Adaptive amplitude modulation screening digital watermark optimization method and system

By analyzing the mutual influence of CMYK four-color channels and mask marking technology, the problems of poor watermark area selection and low image clarity are solved, and high-quality half-tone image watermark hiding and extraction effects are achieved.

CN119006259AActive Publication Date: 2024-11-22XIDIAN UNIV

Patent Information

Application Number
CN202411211492.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-22
Estimated Expiration
2044-08-30

AI Technical Summary

Technical Problem

The prior art fails to fully consider the four-color channel characteristics in the hiding of digital watermarks in halftone images, resulting in poor selection of watermark areas, affecting the hidden effect and extraction effect, and at the same time, the image clarity is not high after embedding the watermark.

Method used

By analyzing the mutual influence of the four-color channels of CMYK, the fitness of the carrier image is calculated, the most suitable watermark area is found, and the mask mark is used to maintain the clarity of the non-watermark area. The watermark embed is embedded using rational tangent meshing technology.

Benefits of technology

It improves the concealment of watermarks and the clarity of the image, ensures the rationality and concealment of the watermark area, and achieves high-quality half-tone target image generation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119006259A_ABST
    Figure CN119006259A_ABST
Patent Text Reader

Abstract

The invention discloses an optimization method and system of a self-adaptive amplitude modulation screening digital watermarking, and the method comprises the steps: converting an input RGB image into a CMYK image suitable for printing, and splitting four color channels of cyan, magenta, yellow and black of the CMYK image; calculating mutual influence among the channels after splitting, setting loss constants of the channels, and calculating fitness; searching a watermark region color block meeting the requirement on the carrier image, and splitting the searched watermark region color block according to an edge algorithm to obtain a channel and a region suitable for embedding a watermark; marking a color block containing a watermark to obtain a carrier image added with a mask; amplitude modulation screening operation is carried out on the carrier image, watermarks are embedded in channels and areas suitable for watermark embedding, and a final halftone target image is generated; the system comprises a color space conversion module, a fitness statistics module, a watermark area selection module, a mask processing module and a watermark embedding and amplitude modulation screening module. The system has the advantages of being good in image quality, flexible in printing resolution, high in watermark imperceptibility and the like.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of digital image watermarking, and in particular to an optimization method and system for adaptive amplitude modulation screening digital watermarking. Background Art

[0002] In recent years, with the widespread popularity of digital images and the rapid development of network technology, the dissemination and sharing of a large amount of digital content has become an important part of daily life. However, this convenient digital environment has also brought about a series of problems such as copyright protection and content security. In this context, digital watermarking technology came into being. As an effective means of intellectual property protection, it has gradually attracted people's attention and attention, and has become a research hotspot in the field of digital content security.

[0003] Halftone images are binary images obtained by screening continuous tone images, and are usually used in the printing field, such as newspapers and magazines. Halftone, also known as grayscale, refers to the tone of the picture expressed by the size or density of dots. The basic element of halftone images is the exposure point, and there are only two states, colored and colorless, which can be directly used for printing or printing. Minimizing the damage to the carrier as much as possible without affecting the value of the carrier is the main research goal of the current halftone image digital watermark hiding method.

[0004] Among the halftone image digital watermark hiding technologies, the halftone image watermark hiding technology based on amplitude modulation screening has always been a hot research direction at home and abroad. Amplitude modulation screening technology uses the size of the dot area to reproduce the gradation changes of the continuous tone image. At present, there are three main digital image watermark hiding technologies based on amplitude modulation screening: first, dot offset watermark hiding technology, which offsets specific dots at a specified angle and distance to achieve the hiding of watermark information. During detection, the watermark information is reproduced by grating coverage using the moiré effect; second, micro-structure dot watermark hiding technology, which uses special structure dots for digital screening, offsets the dots at the specified position to achieve watermark hiding, and the grating composed of special dots is covered on the image to be detected to achieve watermark extraction; third, artistic dot chaotic scrambling information hiding technology, which uses special structure dots to replace the specified dots in the halftone image to complete watermark hiding. During detection, it is only necessary to find and record the position of the special dots to achieve watermark detection.

