TWI823814B - QR code verification system and method for protecting digital images - Google Patents

Abstract

The present invention discloses a QR code verification system and method that can protect digital images. It converts digital images into frequency domain blocks and embeds the first irregular QR code sequence into the DC area coefficient in the frequency domain of the hidden block image. , causing the digital image to generate a dominant watermark, embedding the second irregular QR code sequence into the mid-frequency area coefficient hidden in the frequency domain of the block image, causing the digital image to generate a recessive watermark, and hiding the first irregular QR code sequence The block image frequency domain of the code sequence and the second informal QR code sequence is converted back to a digital image, so that the digital image becomes a watermark image with dual protection of explicit QR code watermark and recessive watermark, so that it can be used through explicit The establishment of double protection QR code verification technology with hidden watermark changes the internal RS code design of the QR code to generate a new non-standard QR code to ensure the safety and protection of the use of electronically painted image works.

Description

QR code verification system and method for protecting digital images

The present invention relates to a QR code verification system and method that can protect digital images. In particular, it refers to a QR code that can protect the security of digital images in use through dual protection QR code verification technology of explicit and implicit watermarks. Verification technology.

Press, in many application fields, the demand for efficient and reliable information hiding technology is constantly increasing. Some related paper research points out the theory and design of codes that hide or embed information in multimedia such as images, videos, audios, graphics, and text. Such codes are also called watermark codes; they can be used in a variety of applications, including copyright protection of digital media, media forensics, content authentication, data binding, and covert communications, such as references [1-5]. Secondly, the emergence of the Internet and public communication networks has given rise to many new technological applications, among which information hiding technology plays an important role. The main goal of these technologies is to make multimedia image data more reliable and secure.

Digital watermarking is an emerging technical method in various fields. It is a process of embedding information called watermarks into multimedia objects such as images, audios, videos, etc., so that the watermarks can be extracted. A watermark may be a visible portion of an image placed on it to identify copyright. Generally speaking, a digital watermark system consists of several parts: (1) watermark, (2) encoder; and (3) decoder and comparator, as shown in a paper review in Reference [6] The theory and design of watermark scheme codes, such as references [7-8], discuss some of the available video watermarking technologies. For example, Reference [9-12] proposed a video standard watermarking scheme. Watermark message or watermark request Either embedded into the encoded video stream or can be retrieved from the codec video. The main idea of video watermarking is to add a watermark message in the video. When the message is lower than the perception threshold, it cannot be recognized or verified, so it is deleted without knowing the parameters of the watermark algorithm.

Furthermore, the digital image verification system is a very important part of network security. Among various digital image formats, JPEG encoding is an important and indispensable encoding format. Generally, when digital images are transmitted on the Internet or in digital systems, they often encounter infringements in the form of remakes and copies. Explicit watermarking methods in the world are mainly used to prevent direct theft of digital images, while some protection methods use hidden watermarks. The main purpose of protecting digital watermarks is to prevent various signal processing (such as cropping, pressing, filtering, etc.) from infringing on digital images. Therefore, how to develop a method that can avoid infringement of digital images? Behavioral image QR code verification technology has become a technical issue that industry, government and scholars in related technical fields are eager to solve and challenge.

In view of this, based on the urgent needs of related industries, the inventors of the present invention relied on many years of practical experience and relevant professional knowledge, and after continuous efforts in research and development, they finally developed a method that is different from the above-mentioned conventional technology and the previous ones. Uncover the patented invention.

The main purpose of the present invention is to provide a QR code verification system and method that can protect digital images, mainly by changing the internal RS code design of the QR code through the establishment of explicit and implicit watermark dual protection QR code verification technology. New informal QR code to ensure the safety and protection of the use of electronically painted image works. The technical means to achieve the main purpose of the present invention is to convert the digital image into a frequency domain block, and embed the first irregular QR code sequence into the DC area coefficient in the frequency domain of the hidden block image, so that the digital image generates a dominant watermark. Will the second informal QR The code sequence is embedded in the mid-frequency area coefficient hidden in the frequency domain of the block image, causing the digital image to generate a hidden watermark. The first irregular QR code sequence and the second irregular QR code sequence are embedded in the frequency domain of the block image. Convert it back to a digital image, turning the digital image into a watermark image with dual protection of explicit QR code watermark and recessive watermark.

