is only a part of the embodiments of the present invention, not all.

[0036] The present invention provides a multi-color mode quantum image blind watermarking method based on the TMQIR model. It mainly includes a watermark embedding stage and a watermark extraction stage. The specific steps are described as follows:

[0037] The flowchart of the watermark embedding phase is as follows: Figure 1 As shown, it mainly includes the following 5 steps.

[0038] Emb1, a 24-bit RGB carrier image of size H×W and 2 n ×2 n (2 n ≤min(H,W)) as an example, the TMQIR model is used to represent the binary grayscale QR code watermark image as |I C >and|I W >.

[0039]

[0040] in, z k ,y i ,x j ∈{0,1},, i∈{0,1,…,h-1}, j∈{0,1,…,w-1}, k∈{0,1,2}.

[0041]

[0042] Among them, |YX>=|Y>|X>=|y n-1 y n-2 …y 0 >,|x n-1 x n-2 …x 0 >, n = min{h,w}, |C YX >∈{0,1} is used to represent the binary color of the corresponding pixel |YX> position.

[0043] Emb2, put TMQIR color carrier map |I C >Convert to TMQIR grayscale image. Based on the three color channels of red, green and blue of true color I R ,I G and I B and weight W R ,W G and W B The weights are used to calculate the effective grayscale value of each pixel. The formula is as follows:

[0044] I gray =W R ×I R +W G ×IG +W B ×I B

[0045] Where W R +W G +W B =1.

[0046] Introducing Quantum Color Map |M Cg >, as shown in the following formula:

[0047]

[0048] Among them, |C>=|c 2 c 1 c 0 z 2 z 1 z 0 >,|c j >,|z j >∈{0,1},j∈{0,1,2}, respectively representing the color information and its position information, i∈{0,1,2}, respectively representing the grayscale information and its location information.

[0049] According to the conversion circuit of RGB true color TMQIR image to grayscale TMQIR (i.e. Figure 3 As shown), based on the quantum color map |M Cg >, put the color image |I C >Convert to grayscale image|I Cg >,|I Cg >As shown in the following formula.

[0050]

[0051] in,

[0052] |ZYX>=|z 2 z 1 z 0 >|y h-1 y h-2 …y 0 >|x w-1 x w-2 …x 0 >, c ZYX ,z k ,y i ,x j ∈{0,1}, i∈{0,1,…,h-1}, j∈{0,1,…,w-1}, k∈{0,1,2}, |C ZYX >∈{0,1} represents the grayscale information of the corresponding pixel.

[0053] Emb3, according to the improved binary quantum image Arnold scrambling algorithm, change the binary TMQIR watermark image |I W >pixel coordinates, making it a scrambled binary watermark image |I Ws >.

[0054] Emb4, if the scrambled 2 n ×2 n Binary TMQIR watermark image |I Ws > According to the nearest neighbor interpolation method in the literature, |I Ws > Zoom to get the watermark image Size

[0055] For ease of description, we put the front of the carrier map OK, go ahead The area of ​​the column is recorded as an image exist and At the same coordinate position of the two, exchange the lowest bit of the gray value at the coordinate point, and at the same time Set the coordinates of 0 to complete Embedded in carrier image In the lowest bit of the corresponding pixel, the watermark is embedded Size Finally, we get a grayscale image of H×W with watermark |I CWg >, TMQIR is represented as follows:

[0056]

[0057] in Specific quantum circuits such as Figure 4 shown.

[0058] Emb5, TMQIR grayscale image with watermark |I CWg >Convert to TMQIR color image|I CW >. The conversion formula is as follows:

[0059]

[0060] With the help of quantum color map |M Cg >, in the original TMQIR color map |I C >, convert the watermarked TMQIR grayscale image to |I CWg > is the TMQIR color map |I CW >. The quantum circuit that realizes the conversion is as follows Figure 5 shown.

[0061] The overall process of the watermark extraction stage is as follows: Figure 2 As shown, it mainly includes the following 4 steps.

[0062] Ext1, the watermark image is embedded. Take the RGB true color image with a size of H×W as an example, and use TMQIR to represent it, denoted as |I CW >, the expression is as follows:

[0063]

[0064] here,

[0065] |ZYX>=|z 2 z 1 z 0 >|y h-1 y h-2 …y 0 >|x w-1 x w-2 …x 0 >, z k ,y i ,x j ∈{0,1}, i∈{0,1,…,h-1}, j∈{0,1,…,w-1}, k∈{0,1,2}, It represents the color information of the three channels R, G, and B at the corresponding pixel |ZYX> respectively.

[0066] Ext2, the TMQIR image with watermark |I CW >Convert to TMQIR grayscale image|I CWg >, the conversion process refers to the watermark embedding process Emb2. The quantum circuit diagram used for the conversion is the same as Figure 3 Similar. The obtained watermarked grayscale image |I CWg >The TMQIR is represented as follows:

[0067]

[0068] in |C ZYX >∈{0,1} represents the grayscale information of the corresponding pixel.