[0005] In digital image watermarking technology, the selection of watermark area needs to comprehensively consider factors such as the characteristics of the image, the requirements of the application scenario, and the performance of the watermark algorithm to ensure the effective embedding and extraction of the watermark. The selection of the watermark area directly affects the reliability, stability and security of the watermark. The correct selection of the watermark area can maximize the concealment of the watermark, reduce the detectability of the watermark, and protect the watermark from attack and tampering. Therefore, it is crucial to find the appropriate watermark embedding area.

[0006] The patent application document with publication number CN108921768A discloses a grating orientation method for improving digital dot generation. By analyzing the color conditions and histogram conditions of the carrier image, the embedding channel is selected, and then different screening methods are selected according to different color values. The half-reversal method is used for the darker area, and finally the watermark image is embedded to obtain the halftone target image. This scheme achieves a better hiding effect for mid-tone and dark-tone images, but the scheme does not deeply analyze the relationship between the four color channels. It only selects the appropriate channel based on the analysis of the image color and histogram. The appropriate embedding position is not found, and the hiding effect is not optimal. At the same time, the watermark embedding area and the non-embedded area are not strictly distinguished, resulting in the target image after the watermark is embedded not achieving the best clarity effect.

[0007] The patent application document with publication number CN103996166A discloses a holographic anti-counterfeiting method based on dot shape, which generates a halftone image by using the minimum threshold screening technology on the original digital image, transforms the binary watermark image to generate Fourier transform encrypted holographic data, splices the data and binarizes it to generate a new data matrix, and then embeds the watermark into the generated halftone image according to the generated data matrix, and finally obtains the target image containing anti-counterfeiting information. This scheme realizes the encryption of the watermark image and ensures the concealment and security of the watermark embedding, but this scheme directly guides the watermark embedding through the binary data matrix of the watermark image without excessive processing, resulting in poor clarity of the target image after the watermark is embedded.

[0008] The problems and defects of the above scheme are mainly manifested in two aspects: first, the selection of the watermark area does not fully take into account the four-color channel characteristics of the printing field, resulting in the selected watermark embedding area not always being the most suitable, resulting in poor hiding effect and poor watermark extraction effect; second, after the watermark is embedded, in order to ensure the watermark hiding effect, the image usually uses a lower screen line number, which leads to the problem of low image clarity. Summary of the invention

[0009] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide an optimization method and system for adaptive amplitude modulation screening digital watermarking, which obtains the adaptability of the carrier image by analyzing the CMYK four-color channels of the image, thereby solving the problem of poor watermark area selection; by generating a mask image and marking the watermarked color block, the problem of low clarity of the part without embedded watermark is solved; and thus, high-quality and clear halftone target image generation can be achieved in different environments.

[0010] In order to achieve the above object, the technical solution adopted by the present invention is:

[0011] An optimization method for adaptive amplitude modulation screening digital watermarking includes the following steps:

[0012] Step 1, converting the carrier image color space: converting the input RGB image into a CMYK image suitable for printing, and splitting the cyan, magenta, yellow, and black channels of the CMYK image;

[0013] Step 2, statistically analyzing the fitness of the carrier image: calculating the mutual influence between the channels after the split in step 1, setting the loss constant of each channel, and calculating the fitness based on the obtained calculation results and the set loss constant;

[0014] Step 3, find the channel and area suitable for embedding watermark: Based on the fitness obtained in step 2, search for the watermark area color block that meets the requirements on the carrier image, that is, under the premise that the shape and area of ​​the watermark area color block are equal to the watermark image, the fitness mean of the color block is the highest, and the color block of the found watermark area is split according to the edge algorithm to obtain the channel and area suitable for embedding watermark;

[0015] Step 4, adding a mask to the carrier image: generating a mask according to the channel and area suitable for embedding the watermark obtained in step 3, and marking the color block containing the watermark to obtain the carrier image with the mask added;

[0016] Step 5, watermark embedding and AM screening: Based on the carrier image with mask added obtained in step 4, AM screening operation is performed on the carrier image, and the watermark is embedded in the channel and area suitable for embedding watermark found in step 3 to generate the final halftone target image.