10:Image website platform

20:Network link module

30: Verify the servo platform

40:Terminal device

50: Random sequence generator

DC: DC area coefficient

IF: Intermediate frequency area coefficient

Figure 1 is a schematic diagram of the specific implementation of the system architecture of the present invention.

Figure 2 is a schematic diagram of the basic architecture of the present invention for embedding QR code watermark protection into digital images.

Figure 3 is a schematic diagram of the specific architecture implementation of the present invention for embedding QR code watermark protection into digital images.

Figure 4 is a schematic diagram of the intensity adjustment implementation of the watermark image according to the present invention.

Figure 5 is a schematic diagram of the implementation of embedding DC low-frequency coefficients and intermediate-frequency coefficients into digital images according to the present invention.

Figure 6 is a schematic diagram of the specific implementation of the random sequence generator of the present invention.

In order to allow your review committee to further understand the overall technical characteristics of the present invention and the technical means to achieve the purpose of the present invention, specific embodiments are described in detail with the drawings as follows: Please refer to Figures 1 to 3. In order to achieve the purpose of the present invention, The first embodiment of the digital image protection method that is the main object of the invention uses a QR code verification system, which includes an image website platform 10, a network link module 20 and a verification server platform 30. The image website platform 10 is used to provide digital images. The verification server platform 30 is built with a first irregular QR code sequence and a second irregular QR code sequence, and is connected to the image website platform 10 through the network link module 20 to obtain digital images, and then converts the digital images into is a frequency domain block, and the first irregular QR code sequence is embedded into the DC area coefficient DC in the frequency domain of the hidden block image to produce an explicit watermark on the digital image, and then the second irregular QR code sequence is embedded The intermediate frequency coefficient IF hidden in the block image frequency domain is used to generate a hidden watermark in the digital image, and embeds the block image frequency that hides the first irregular QR code sequence and the second irregular QR code sequence. The domain is converted back to a digital image, making the digital image a watermark image with dual protection of explicit QR code watermark and recessive watermark.

In a specific implementation based on the first embodiment, the verification server platform 30 communicates with multiple computers through a network connection module 20 (such as a combination of the Internet and a networked device; or a combination of a mobile communication network, a networked device, and the Internet). The terminal device 40 information is connected. When one of the terminal devices 40 makes a download image request, the verification server platform 30 adjusts the λ value of the protected watermark image so that the explicit QR code watermark strength of the watermark image is sufficient for the terminal device. 40's optical recognition module (such as a camera module) recognizes the content of the explicit QR code watermark, and uses the optical recognition module to detect the explicit QR code watermark, decode it, and then decode the decoded content. According to the hidden intermediate frequency coefficient IF embedded in the recessive watermark pointed out by the decoding content, the second irregular QR code sequence is taken out at that position, and then the data of the second irregular QR code sequence is explicitly floated. The watermark information on the DC area coefficient DC of the corresponding position of the watermark has been offset, so that the terminal device 40 can display or store the digital image with the dominant watermark offset.

Please refer to Figures 2 to 3. In order to achieve the main purpose of the present invention, the digital image protection method uses a second embodiment of the QR code verification method, which includes the following steps:

(a) Preparation step: provide digital image, first irregular QR code sequence and second irregular QR code sequence.

(b) The image frequency domain conversion step is to convert the digital image into frequency domain blocks; specifically, the at least one digital image color is converted into a brightness layer (Y), a chroma layer (Cb) and a chroma layer (Cr). ) and other three layers, and cut the brightness layer (Y) into 64 sub-blocks of 8×8, and perform discrete cosine transform (DCT) on the 64 sub-blocks to generate coefficients in the frequency domain of the block image.

(c) The explicit watermark embedding step is to embed the first informal QR code sequence into a direct current area coefficient DC in the frequency domain of the hidden block image to generate an explicit watermark in the digital image; specifically, the non-regular QR code sequence is adjusted to The size of the regular QR code sequence becomes the first irregular QR code sequence. After selecting the embedded sub-block of the DC area coefficient DC, the 0 and 1 digits in the first irregular QR code sequence are converted into -1 and +1. , then multiply it by the λ value, adjust its +1 and -1 sizes, and add the DC coefficient of the original frequency domain (original value S ) to form the digital image with explicit watermark protection, explicit watermark protection The digital image is represented by the following formula:

S _w = S + λ (2 M _re -1), where S is the original value of the DC coefficient embedded sub-block, the λ value is a constant that can be adjusted, and M _re is the irregular QR code sequence. .

