Image watermark processing method and device, equipment, medium and product

Abstract

The invention discloses an image watermark processing method, device and equipment, a medium and a product, and relates to the technical field of computer data security. The method comprises the following steps: acquiring a to-be-embedded image, watermark text information and a quantization step size; performing channel separation on the to-be-embedded image, and determining a first channel image, a second channel image and a third channel image; the first channel image is an R channel image; the second channel image is a G channel image; the third channel image is a B channel image; performing discrete cosine transform on the third channel image to obtain a first DCT coefficient matrix; performing binary conversion on the watermark text information to obtain watermark coding information; embedding the watermark coding information into an intermediate frequency region of the first DCT coefficient matrix to obtain a second DCT coefficient matrix; performing inverse discrete cosine transform on the second DCT coefficient matrix to obtain a fourth channel image; and obtaining a target image based on the fourth channel image, the first channel image and the second channel image. Therefore, the watermark processing efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of computer data security technology, and in particular to an image watermarking processing method, apparatus, device, medium and product. Background Technology

[0002] Traditional spatial watermarking methods are computationally simple but lack robustness, making them highly susceptible to corruption by common image processing operations such as compression and filtering. Frequency domain watermarking, particularly methods based on Discrete Cosine Transform (DCT), achieves a good balance between invisibility and robustness by embedding watermark information into key frequency coefficients of the image. However, traditional DCT watermarking schemes typically require frequency domain transformation and calculation for all color channels of a full-size image, resulting in high computational complexity and significant processing latency, making them unsuitable for real-time or high-volume image processing. Summary of the Invention

[0003] This invention provides an image watermarking processing method, apparatus, device, medium, and product to solve the problem of low watermarking processing efficiency.

[0004] According to one aspect of the present invention, an image watermarking processing method is provided, comprising: Obtain the image to be embedded, the watermark text information, and the quantization step size; The image to be embedded is subjected to channel separation to determine a first channel image, a second channel image, and a third channel image; the first channel image is an R channel image; the second channel image is a G channel image; and the third channel image is a B channel image. Perform a discrete cosine transform on the third channel image to obtain the first DCT coefficient matrix; The watermark text information is converted into binary to obtain the watermark encoding information; The watermark encoding information is embedded into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix; Perform inverse discrete cosine transform on the second DCT coefficient matrix to obtain the fourth channel image; The target image is obtained based on the fourth channel image, the first channel image, and the second channel image.

[0005] According to another aspect of the present invention, an image watermarking processing apparatus is provided, comprising: The first data acquisition module is used to acquire the image to be embedded, watermark text information, and quantization step size; The first channel determination module is used to perform channel separation on the image to be embedded, and determine a first channel image, a second channel image, and a third channel image; the first channel image is an R channel image; the second channel image is a G channel image; and the third channel image is a B channel image. The first matrix acquisition module is used to perform discrete cosine transform on the third channel image to obtain the first DCT coefficient matrix; The watermark encoding acquisition module is used to perform binary conversion on the watermark text information to obtain watermark encoding information; The second matrix acquisition module is used to embed the watermark encoding information into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix. The fourth channel acquisition module is used to perform inverse discrete cosine transform on the second DCT coefficient matrix to obtain the fourth channel image; The target image acquisition module is used to obtain a target image based on the fourth channel image, the first channel image, and the second channel image.

[0006] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the image watermarking processing method according to any embodiment of the present invention.

[0007] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the image watermarking processing method according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer program product is provided, comprising a computer program / instructions that, when executed by a processor, implement the image watermarking processing method as described in any embodiment of the present invention.