[0017] In the carrier image color space conversion operation of step 1, the carrier image color space conversion operation is skipped by inputting a CMYK image, and the four color channels of cyan, magenta, yellow and black of the input CMYK image are split.

[0018] The specific method of step 2 is:

[0019] Traverse the carrier image pixel by pixel, and calculate the mutual influence of each pixel in each channel of the four color channels split in step 1 between each layer. The expression is:

[0020]

[0021] Among them, i is the pixel value of the current traversed pixel, c1, c2, c3, c4 are constants, ES, EL, EB are the same layer influence, high layer influence, and bottom layer influence respectively;

[0022] According to different requirements, different loss constants are set for different channels. Finally, according to the mutual influence between each channel, the final fitness of each pixel in each channel of the four-color channel is calculated. The expression is:

[0023] E=min(((255-|i-c5|×c6×255)×LC×ES×EL×EB), 255)

[0024] Among them, i is the pixel value of the current traversed pixel, c5 and c6 are constants, LC is the loss constant, ES, EL, and EB are the same-layer influence, high-layer influence, and bottom-layer influence respectively, and E is the carrier image fitness.

[0025] In step 3, according to the size and shape of the watermark image, the carrier image is traversed pixel by pixel to find the channel and area suitable for embedding the watermark.

[0026] In step 4, when a mask is added to the carrier image, the watermark area and the non-watermark area are marked.

[0027] In the step 5, the watermark is embedded in the AM screening, and the rational tangent screening method is used to generate the final halftone target image.

[0028] The present invention provides an optimization system for adaptive amplitude modulation screening digital watermarking, comprising:

[0029] The color space conversion module is used to convert the input RGB image into a CMYK image suitable for printing, and to split the cyan, magenta, yellow and black channels of the CMYK image;

[0030] The fitness statistics module is used to calculate the mutual influence between the split channels, set the loss constant of each channel, and calculate the fitness based on the calculated results and the set loss constant;

[0031] The watermark area selection module is used to search for watermark area color blocks that meet the requirements on the carrier image according to the fitness calculated by the fitness statistics module, that is, under the premise that the shape and area of ​​the watermark area color blocks are equal to the watermark image, the fitness mean of the color blocks is the highest. The color blocks of the watermark area found are split according to the edge algorithm to obtain channels and areas suitable for embedding watermarks;

[0032] The mask processing module is used to generate a mask according to the appropriate watermark embedding area selected by the watermark area selection module, and mark the color block containing the watermark to obtain a carrier image with the mask added;

[0033] The watermark embedding and AM screening module performs AM screening on the carrier image based on the masked carrier image obtained by the mask processing module, embeds the watermark in the channel and area suitable for watermark embedding found by the watermark area selection module, generates the final halftone target image, and makes a full-color print.

[0034] The color space conversion module comprises:

[0035] A carrier image reading module, used for reading a carrier image from a computer;

[0036] A CMYK generation module, used for converting a carrier image in an RGB color space into a carrier image in a CMYK color space;

[0037] The printing plate splitting module is used to split the carrier image into four CMYK monochrome plates;

[0038] The mask processing module comprises:

[0039] A mask generation module is used to generate a mask corresponding to the found watermark embedding area;

[0040] A mask marking module is used to apply the generated mask to the carrier image to mark the watermark area and keep the clarity of the non-watermark area unchanged;

[0041] The watermark embedding and amplitude modulation screening module includes:

[0042] A threshold matrix generation module is used to generate a corresponding threshold matrix according to the set dot shape;

[0043] A screening generation module is used to perform a screening operation on a carrier image using a threshold matrix to generate a screened carrier image, and then embed a watermark image to generate a final halftone target image;

[0044] A file storage module, used for storing the halftone target image to a user's designated location;

[0045] The plate making and printing module is used to make plates and print high-definition full-color prints with invisible watermarks.

[0046] The fitness statistics module includes:

[0047] Channel influence calculation module, used to calculate the mutual influence between each channel of the four color channels pixel by pixel;

[0048] The overall influence calculation module is used to calculate the fitness based on the mutual influence of channels and the loss constant.