(d) The implicit watermark embedding step is to embed the second informal QR code sequence into the intermediate frequency area coefficient IF hidden in the frequency domain of the block image to generate an implicit watermark in the digital image; specifically, the second The non-normal QR code sequence is the non-normal QR code sequence. In the implicit watermark embedding step, after selecting the embedded sub-block of the intermediate frequency region coefficient IF, the second non-normal QR code sequence is sequentially replaced. The original frequency domain coefficient embedded in the sub-block of the intermediate frequency region coefficient IF.

(e) Image double protection synthesis step, convert the frequency domain of the block image that has embedded the first irregular QR code sequence and the second irregular QR code sequence back into a digital image, so that the digital image becomes an explicit QR code watermark A watermark image with dual protection of a hidden watermark; specifically, the discrete cosine transform (DCT) is used to convert the frequency domain of the block image that has embedded the first irregular QR code sequence and the second irregular QR code sequence back into Luminance layer (Y), and then combine the information of the brightness layer (Y) with the chroma layer (Cb) and the chroma layer (Cr) to make the at least one digital image become a dominant QR code and a recessive QR code. This watermark image is double protected by watermark.

In a specific embodiment based on the second embodiment, the first irregular QR code sequence lengthens the length of the irregular QR code sequence by K times. The first irregular QR code sequence is represented by the following formula:

，K表示為非正規QR碼序列的放大 倍數。 M _re ={ m ₁ , m ₂ ,L, m _{N × K} }, the pixel of the digital image is L , and the pixel whose length is lengthened to K times is N. The calculation of K is,

, K represents the amplification factor of the informal QR code sequence.

A specific embodiment based on the second embodiment further includes a watermark image reply step, which includes the following steps:

(f) Step: Adjust the λ value of the protected watermark image so that the explicit QR code watermark intensity of the watermark image is sufficient for an optical recognition module to identify the content of the explicit QR code watermark, such as As shown in Figure 4.

(g) In step (g), the optical recognition module detects the explicit QR code watermark and then decodes the decoded content.

(h) Step: According to the intermediate frequency region coefficient IF embedded in the hidden watermark pointed out by the decoded content, and go to the position to retrieve the second irregular QR code sequence.

Step (i): Offset the watermark information on the DC area coefficient DC at the position corresponding to the explicit watermark of the retrieved second informal QR code sequence data.

Please refer to FIG. 6 , which is a specific embodiment based on the second embodiment. This embodiment further includes an informal QR code generation step, which uses a random sequence generator 50 to generate a random number, and uses The random number is selected as a random polynomial to generate a binary sequence with a length of up to 2 ^m -1 and an irregular decimal QR code sequence, where m is the level of the displacement register of the random sequence generator 50 Count.

This invention utilizes the currently very popular QR code to verify JPEG digital images with private copyright, which is the main contribution of the invention. Since QR codes allow users to obtain information simply and conveniently through optical capture, QR codes can also be achieved using the camera function of mobile phones. Capture purpose is a very convenient optical module technology. The present invention proposes a method of using QR code verification for security protection of JPEG digital images. The QR code generator used in the present invention is an informal QR code encoding method. Therefore, it has a better confidentiality effect in verification than the standard QR code. Based on the proposed method, the present invention generates and converts irregular QR codes into sequence patterns and embeds them in the frequency domain of JPEG digital images. The embedding method uses two sets of QR code sequences. One of them is hidden in the DC area coefficient. The main function is One is to produce a dominant watermark. This purpose can be used to infringe the infringement of directly reproducing images on the Internet. The other is to hide in the frequency domain coefficients of JPEG images in the mid-band. Its main function is to generate a hidden watermark. Its purpose The main purpose is to prevent infringement of digitally stored images.