[0009] This invention significantly reduces the computational load required for DCT transformation and embedding operations by introducing image scaling and single-channel processing strategies, making the watermark embedding process more efficient. By selecting the blue channel, which is less sensitive to the human eye, for embedding, it ensures that the introduction of the watermark will not have a significant impact on the subjective visual quality of the image. Embedding the watermark in the mid-frequency region enables the watermark to resist a certain degree of common attacks such as compression and filtering. It can effectively improve the efficiency of watermark processing, take into account the invisibility and robustness of watermark processing, and meet the needs of real-time or large-scale image processing, such as batch watermarking of cloud photo albums and real-time video stream embedding.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of an image watermarking processing method provided in an embodiment of the present invention; Figure 2 This is a flowchart of another image watermarking method provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an image watermarking processing device provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements the image watermarking processing method of this invention. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0015] Furthermore, it should be noted that the information collected in the technical solution of this invention is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data all comply with the relevant laws, regulations and standards of relevant countries and regions, necessary confidentiality measures have been taken, and public order and good morals are not violated. Corresponding operation entry points are provided for users to choose to authorize or refuse.

[0016] Figure 1 This is a flowchart of an image watermarking processing method provided by an embodiment of the present invention. This embodiment is applicable to situations involving watermark processing. The method can be executed by an image watermarking processing device, which can be implemented in hardware and / or software. This device can be configured in an electronic device with corresponding data processing capabilities, such as a server. Figure 1 As shown, the method includes: S110: Obtain the image to be embedded, watermark text information, and quantization step size.

[0017] In this context, the image to be embedded is the image into which the watermark is to be embedded. The watermark text information refers to the content of the watermark. The quantization step size is a pre-set adjustment parameter. The quantization step size is used to determine the embedding strength and robustness of the watermark.

[0018] Specifically, it obtains the image to be embedded, watermark text information (such as copyright information, certification mark information, etc.), and quantization step size.

[0019] S120. Perform channel separation on the image to be embedded to determine the first channel image, the second channel image, and the third channel image; the first channel image is the R channel image; the second channel image is the G channel image; and the third channel image is the B channel image.

[0020] The image to be embedded is a color image. Channel separation refers to the process of splitting a color image into three independent single-channel grayscale images according to its color components. The color components include the R (Red) component, the G (Green) component, and the B (Blue) component. Each component has a value ranging from 0 to 255, representing the intensity of the corresponding color. The first channel image is a single-channel grayscale image of the Red component. The second channel image is a single-channel grayscale image of the Green component. The third channel image is a single-channel grayscale image of the Blue component.

[0021] Specifically, each pixel of the image to be embedded is composed of three color components: red (R), green (G), and blue (B). By extracting the dimensional data of the corresponding color components in the image, the three channels are independently split to obtain the first channel image, the second channel image, and the third channel image. The first channel image is the R channel image, the second channel image is the G channel image, and the third channel image is the B channel image.

[0022] S130. Perform discrete cosine transform on the third channel image to obtain the first DCT coefficient matrix.

[0023] The Discrete Cosine Transform (DCT) utilizes the orthogonality of the cosine function to transform a two-dimensional pixel matrix represented by pixel values ​​(0-255) in the spatial domain into a two-dimensional frequency domain coefficient matrix represented by the intensity of frequency components in the frequency domain. The two-dimensional frequency domain coefficient matrix is ​​the DCT coefficient matrix, where each element is called a DCT coefficient, and the value of the coefficient represents the intensity of the corresponding frequency component. The first DCT coefficient matrix is ​​obtained by performing a Discrete Cosine Transform on the third channel image (a single-channel grayscale image of the blue component).

[0024] Specifically, the Human Vision System (HVS) is least sensitive to changes in the brightness of the blue channel, and modifying the coefficients of the B channel image has the least negative impact on the image's visual quality. By performing frequency domain transformation on the third channel image, the image is converted from the spatial domain to the frequency domain, resulting in the first DCT coefficient matrix, which lays the image foundation for subsequent watermark embedding.

[0025] S140. Perform binary conversion on the watermark text information to obtain the watermark encoding information.

[0026] The watermark encoding information refers to the binary sequence obtained after converting the watermark text information according to a preset character encoding standard. The watermark encoding information consists of 0 and 1 bit symbols arranged in a fixed order. Each watermark text corresponds to a unique watermark encoding information.

[0027] Specifically, each character (including letters, numbers, symbols, etc.) in the watermark text information is converted into a corresponding fixed-length (8-bit) binary number according to a preset character encoding standard (such as ASCII code). The binary numbers corresponding to all characters are then concatenated according to the character order of the watermark text information to form a continuous binary sequence, which is the watermark encoding information.