[0069] Ext3, from the TMQIR grayscale image of size H×W |I CWg >Previous Line and Column (the area marked ), according to the LSB steganographic algorithm, the lowest bit is extracted and processed as 2 n ×2 n Binary TMQIR watermark image |IWs >, where The detailed process is as follows:

[0070] First The qubit is initialized to |0>, which is used to represent the watermark image to be extracted. in The qubits are used to represent the row and column coordinates of the watermark image, and the states are set to the superposition of |0> and |1>. The remaining qubit is used to represent the binary grayscale of the corresponding pixel.

[0071] Then in the carrier image The coordinates and watermark image At the same coordinate position of the corresponding coordinates, exchange the lowest bit of the gray value of the coordinate point. The corresponding coordinates of are set to |0>. Thus, the size is obtained Binary watermark image Its TMQIR is expressed as follows: The quantum circuit for extracting the QR code watermark image from the carrier image based on LSB is as follows: Figure 6 shown.

[0072]

[0073] in |C YX >∈{0,1} is used to represent the binary color information for the pixel position |YX>.

[0074] The last extracted Watermark image According to the nearest neighbor interpolation method in the literature, the 2 n ×2 n Binary TMQIR watermark image |I Ws >.

[0075] Ext4, extract 2 n ×2 n Watermark image |I Ws >Perform inverse scrambling, that is, perform the inverse process of the scrambling algorithm of the embedding process Emb3 to restore the original binary TMQIR watermark image |I W >.

Claims

1. A multi-color mode quantum image blind watermark embedding method based on TMQIR, characterized by , including the following steps: Emb1: a vector image of arbitrary size H×W and 2 n ×2 n (2 n The binary grayscale QR code watermark image with a size of ≤min(H,W)) is represented by the TMQIR model, and the carrier image is a 24-bit true color carrier image; Emb2: Convert TMQIR color carrier image to TMQIR grayscale image, red, green and blue three color channels in TMQIR R ,I G and I B The corresponding weight values ​​W R ,W G and W B Given two, with the help of quantum color mapping, according to formula I gray =W R ×I R +W G ×I G +W B ×I B Get the TMQIR grayscale carrier image; Emb3: Improved Arnold scrambling of binary TMQIR watermark image; Emb4: Embed the quantum watermark image based on the LSB steganographic algorithm, appropriately scale the scrambled TMQIR watermark image, and then exchange the lowest bit of the color of the corresponding pixel with the TMQIR grayscale carrier image to embed the watermark; Emb5: Based on the set color weights, the TMQIR grayscale image embedded with the watermark is converted into the original color mode with the help of the quantum color map.

2. A multi-color mode quantum image blind watermark extraction method based on TMQIR, characterized in that The following steps are involved: E xt1: The carrier image of any size H×W with embedded watermark is represented by TMQIR model, and the carrier image is a 24-bit true color carrier image; Ext2: Convert the TMQIR watermark image to a grayscale image and set the weight value W of the three color channels RGB R ,W G and W B Any two of them, with the help of quantum color mapping, according to formula I gray =W R ×I R +W G ×I G +W B ×I B , convert the TMQIR color image embedded with watermark into TMQIR grayscale image; Ext3: extracting the watermark image from the grayscale TMQIR image embedded with the watermark, extracting the watermark from the corresponding pixel area based on the LSB technology, and appropriately adjusting the image size to obtain the TMQIR watermark image. This process is the inverse process of the embedding process in step Emb4 in claim 1; Ext4: Perform Arnold inverse scrambling on the extracted watermark image to restore the original TMQIR watermark image. This process is the inverse process of Emb3 in claim 1.

3. The watermark embedding quantum circuit design method according to claim 1, characterized in that To implement the watermark embedding process described in step Emb4 of claim 1, based on the LSB algorithm, the binary scrambled TMQIR watermark image I Ws >Embedded into TMQIR grayscale carrier image I Cg >In the least significant bit of the corresponding pixel.

4. The watermark extraction quantum circuit design method according to claim 2, characterized in that To implement the watermark extraction process described in step Ext3 of claim 2, based on the LSB algorithm, in the TMQIR grayscale image I of size H×W CWg >, extract the TMQIR watermark image from the lowest bit of the corresponding pixel area.

5. The method for designing a quantum circuit for mutual conversion between true color TMQIR images and grayscale TMQIR images according to claims 1 and 2, characterized in that : Based on the weight value W of the RGB three-color channels R ,W G and W B Any two of them can be used to realize the mutual conversion between the RGB true color image and the grayscale image represented by TMQIR with the help of quantum color mapping.

Images