The block diagram of the JPEG digital image protection system of the present invention is shown in Figure 2. This invention converts JPEG digital images into three levels: Y, Cb and Cr. This invention uses the brightness layer Y and then cuts its plane into 8 by 8 sub-blocks for discrete cosine transform (DCT) to generate Coefficients in the frequency domain of the block image. The present invention uses informal QR codes for hiding. The hiding areas are the DC area and the mid-frequency band area in the frequency domain. The QR codes hidden in the DC area mainly produce explicit watermark functions. In addition, the QR codes embedded in the mid-frequency band The main function of the QR code is to generate a recessive watermark, and then convert the hidden frequency domain block back to the Y level and then combine it with the information at the two chroma levels of Cb and Cr to generate explicit and recessive watermarks. Double protection of color image data.

In Figures 2~3, frequency domain conversion uses DCT conversion to generate frequency domain information for each sub-block. The frequency domain description of each sub-region is shown in Figure 5.

In Figure 5, the main hiding locations of the present invention are the direct current area (DC area) and the intermediate frequency area (IF area). First, the explicit watermark hiding method is as follows: after selecting the appropriate embedding block through the adjusted informal QR code sequence M _re , the present invention converts the 0 and 1 digits in the QR code sequence into -1 and + 1, then multiply by the λ value, adjust its size of +1 and -1, and finally add the DC coefficient of the original frequency domain (original value S in the red area) to form the final JPEG watermark protected digital image, as follows Mode:

S _w = S + λ (2 M _re -1)

The above λ value is a constant that can be adjusted. Its function is to adjust the intensity of the explicit watermark to the intensity that the optical system can take pictures for subsequent QR code decoding. In addition, the embedding method of the implicit watermark is as follows: first, the informal QR code sequence is sequentially replaced by the original frequency domain coefficient at the IF position in Figure 5 in the required sub-block. Although the frequency domain coefficient at the IF position is replaced by the QR code The value replacement of the sequence produces frequency distortion in the mid-band, but the change cannot be clearly detected in human vision. Figure 3 illustrates how to perform explicit and implicit watermarking. The red area is mainly the location where explicit watermarks are hidden. Assume that the QR code sequence is:

M ={ m ₁ , m ₂ ,L, m _N }

The present invention lengthens its sequence length to K times, as follows:

M _re ={ m ₁ , m ₂ ,L, m _{N × K} }

Assume that the original JPEG digital image pixel is L , and the pixel length extended to K times is N. The calculation of K in the above is as follows:

K represents the multiple that the original QR code sequence should be enlarged to be closest to the QR code explicit watermark that is similar to the original image after adjustment. For the hidden watermark, the original QR code M sequence is used for intermediate frequency to randomly select a hiding position for replacement, such as the IF position in Figure 5. The present invention should select the corresponding yellow color in N 8×8 frequency domain sub-blocks. The position is replaced by the M sequence value. The yellow position corresponding to this mid-frequency band is a secret position. Most people will not know where this position is. The M sequence has two main functions: 1. It is used for authentication of hidden watermarks. 2. The M sequence in the mid-frequency band can be used to remove the explicit watermark on the digital image that has been concealed.

Furthermore, in order to prevent the embedded explicit watermark QR code from being scanned and decoded by ordinary scanners (such as the camera module built into a smartphone), the present invention uses the following method to design the QR code to obtain an informal QR code. Code decoding method to increase its security. It is known that a polynomial can be decomposed as follows:

The current standard QR code generation polynomial is as follows:

g ( x )=< m ₁ ( x ), m ₂ ( x ),L, m _{2 t} ( x )>

The code generated by RS is:

c ( x ) = u ( x ) g ( x )

The above u ( x ) is the message polynomial. In the following, the present invention uses a random sequence generator to generate a random number, and uses this random number to select a random formula to generate a polynomial. The random sequence generator used in the present invention is shown in Figure 6.

The above structure can generate binary sequences and decimal sequences with a length of up to 2 ^m -1. The present invention uses the following example to illustrate. Assuming a virtual sequence generator as shown in Figure 6, it can generate a binary sequence with a length of 15 bits, such as: 000100110101111; a decimal sequence with a length of 15 bits, such as: R=1,2, 4,8,3,6,13,10,5,11,7,15,14,12,8,…, and use decimal random sequences to perform denormalized polynomial design of QR code generation, if multinomial m _i ( x ) is any open continuous root polynomial, then the decimal value in the R sequence can be selected as the starting continuous root polynomial m _i ( x ). For example, if the random decimal value is 6, an irregular QR code will be generated. The polynomial is:

g ₆ ( x )=< m ₆ ( x ), m ₇ ( x ), m ₈ ( x ), m ₉ ( x )>

The above-mentioned generating polynomial uses four polynomial roots, so the generating polynomial g ₆ ( x ) has the ability to correct two errors, and the generating polynomial is randomly selected, unlike the standard QR code generating polynomial which has a fixed pattern. The encoding method of informal QR codes uses a virtual sequence generator to randomly generate polynomials, which can more effectively improve the security of embedded QR code information.