[0028] S150. Embed the watermark encoding information into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix.

[0029] The mid-frequency region is the area in the DCT coefficient matrix where both row and column indices fall within a predetermined middle range. In the DCT coefficient matrix, the elements in the upper left corner are low-frequency coefficients with larger values, corresponding to the low-frequency region of the matrix, which represents the fundamental energy of the image. From the upper left to the lower right corner, the frequencies of the coefficients gradually increase, while their values ​​decrease. The elements in the lower right corner are high-frequency coefficients with smaller values, corresponding to the high-frequency region of the matrix, which represents subtle details or noise in the image and is easily filtered out. The elements in the middle region of the matrix are mid-frequency coefficients (such as the area with row or column indices 3-5), with values ​​between those of the low-frequency and high-frequency coefficients. The second DCT coefficient matrix contains watermark encoding information.

[0030] Specifically, after embedding the watermark encoding information into the mid-frequency region of the first DC coefficient matrix, a two-dimensional DCT coefficient matrix is ​​obtained that retains the original frequency domain characteristics. Only some coefficients in the mid-frequency region are modified due to the watermark embedding, while the coefficients in the low-frequency and high-frequency regions remain unchanged, ensuring the integrity of the image visual quality and frequency domain structure. The second DCT coefficient matrix inherits the frequency characteristics of the first DCT coefficient matrix and also implicitly contains the watermark encoding information, achieving the concealment of watermark information and lossless preservation of image visual quality.

[0031] S160. Perform inverse discrete cosine transform on the second DCT coefficient matrix to obtain the fourth channel image.

[0032] The inverse discrete cosine transform (DCT) is a linear orthogonal transform that is the inverse of the discrete cosine transform. It utilizes the orthogonality and invertibility of the DCT transform to restore the DCT coefficient matrix in the frequency domain to a two-dimensional image matrix represented by pixel values ​​in the spatial domain. The fourth channel image refers to the single-channel grayscale image in the spatial domain obtained after performing the inverse discrete cosine transform on the second DCT coefficient matrix.

[0033] Specifically, the second DCT coefficient matrix is ​​a watermarked DCT coefficient matrix. By performing an inverse discrete cosine transform on the second DCT coefficient matrix, a watermarked spatial domain single-channel grayscale image is obtained, which is then used as the fourth channel image. The fourth channel image, acting as a "watermarked B channel carrier," carries the watermark information from the second DCT coefficient matrix, providing the foundation for subsequent merging with the first channel image (R channel) and the second channel image (G channel) to form a complete watermarked RGB color image.

[0034] S170. Based on the fourth channel image, the first channel image, and the second channel image, the target image is obtained.

[0035] The target image is an image with an embedded watermark.

[0036] Specifically, based on the first channel image (R channel single-channel grayscale image), the second channel image (G channel single-channel grayscale image), and the fourth channel image (B channel single-channel grayscale image with watermark), a complete RGB color image with watermark is obtained, which is the target image.

[0037] Optionally, channel separation is performed on the image to be embedded to determine a first channel image, a second channel image, and a third channel image, including: reducing the image to be embedded according to a first ratio to obtain a first image; and performing channel separation on the first image to obtain the first channel image, the second channel image, and the third channel image.

[0038] The first ratio is a preset scaling parameter used to resize the image to be embedded. The first image is an RGB color image obtained by scaling down the image to be embedded according to the first ratio.

[0039] Specifically, interpolation algorithms (such as bilinear interpolation and bicubic interpolation to ensure that the image is not significantly distorted after scaling) can be used to reduce the size of the image to be embedded according to the first ratio to obtain the first image. After processing, while retaining the core visual features and color model of the original image, the total number of pixels in the image is reduced, and the computational load of all subsequent pixel-based operations (such as DCT) will also be reduced. Channel separation is performed on the first image to obtain the first channel image, the second channel image, and the third channel image.