[0049] The watermark area selection module comprises:

[0050] A watermark image reading module is used to select and read a watermark image from a computer file system;

[0051] The area search module is used to find the area suitable for embedding watermark according to the shape, area and fitness of the watermark image;

[0052] The color block splitting module is used to split the appropriate watermark area from the carrier image according to the edge algorithm.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. The present invention analyzes the mutual influence of each layer of the four-color channel on watermark hiding, and combines the affinity of the image to each channel to calculate the fitness of each pixel of each channel of the carrier image, thereby finding the most suitable area for embedding watermarks, which can greatly improve the concealment of watermarks. This effectively solves the problem of poor watermark hiding effect caused by the inability to find a suitable area for embedding watermarks in the carrier image.

[0055] 2. The present invention uses an edge algorithm to select the watermark area and generates a mask to mark the watermark area and the non-watermark area, ensuring that the dot arrangement of the non-watermark part remains unchanged, so that the non-watermark area can still maintain good clarity. This effectively solves the problem of low clarity of the target image after embedding the watermark, and makes the final target image have a good visual effect.

[0056] In summary, the present invention finds a suitable area for embedding watermarks through the digital watermark adaptive optimization method and the generation of mask marks, and allows the non-embedded watermark area to maintain a clearer image quality, thereby ensuring the rationality of the watermark embedding area and improving the concealment of the watermark. It has the advantages of good image quality, flexible printing resolution, and high watermark imperceptibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a flow chart of the method of the present invention.

[0058] Figure 2 It is a system structure diagram of the present invention.

[0059] Figure 3 It is the fitness image of four channels generated by the present invention.

[0060] Figure 4 It is a mask effect image generated by the present invention.

[0061] Figure 5 The target image generated by the present invention is compared with the target image without adding a mask at the same position; wherein, Figure 5 (a) is the target image with mask added, Figure 5 (b) is the target image without mask.

[0062] Figure 6 The target image generated by the present invention is compared with the target image with the watermark embedded in other positions; wherein, Figure 6 (a) is the target image with watermark embedded in the appropriate position, Figure 6 (b) is the target image where the watermark is embedded in an inappropriate position.

[0063] Figure 7It is the local effect of embedding a single channel of watermark in the target image in the non-suitable watermark area.

[0064] Figure 8 It is a watermark extraction effect image of a target image generated by the present invention, in which a watermark image is embedded in an area suitable for embedding a watermark and an area not suitable for embedding a watermark; wherein, Figure 8 (a) is the target image with watermark embedded in the appropriate position, and the watermark extraction effect image is shown. Figure 8 (b) is the watermark extraction effect image of the target image with the watermark embedded in an inappropriate position. DETAILED DESCRIPTION

[0065] In order to more clearly illustrate the purpose, technical solutions and advantages of the present invention, the following is further described in detail in conjunction with specific embodiments. It should be noted that the specific embodiments described below are only used to illustrate the present invention, rather than to limit the present invention.

[0066] An optimization method for adaptive amplitude modulation screening digital watermarking, such as Figure 1 In this embodiment, the watermark image is embedded into a single color channel of the carrier image, and the watermark image is extracted respectively by the grating extraction method. The specific method is as follows:

[0067] An optimization method for adaptive amplitude modulation screening digital watermarking includes the following steps:

[0068] Step 1, convert the carrier image color space: convert the input RGB image into a CMYK image suitable for printing, and split the cyan, magenta, yellow and black channels of the CMYK image; specifically:

[0069] Read the carrier image in RGB color space and convert it to CMYK color space suitable for four-color printing, and separate the channels. In this process, sRGB2014.icc is used as the RGB profile, and JapanColor2001Coated.icc is used as the CMYK profile.