In addition, the present invention demonstrates the proposed watermark embedding method, as shown in Figure 2. This work uses some DCT blocks to insert watermarks. The image watermark uses 8×8 partition blocks to obtain the DCT of the original image. Use QR codes to select and modify coefficients corresponding to low and mid frequencies, i.e. watermarks generated by resizing the original image. In Figure 5, the watermark looks like an artifact of the original JPEG image. The principle of the watermarking method proposed in Figure 5. This work in Figure 3 computes DCT transforms of the entire image and limits their embedding process to low frequencies to achieve explicit watermarking. The embedding method directly embeds the watermark in the DC value of the DCT coefficient. For each bit of the informal QR code sequence Mre to be embedded, a group of blocks is concentrated in a low-frequency value. The explicit watermark embedding system of the present invention can be represented by the block diagram shown in Figure 5. Figure 5 The embedding process of the proposed system; in the JPEG image, the quantized DCT coefficients in the DC value S are encoded using M _re -1 QR encoding, as shown on the right: W=S+ λ (2M _re -1). The λ value represents the scale factor of the watermark. Based on the low frequency of DCT, a masking threshold is required for each block. Assume that the size of the QR code is L×L and the size of the JPEG image is N×N. A single block of explicit watermark with QR code is a resized sequence, L[N/L×L[N/L].

For implicit watermarking, the embedding of the QR code M sequence needs to be embedded into the secret channel. This channel is located in the mid-frequency area of the DCT frequency domain so that JPEG images have better quality. As shown in Figure 4, channels can be selected in the intermediate frequency area of the DCT frequency domain. The purpose of this step process is to restore the explicit watermark image to the original image.

This experimental example discusses the performance between original JPEG images and watermarked JPEG images. As mentioned in image quality estimation, the embedding process can estimate the quality of the watermarked image. We will estimate the peak signal-to-noise ratio (PSNR) of the test images of the proposed method. Experimental results are simulated on images (512×512 pixels) used for testing and message QR codes for each image. The peak signal-to-noise ratio of the watermark image is defined as:

Among them, MSE represents the mean square error between the original image Io and the watermarked image Iwo.

Table 1 shows simulated watermarked JPEG images with various categories. As described in the present invention, the copyright protection and authentication method uses a private QR code that cannot be decoded using standard decoding algorithms. The total number of explicit and implicit watermarks is:

For various test images, the watermark intensity factor is about 60, and the watermark image quality is about 34.

The results report that the proposed method is an effective authentication scheme, as shown in Table 1.

The experimental results of the present invention demonstrate the authentication performance of various levels of intensity factor λ , and enhance the restored image for the purpose method to a sufficient intensity level for decoding. The method was simulated to test their verified QR code, as shown in Figure 4. Figure 4: Image representing the intensity of the embedded watermark. The watermark system shown in Figure 4 simulates the results of various test JPEG images. Embed watermark images of size 21×21 into various test images. Compared with the original JPEG images shown in Table 1, the quality of the watermarked images will be much lower. Figure 4 shows the certification test of different levels of intensity factor λ for various JPEG images. The implementation of the proposed method is applied in the field of web design, and the present invention proposes a family of alternative verifiable watermark and copyright protection schemes. The advantage of the proposed method is that it exploits both explicit and implicit features to protect images without computation. And based on the theoretical decomposition, each image element in the image can be transformed into the frequency domain, and the watermark of the QR code can be embedded in the JPEG image. In the experimental example, a QR code with an RS code generating polynomial private decoding structure is used to construct the watermark image. Therefore, the DCT frequency domain conversion method proposed by the present invention does provide a simple coefficient embedding process.

After the description of the above specific embodiments, the present invention can indeed change the internal RS code design of the QR code to generate a new non-normalized QR code through the establishment of the explicit and implicit watermark dual protection QR code verification technology to ensure that the electronic The safety and protection of the use of graphic works.