[0040] Optionally, obtaining a target image based on the fourth channel image, the first channel image, and the second channel image includes: merging the fourth channel image, the first channel image, and the second channel image to obtain a second image; and enlarging the second image according to a second ratio to obtain the target image; wherein the ratio of the first ratio to the second ratio is the reciprocal of each other.

[0041] Channel merging is the process of combining three independent single-channel grayscale images into a single color image. The second ratio is a preset parameter used to enlarge the second image to the size of the image to be embedded.

[0042] Specifically, the fourth channel image, the first channel image, and the second channel image are merged to obtain the second image; the second image is enlarged according to the second ratio to obtain the target image; the ratio of the first ratio and the second ratio are reciprocals of each other, thus forming a size closed loop of "the image to be embedded is reduced according to the first ratio and enlarged according to the second ratio to obtain the target image" through the inverse relationship between the second ratio and the first ratio, ensuring that there is no difference in image size before and after watermark embedding.

[0043] This invention significantly reduces the computational load required for DCT transformation and embedding operations by introducing image scaling and single-channel processing strategies, making the watermark embedding process more efficient. By selecting the blue channel, which is less sensitive to the human eye, for embedding, it ensures that the introduction of the watermark will not have a significant impact on the subjective visual quality of the image. Embedding the watermark in the mid-frequency region enables the watermark to resist a certain degree of common attacks such as compression and filtering. It can effectively improve the efficiency of watermark processing, take into account the invisibility and robustness of watermark processing, and meet the needs of real-time or large-scale image processing, such as batch watermarking of cloud photo albums and real-time video stream embedding.

[0044] Figure 2 This is a flowchart of another image watermarking processing method provided by an embodiment of the present invention. Based on the above embodiments, this embodiment optimizes the step of "embedding the watermark encoding information into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix," providing an optional implementation scheme. For example... Figure 2 As shown, the method includes: S210, Obtain the image to be embedded, watermark text information, and quantization step size.

[0045] S220. Perform channel separation on the image to be embedded to determine the first channel image, the second channel image, and the third channel image; the first channel image is the R channel image; the second channel image is the G channel image; and the third channel image is the B channel image.

[0046] S230. Perform discrete cosine transform on the third channel image to obtain the first DCT coefficient matrix.

[0047] S240. Perform binary conversion on the watermark text information to obtain the watermark encoding information.

[0048] S250. Embed the watermark encoding information into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix.

[0049] S260. Perform inverse discrete cosine transform on the second DCT coefficient matrix to obtain the fourth channel image.

[0050] S270. Based on the fourth channel image, the first channel image, and the second channel image, the target image is obtained.

[0051] Optionally, embedding watermark encoding information into the intermediate frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix includes: for the intermediate frequency region of the first DCT coefficient matrix, obtaining at least one coefficient and its corresponding coordinate value in the intermediate frequency region; performing position mapping between the coordinate value and the watermark encoding information, and determining the watermark bit to be embedded for the coefficient from the watermark encoding information; using a coefficient quantization formula, quantizing the coefficient based on the quantization step size to determine the initial quantization value corresponding to the coefficient; processing the initial quantization value based on the watermark bit to be embedded and the quantization step size using an even / odd coding rule to obtain the target quantization value; and updating the first DCT coefficient matrix based on at least one target quantization value to obtain the second DCT coefficient matrix.

[0052] In this system, the watermark bit to be embedded is the smallest embedding unit of watermark encoded information. Each coefficient in the intermediate frequency region carries only one watermark bit to be embedded, which is a single binary bit (0 or 1). The coefficient quantization formula is a preset mathematical calculation formula used to quantize the coefficients. The initial quantization value is the value obtained after preliminary quantization of the coefficients. The parity coding rule is a preset operational rule for adjusting the parity of the quantized values ​​based on the value of the watermark bit to be embedded. The target quantization value is the final quantization result of the coefficients.