[0070] Step 2, statistical carrier image fitness: calculate the mutual influence between the channels after splitting in step 1 and set the loss constant of each channel to calculate the fitness; specifically:

[0071] Step 2.1, since cyan and magenta belong to the same level channel, yellow is a lower level channel, and black is a higher level channel, the high-level influence, same-level influence, and low-level influence of each pixel in each channel are calculated based on the grayscale images of the four channels separated in step 1. The calculation formula is:

[0072]

[0073] Among them, i is the pixel value of the current traversed pixel, c1, c2, c3, c4 are constants, ES, EL, EB are the same layer influence, high layer influence, and bottom layer influence respectively;

[0074] Step 2.2, first calculate the single channel fitness, assuming that the more the single channel pixel value deviates from c5, the lower its single channel fitness. Then, based on the mutual influence calculated in step 2.1 and the set loss constant of the current image on the four-layer channel, calculate the overall influence as the fitness. The calculation formula is:

[0075] E=min(((255-|i-c5|×c6×255)×LC×ES×EL×EB), 255)

[0076] Among them, i is the pixel value of the current traversed pixel, c5 and c6 are constants, ES, EL, and EB are the same-layer influence, high-layer influence, and bottom-layer influence, respectively, LC is the loss constant of each channel, and E is the fitness of the carrier image. The calculation results are as follows: Figure 3 As shown, the white area is the area suitable for embedding a watermark.

[0077] Step 3: Find the channel and area suitable for embedding watermarks:

[0078] Step 3.1, read the watermark image, traverse the carrier image pixel by pixel according to the size and shape of the watermark image, and find the area with the highest fitness obtained in step 2.2, that is, search for the area suitable for embedding the watermark;

[0079] Step 3.2, watermark area determination operation: the area suitable for embedding watermarks found in step 3.1 is split out using an edge algorithm to facilitate marking and distinguishing watermark areas from non-watermark areas.

[0080] Step 4, add a mask to the carrier image: According to the marking in step 3.2, generate a mask of the carrier image, where the non-watermark area is represented by black, and the searched area suitable for embedding watermarks is marked with white, such as Figure 4 As shown, the mask is then marked on the carrier image. In subsequent operations, the white area is used as the watermark area. In order to ensure the hidden effect of the watermark, a lower screen count can be used. The black area is used as the non-watermark area. In order to ensure the clarity of the image, a high screen count is used.

[0081] Step 5, watermark embedding and AM screening: Based on the carrier image with mask added obtained in step 4, generate corresponding dot threshold matrices for the watermark area and non-watermark area, use conventional AM screening algorithm, perform AM screening on all dots, and embed the watermark image on the carrier image according to the area suitable for watermark embedding found in step 3.2 to generate the final target image, as shown in FIG. Figure 5 (a) shown.

[0082] The target image generated in step 5 is stored in a pre-set folder and input into the printing equipment to realize the final plate-making and printing to obtain a high-resolution full-color print.

[0083] At this stage, the watermark image is already hidden in the full-color print, and the hidden watermark information can be extracted through corresponding detection methods. These detection methods include machine scanning, optical reflection and special optical tools. In this example, the common grating extraction method is used. Its principle is to use the spectral effect of the grating sheet to detect the hidden information. The extraction effect is as follows: Figure 8 (a) shown.

[0084] An optimization system for adaptive amplitude modulation screening digital watermarking, such as Figure 2 The system includes: a color space conversion module M201, a fitness statistics module M202, a watermark region selection module M203, a mask processing module M204, and a watermark embedding and amplitude modulation screening module M205.

[0085] The color space conversion module M201 has the main function of reading the carrier image and converting it into a carrier image of a CMYK color space suitable for printing, and splitting the cyan, magenta, yellow and black channels of the CMYK image so that subsequent modules can process it.

[0086] The color space conversion module M201 specifically includes: a carrier image reading module M2011, used to select and read a carrier image from a computer file system; a CMYK generation module M2012, used to convert a carrier image in an RGB color space into a carrier image in a CMYK color space; and a printing plate splitting module M2013, used to split the carrier image into four CMYK monochrome plates.

[0087] The fitness statistics module M202 has the main function of calculating the fitness of each pixel point in each channel of the carrier image to identify the area suitable for watermark embedding.

[0088] The fitness statistics module M202 specifically includes: a channel influence calculation module M2021, which is used to calculate the mutual influence ES, EL, and EB between each channel of the four-color channels pixel by pixel; an overall influence calculation module M2022, which is used to calculate the fitness based on the single-channel influence, the influence between channels, and the loss constant to generate an analysis result.

[0089] The watermark area selection module M203 has the main function of finding and marking the area of ​​the carrier image suitable for embedding watermarks according to the fitness.