The above are only possible embodiments of the present invention and are not intended to limit the present invention. Within the patent scope of the present invention, any equivalent implementation with other changes based on the content, features and spirit described in the following claims shall be included in the patent scope of the present invention. The structural features specifically defined in the claim of the present invention have not been found in similar articles, and are practical and progressive. They have met the requirements for an invention patent. I file an application in accordance with the law. I sincerely request the Office to approve the patent in accordance with the law to protect this invention. The legitimate rights and interests of the applicant.

References

[1] FAP Petitcolas, RJ Anderson and MG Kuhn, "Information hiding-a survey," in Proceedings of the IEEE , vol. 87, no. 7, pp. 1062-1078, July 1999, doi: 10. 1109/5 .771065.

[2] Y. Tew and K. Wong, "An Overview of Information Hiding in H. 264/AVC Compressed Video," in IEEE Transactions on Circuits and Systems for Video Technology , vol. 24, no. 2, pp. 305- 319, Feb. 2014, doi: 10.1109/TCSVT. 2013. 2276710.

[3] K. Cabaj, L. Caviglione, W. Mazurczyk, S. Wendzel, A. Woodward and S. Zander, "The New Threats of Information Hiding: The Road Ahead," in IT Professional , vol. 20, no. 3, pp. 31-39, May./Jun. 2018, doi: 10.1109/MITP. 2018. 032501746.

[4] S. Rahman et al., "A Comprehensive Study of Digital Image Steganographic Techniques," in IEEE Access , vol. 11, pp. 6770-6791, 2023, doi: 10. 1109/ACCESS. 2023. 3237393.

[5] J. Lubacz, W. Mazurczyk and K. Szczypiorski, "Principles and overview of network steganography," in IEEE Communications Magazine , vol. 52, no. 5, pp. 225-229, May 2014, doi: 10.1109/ MCOM. 2014. 6815916.

[6] O. Evsutin, A. Melman and R. Meshcheryakov, "Digital Steganography and Watermarking for Digital Images: A Review of Current Research Directions," in IEEE Access , vol. 8, pp. 166589-166611, 2020, doi: 10.1109/ACCESS. 2020. 3022779.

[7] IJ Cox, J. Kilian, FT Leighton and T. Shamoon, "Secure spread spectrum watermarking for multimedia," in IEEE Transactions on Image Processing , vol. 6, no. 12, pp. 1673-1687, Dec. 1997 , doi: 10.1109/83. 650120.

[8] IJ Cox, ML Miller and AL McKellips, "Watermarking as communications with side information," in Proceedings of the IEEE , vol. 87, no. 7, pp. 1127-1141, July 1999, doi: 10.1109/5. 771068.

[9] H. Mareen et al., "Camcording-Resistant Forensic Watermarking Fallback System Using Secondary Watermark Signal," in IEEE Transactions on Circuits and Systems for Video Technology , vol. 31, no. 9, pp. 3403- 3416, Sept . 2021, doi: 10. 1109/TCSVT. 2020. 3042882.

[10] W. Huan, S. Li, Z. Qian and X. Zhang, "Exploring Stable Coefficients on Joint Sub-Bands for Robust Video Watermarking in DT CWT Domain," in IEEE Transactions on Circuits and Systems for Video Technology , vol. . 32, no. 4, pp. 1955-1965, April 2022, doi: 10. 1109/TCSVT. 2021. 3092004.

[11] H. Mareen, M. Courteaux, J. De Praeter, M. Asikuzzaman, G. Van Wallendael and P. Lambert, "Rate- Distortion-Preserving Forensic Watermarking Using Quantization Parameter Variation," i n IEEE Access , vol. 8, pp. 63700-63709, 2020, doi: 10. 1109/ACCESS. 2020. 2984354.

[12] R. Ullah, SD Khan, M. Ullah, F. Al-Machot and H. Ullah, "Toward Authentication of Videos: Integer Transform Based Motion Vector Watermarking," in IEEE Access , vol. 10, pp. 75063- 75073, 2022, doi: 10.1109/ACCESS.2022.3191667.