[0053] Specifically, for the intermediate frequency (IF) region of the first DCT coefficient matrix, at least one coefficient and its corresponding coordinate value are obtained. Based on a preset position mapping rule (e.g., starting from coordinate (5,5) and scanning row by row), a one-to-one mapping relationship is established between the coefficient coordinates and the bits of the watermark encoding information within the IF region. At least one coefficient is traversed, and according to the coordinate value of the current coefficient, the corresponding single binary bit is extracted from a fixed position in the watermark encoding information. This bit is the watermark bit to be embedded for the current coefficient. For example, if the watermark encoding information is "00011100", and the coefficient with coordinate value (3,3) in the IF region corresponds to the 4th bit, then the watermark bit to be embedded for this coefficient is 1. The coefficient quantization formula can be: ; Where quantized_value is the initial quantization value, coefficient is the coefficient, quantization_step is the quantization step size, and round is the rounding function. The round function is used to round the numerical value to the nearest integer. The parity coding rule is used, which uses the initial quantization value and quantization step size to obtain the odd or even attribute of the initial quantization value. Based on the one-to-one correspondence between the odd or even attribute and the "1 or 0" of the watermark bit, the initial quantization value is adjusted to obtain the target quantization value. The parity of the adjusted target quantization value is strictly matched with the watermark bit to be embedded, thereby realizing the hiding of watermark information.

[0054] Optionally, the initial quantization value is processed based on the watermark bit to be embedded corresponding to the coefficient and the quantization step size using parity coding rules to obtain the target quantization value. This includes: determining the quotient based on the initial quantization value and the quantization step size; determining the target quantization value based on the initial quantization value and the quantization step size using parity coding rules if the watermark bit to be embedded corresponding to the coefficient is 1 and the quotient is even, or the watermark bit to be embedded corresponding to the coefficient is 0 and the quotient is odd; and determining the initial quantization value as the target quantization value if the watermark bit to be embedded corresponding to the coefficient is 1 and the quotient is odd, or the watermark bit to be embedded corresponding to the coefficient is 0 and the quotient is even.

[0055] In this process, the initial quantization value is divided by the quantization step size to obtain the quotient. In the parity coding rule, odd numbers correspond to bit "1" and even numbers correspond to bit "0". If the watermark bit to be embedded corresponding to a coefficient is 1 and the quotient is even, or the watermark bit to be embedded corresponding to a coefficient is 0 and the quotient is odd, then the sum of the initial quantization value and the quantization step size is determined as the target quantization value. If the watermark bit to be embedded corresponding to a coefficient is 1 and the quotient is odd, or the watermark bit to be embedded corresponding to a coefficient is 0 and the quotient is even, then the initial quantization value is determined as the target quantization value. For example, if the initial quantization value is 130, the quantization step size is 10, and the watermark bit to be embedded is 0, the quotient obtained by dividing the initial quantization value by the quantization step size is 13. Since 13 is odd, the target quantization value is determined to be 14. The quotient obtained by dividing the target quantization value by the quantization step size is 14, which is even, satisfying the rule that even numbers correspond to bit "0".

[0056] Optionally, it also includes: determining a watermark bit depth threshold based on the region size of the intermediate frequency region and the image size of the third channel image; if the number of bits in the watermark encoding information is greater than the watermark bit depth threshold, then updating the watermark encoding information.

[0057] The watermark bit threshold is the maximum number of watermark bits carried by the third channel image.

[0058] Specifically, a watermark bit depth threshold is determined based on the size of the mid-frequency region and the image size of the third channel image. If the number of bits in the watermark encoding information exceeds the threshold, the watermark encoding information is updated so that the number of bits in the updated watermark encoding information is less than or equal to the threshold. This ensures that the watermark encoding information can be completely embedded in the mid-frequency region of the third channel image, avoiding embedding failure or image distortion due to insufficient capacity. If the number of bits in the watermark encoding information exceeds the threshold, a targeted update strategy can be triggered.