[0090] The watermark area selection module M203 specifically includes: a watermark image reading module M2031, used to select and read the watermark image from the computer file system; an area search module M2032, used to search for an area in the carrier image that can accommodate the watermark image and has a high adaptability and is suitable for watermark embedding; a color block splitting module M2033, used to mark the searched watermark embedding area.

[0091] The mask processing module M204 has the main function of generating a mask and marking the carrier image so that the clarity of the non-watermark area will not be affected.

[0092] The mask processing module M204 specifically includes: a mask generating module M2041, used to generate a mask corresponding to the found watermark embedding area; a mask marking module M2042, used to mark the generated mask onto the carrier image, marking the watermark area so that the clarity of the non-watermark area remains unchanged.

[0093] The main function of the watermark embedding and amplitude modulation screening module M205 is to generate the final halftone target image and print out color prints.

[0094] The watermark embedding and amplitude modulation screening module M205 specifically includes: a threshold matrix generation module M2051: used to generate a corresponding dot threshold matrix according to the set dot shape; a screening generation module M2052: used to use the dot threshold matrix to perform a screening operation on the carrier image, generate a screened carrier image, and then embed the watermark image to generate a final halftone target image; a file storage module M2053: used to store the halftone target image to a user-specified location; a plate making and printing module M2054: used to produce and print high-definition full-color prints containing invisible watermarks.

[0095] The present invention provides an optimization method and system for adaptive amplitude modulation screening digital watermarking, the method includes finding the most suitable area for embedding watermarks by counting the fitness of each channel of the carrier image, solving the problem of poor watermark concealment caused by unreasonable watermark embedding position; marking the watermark area of ​​the carrier image with a mask, solving the problem of low image clarity after embedding the watermark. The system includes a color space conversion module, a fitness statistics module, a watermark area selection module, a mask processing module, and a watermark embedding and amplitude modulation screening module. The optimization method and system for adaptive amplitude modulation screening digital watermarking proposed by the present invention can obtain a clearer target image under the same number of extracted grating lines, while ensuring the rationality of the watermark embedding area, improving the concealment of the watermark, and overall having the characteristics of good image quality, flexible printing resolution, and high watermark imperceptibility.

[0096] In order to verify the effect of the adaptive AM screening digital watermarking algorithm on the target image quality and watermark trace concealment, the embodiment of the present invention designed a series of subjective and objective evaluation experiments. For image clarity, the present invention compares the halftone image quality of the present invention with and without adding a mask based on the same number of extracted raster lines. Figure 5 shown.

[0097] from Figure 5 It can be seen that there is a significant difference in image clarity between the halftone images with and without mask. Figure 5 The watermark is added in the green area in the upper left corner. Figure 5 (a) The target image has clearer colors except for the green area. Figure 5 (b) shows a more obvious graininess. Therefore, the mask marking algorithm proposed in the present invention performs better in terms of image clarity and can better preserve the original detail information of the image.

[0098] Regarding the concealment of the watermark image, the present invention compares the image quality of the watermark embedded in the embedding area searched by the present invention and the image quality of the watermark embedded in other areas based on the same number of extracted raster lines. Figure 6 As shown in Figure 2, the watermark extraction effect is as follows: Figure 8 shown.

[0099] from Figure 6 It can be seen that there is an obvious difference in watermark concealment whether or not the present invention is applied to find a suitable area for embedding into an image. Figure 6 In (a), the watermark image is hidden in the green area of ​​the cyan channel in the upper left corner. Figure 6 In (b), the watermark image is hidden in the green area of ​​the upper left magenta channel. Figure 6 (a) has a better hiding effect. Figure 7 This is the target image effect of the magenta channel, and you can see the subtle watermark image texture. Figure 8 It can be seen that after the watermark area selection algorithm, the watermark is embedded in the appropriate embedding area, and the watermark extracted by the grating has a higher quality. In contrast, the watermark embedding algorithm based on fitness proposed in the present invention performs better in terms of watermark image concealment and grating extraction effect, and can better hide the watermark content.