10:Image website platform

20:Network link module

30: Verify the servo platform

40:Terminal device

Claims (8)

A QR code verification system that can protect digital images, which includes: an image website platform for providing at least one digital image; a network link module; and a verification server platform with at least one first non- The regular QR code sequence and at least one second irregular QR code sequence are connected to the image website platform information through the network link module to obtain the at least one digital image, and then the at least one digital image is converted into the frequency domain. block, and embed the first irregular QR code sequence into a direct current area coefficient that hides the block image frequency domain, so as to generate an explicit watermark on the at least one digital image, and then embed the second irregular QR code sequence into the block image. The code sequence embeds an intermediate frequency region coefficient hidden in the frequency domain of the block image to generate a hidden watermark on the at least one digital image, and embeds and hides the first irregular QR code sequence and the second irregular QR code sequence. The frequency domain of the block image of the regular QR code sequence is converted back to the digital image, so that the digital image becomes a watermark image with dual protection of explicit QR code watermark and recessive watermark. A QR code verification method that can protect digital images, which includes: a preparation step of providing at least one digital image, at least a first non-normal QR code sequence and at least a second non-normal QR code sequence; an image frequency domain conversion step of converting The at least one digital image is converted into a frequency domain block; the explicit watermark embedding step is to embed the first irregular QR code sequence into a direct current region coefficient that hides the block image in the frequency domain, so that the at least one digital The image generates a dominant watermark; the recessive watermark embedding step is to embed the second irregular QR code sequence into an intermediate frequency region coefficient hidden in the frequency domain of the block image to generate a recessive watermark on the at least one digital image. watermark; and an image double protection synthesis step, converting the frequency domain of the block image that has been embedded and hiding the first irregular QR code sequence and the second irregular QR code sequence back into the digital image, so that the digital image It becomes a watermark image with dual protection of explicit QR code watermark and recessive watermark. The QR code verification method that can protect digital images as described in claim 2, wherein in the image frequency domain conversion step, the at least one digital image color is converted into a brightness layer (Y) and a chroma layer (Cb) and chroma layer (Cr), and cut the brightness layer (Y) into 64 sub-blocks of 8×8, and perform discrete cosine transform (DCT) on the 64 sub-blocks to generate the Coefficients in the frequency domain of the block image. The QR code verification method capable of protecting digital images as described in claim 3, wherein in the image double protection synthesis step, the first irregular QR code sequence and the second irregular QR code sequence are embedded and hidden. The frequency domain of the block image is converted back to the brightness layer (Y), and then the brightness layer (Y) is combined with the information from the three levels of the chroma layer (Cb) and the chroma layer (Cr), so that the At least one digital image becomes the watermark image with dual protection of an explicit QR code and a recessive watermark. The QR code verification method that can protect digital images as described in claim 2, wherein in the explicit watermark embedding step, the size of the irregular QR code sequence is adjusted to become the first irregular QR code sequence. After the DC area coefficient is embedded in the sub-block, the 0 and 1 digits in the first informal QR code sequence are converted into -1 and +1, and then multiplied by the λ value to adjust the +1 and -1 digits. size, plus the DC coefficient of the original frequency domain (original value S ), to form the digital image protected by the explicit watermark. The digital image protected by the explicit watermark is expressed as follows: S _w = S + λ (2 M _re -1), where S is the original value of the DC coefficient embedded sub-block, the λ value is a constant that can be adjusted, and M _re is the informal QR code sequence. The QR code verification method capable of protecting digital images as described in claim 2, wherein the first irregular QR code sequence lengthens the length of the irregular QR code sequence by K times, and the representation of the first irregular QR code sequence The following formula: M _re ={ m ₁ , m ₂ ,L, m _{N × K} }, the pixel of the digital image is L , and the pixel whose length is lengthened to K times is N. The calculation of K is,

, K represents the amplification factor of the informal QR code sequence. The QR code verification method that can protect digital images as described in claim 2, wherein the second non-normal QR code sequence is the non-normal QR code sequence, and in the hidden watermark embedding step, the intermediate frequency is selected After the region coefficients are embedded in the sub-block, the second irregular QR code sequence is sequentially replaced by the original frequency domain coefficients in the embedded sub-block of the mid-frequency region coefficients. The QR code verification method that can protect digital images as described in claim 2 further includes an informal QR code generation step, which uses a random sequence generator to generate a random number, and uses the random number to perform random processing. The polynomial is selected to generate a binary sequence with a length of up to 2 ^m -1 and an irregular decimal QR code sequence, where m is the number of stages of the displacement register of the random sequence generator.

Images