[0100] Next, use objective evaluation indicators to evaluate the quality of the algorithm. SSIM (Structural Similarity Index) is a similarity measurement method for image structural information. It takes into account the differences in the perception of image structure by the human visual system to evaluate and calculate the similarity of images. The SSIM measurement index includes multiple aspects such as brightness, structure, and contrast. Therefore, compared with PSNR, SSIM is closer to the evaluation criteria of human vision. Use SSIM to evaluate the quality of the target halftone image, and its calculation formula is:

[0101]

[0102] In the formula, μ x represents the mean value of x, σ x represents the standard deviation of x, σ xy Represents the covariance of x and y. C3 and C4 are positive constants to prevent the structural similarity result from being 0.

[0103] The present invention uses ten images in the DIV2K_train_HR super-resolution reconstruction dataset and our test images for comparative experiments. The SSIM index of the original image is compared using the watermark embedding algorithm in the appropriate position and the watermark embedding algorithm in the inappropriate position respectively; and in order to verify that the algorithm has a better effect on watermark extraction, the SSIM index after adding the grating is also compared with the original image; finally, it can be seen that the SSIM value of the target image after adding the mask is also higher than the target image value without adding the mask. The results are shown in Table 1, and the best result of each image has been marked in bold.

[0104] Table 1 Target image quality evaluation

[0105]

[0106]

[0107] The above whole process includes the loading of carrier image and watermark image, color space conversion, fitness calculation, suitable embedding area search, mask generation and marking, amplitude modulation screening and target image generation.

[0108] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When the use is implemented in whole or in part in the form of a computer program product, the computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL) or wireless (e.g., infrared, wireless, microwave, etc.) mode) to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk SolidState Disk (SSD)), etc.

[0109] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. An optimization method for adaptive amplitude modulation screening digital watermarking, characterized in that: The following steps are involved: Step 1, converting the carrier image color space: converting the input RGB image into a CMYK image suitable for printing, and splitting the cyan, magenta, yellow, and black channels of the CMYK image; Step 2, statistically analyzing the fitness of the carrier image: calculating the mutual influence between the channels after the split in step 1, setting the loss constant of each channel, and calculating the fitness based on the obtained calculation results and the set loss constant; Step 3, find the channel and area suitable for embedding watermark: Based on the fitness obtained in step 2, search for the watermark area color block that meets the requirements on the carrier image, that is, under the premise that the shape and area of ​​the watermark area color block are equal to the watermark image, the fitness mean of the color block is the highest, and the color block of the found watermark area is split according to the edge algorithm to obtain the channel and area suitable for embedding watermark; Step 4, adding a mask to the carrier image: generating a mask according to the channel and area suitable for embedding the watermark obtained in step 3, and marking the color block containing the watermark to obtain the carrier image with the mask added; Step 5, watermark embedding and AM screening: Based on the carrier image with mask added obtained in step 4, AM screening operation is performed on the carrier image, and the watermark is embedded in the channel and area suitable for embedding watermark found in step 3 to generate the final halftone target image.

2. The optimization method of adaptive amplitude modulation screening digital watermark according to claim 1, characterized in that: In the carrier image color space conversion operation of step 1, the carrier image color space conversion operation is skipped by inputting a CMYK image, and the four color channels of cyan, magenta, yellow and black of the input CMYK image are split.

3. The optimization method of adaptive amplitude modulation screening digital watermark according to claim 1, characterized in that: The specific method of step 2 is: Traverse the carrier image pixel by pixel, and calculate the mutual influence of each pixel in each channel of the four color channels split in step 1 between each layer. The expression is: Among them, i is the pixel value of the current traversed pixel, c1, c2, c3, c4 are constants, ES, EL, EB are the same layer influence, high layer influence, and bottom layer influence respectively; According to different requirements, different loss constants are set for different channels. Finally, according to the mutual influence between each channel, the final fitness of each pixel point in each channel of the four-color channel is calculated. The expression is: E=min(((255-|i-c5|×c6×255)×LC×ES×EL×EB),255) Among them, i is the pixel value of the current traversed pixel, c5 and c6 are constants, LC is the loss constant, ES, EL, and EB are the same-layer influence, high-layer influence, and bottom-layer influence respectively, and E is the carrier image fitness.

4. The optimization method of adaptive amplitude modulation screening digital watermark according to claim 1, characterized in that: In step 3, according to the size and shape of the watermark image, the carrier image is traversed pixel by pixel to find the channel and area suitable for embedding the watermark.

5. The optimization method of adaptive amplitude modulation screening digital watermark according to claim 1, characterized in that: In step 4, when a mask is added to the carrier image, the watermark area and the non-watermark area are marked.

6. The optimization method of adaptive amplitude modulation screening digital watermark according to claim 1, characterized in that: In the step 5, the watermark is embedded in the AM screening, and the rational tangent screening method is used to generate the final halftone target image.

7. An optimization system for adaptive amplitude modulation screening digital watermarking based on the method according to any one of claims 1 to 6, characterized in that: include: The color space conversion module is used to convert the input RGB image into a CMYK image suitable for printing, and to split the cyan, magenta, yellow and black channels of the CMYK image; The fitness statistics module is used to calculate the mutual influence between the split channels, set the loss constant of each channel, and calculate the fitness based on the calculated results and the set loss constant; The watermark area selection module is used to search for watermark area color blocks that meet the requirements on the carrier image according to the fitness calculated by the fitness statistics module, that is, under the premise that the shape and area of ​​the watermark area color blocks are equal to the watermark image, the fitness mean of the color blocks is the highest. The color blocks of the watermark area found are split according to the edge algorithm to obtain channels and areas suitable for embedding watermarks; The mask processing module is used to generate a mask according to the appropriate watermark embedding area selected by the watermark area selection module, and mark the color block containing the watermark to obtain a carrier image with the mask added; The watermark embedding and AM screening module performs AM screening on the carrier image based on the masked carrier image obtained by the mask processing module, embeds the watermark in the channel and area suitable for watermark embedding found by the watermark area selection module, generates the final halftone target image, and makes a full-color print.

8. The optimization system of adaptive amplitude modulation screening digital watermarking according to claim 7, characterized in that: The color space conversion module comprises: A carrier image reading module, used for reading a carrier image from a computer; A CMYK generation module, used for converting a carrier image in an RGB color space into a carrier image in a CMYK color space; The printing plate splitting module is used to split the carrier image into four CMYK monochrome plates; The mask processing module comprises: A mask generation module is used to generate a mask corresponding to the found watermark embedding area; A mask marking module is used to apply the generated mask to the carrier image to mark the watermark area and keep the clarity of the non-watermark area unchanged; The watermark embedding and amplitude modulation screening module includes: A threshold matrix generation module is used to generate a corresponding threshold matrix according to the set dot shape; A screening generation module is used to perform a screening operation on a carrier image using a threshold matrix to generate a screened carrier image, and then embed a watermark image to generate a final halftone target image; A file storage module, used for storing the halftone target image to a user's designated location; The plate making and printing module is used to make plates and print high-definition full-color prints with invisible watermarks.

9. The optimization system of adaptive amplitude modulation screening digital watermarking according to claim 7, characterized in that: The fitness statistics module includes: Channel influence calculation module, used to calculate the mutual influence between each channel of the four color channels pixel by pixel; The overall influence calculation module is used to calculate the fitness based on the mutual influence of channels and the loss constant.

10. The optimization system of adaptive amplitude modulation screening digital watermark according to claim 7, characterized in that: The watermark area selection module comprises: A watermark image reading module is used to select and read a watermark image from a computer file system; The area search module is used to find the area suitable for embedding watermark according to the shape, area and fitness of the watermark image; The color block splitting module is used to split the appropriate watermark area from the carrier image according to the edge algorithm.

Citation Information

Patent Citations

  • Adaptive visible watermark adding method of digital image of mouse picking-up and adding position on the basis of .NET

    CN104331855A

  • Data hiding for spot colors in product packaging

    CN106716483A

  • Method for setting anti-counterfeiting information in host channel of digital image

    CN108898543A

  • Grating anti-counterfeiting method for improved digital dot generation

    CN108921768A

  • Halftone copying fragile watermarking technical method based on moire effect

    CN110163789A

Cited By

  • Generative AI output-based domestic credible watermark embedding and tracing method and system

    CN120832661A

  • Method and system for localization of trusted watermark embedding and tracing of generative ai output

    CN120832661B