[0059] Optionally, the method further includes: obtaining the watermark text length to be extracted; performing channel separation on the target image to obtain the channel image to be extracted; the channel image to be extracted includes a first channel image to be extracted, a second channel image to be extracted, and a third channel image to be extracted; the first channel image to be extracted is an R channel image; the second channel image to be extracted is a G channel image; and the third channel image to be extracted is a B channel image; determining the number of watermark encoding bits based on the watermark text length; performing discrete cosine transform on the channel image to be extracted to obtain the DCT coefficient matrix to be extracted; obtaining at least one coefficient to be extracted and the coordinate value corresponding to the coefficient to be extracted in the mid-frequency region of the DCT coefficient matrix to be extracted based on the number of watermark encoding bits; determining the watermark position to be determined corresponding to the coordinate value based on the quotient of the coefficient to be extracted and the quantization step size; determining the target watermark position corresponding to the coefficient to be extracted based on the confidence level of the channel image to be extracted and the watermark position to be determined; and performing binary conversion on at least one target watermark position to obtain the watermark text information to be extracted.

[0060] Here, "watermark to be extracted" refers to the watermark that needs to be extracted from the target image containing the watermark. "Watermark text length" refers to the number of characters in the original text corresponding to the watermark to be extracted. "Extractable channel image" is the single-channel grayscale image obtained after performing channel separation on the target image containing the watermark. "Watermark encoding bit length" refers to the total number of bits corresponding to the watermark text length. "Extractable coefficients" are the coefficients corresponding to the watermark to be extracted. "Watermark bit to be determined" is a single binary bit (0 or 1) determined based on the parity of the quotient of the extractable coefficient and the quantization step size. "Confidence level" is a quantitative indicator that measures the reliability of the determination result for a single watermark bit to be determined. "Target watermark bit" is the finally determined binary bit.

[0061] Specifically, the process of reconstructing the hidden watermark text information from the watermarked target image during the embedding stage can be as follows: First, determine the number of watermark encoding bits based on the watermark text length. For example, if the watermark text length is 10 characters and one character corresponds to 8 bits, then the watermark encoding bits are 80. Second, perform channel separation on the target image to obtain the channel image to be extracted. Third, perform Discrete Cosine Transform (DCT) on the channel image to obtain the DCT coefficient matrix to be extracted, which converts the spatial domain channel image to be extracted into a frequency domain DCT coefficient matrix, thus locating the mid-frequency region where the watermark is embedded. Fourth, based on the number of watermark encoding bits, obtain the coefficients and coordinate values ​​to be extracted in the mid-frequency region. That is, select the corresponding number of coefficients and coordinate values ​​from the mid-frequency region of the DCT coefficient matrix to be extracted, based on the required total number of bits to be extracted (e.g., 80). Fifth, determine the watermark bit corresponding to the coordinate value based on the quotient of the coefficients to be extracted and the quantization step size, i.e., reverse the judgment according to the parity rule during embedding (e.g., the embedding convention "0 corresponds to even numbers, 1 to even numbers"). (For odd numbers, the watermark bit to be determined is 0 when the quotient is even and 1 when the quotient is odd.) Based on the confidence level of the channel image to be extracted and the watermark bit to be determined, the target watermark bit corresponding to the coefficient to be extracted is determined. The watermark bits to be determined in the R, G, or B channels are weighted and voted on. The channel with higher confidence has a larger weight, and the bit with the most votes is the target watermark bit. At least one target watermark bit is converted from binary to obtain the watermark text information to be extracted. That is, at least one target watermark bit is converted from binary to character according to the character encoding standard (such as ASCII code) at the time of embedding, and all characters are concatenated to obtain the watermark text information to be extracted.

[0062] Optionally, based on the confidence level of the channel image to be extracted and the watermark position to be determined, the target watermark position corresponding to the coordinate value is determined, including: determining the watermark position to be determined corresponding to the coordinate value in the channel image to be extracted; if the watermark position to be determined is 1, then the channel image to be extracted is determined as the first voting party; if the watermark position to be determined is 0, then the channel image to be extracted is determined as the second voting party; summing the confidence levels of the first voting party to obtain the first voting score, and summing the confidence levels of the second voting party to obtain the second voting score; if the first voting score is greater than the second voting score, then the target watermark position corresponding to the coordinate value is determined to be 1; if the first voting score is less than or equal to the second voting score, then the target watermark position corresponding to the coordinate value is determined to be 0.

[0063] In this context, the first voter refers to the image channel to be extracted that determines the watermark bit corresponding to the coordinate value as 1. The second voter refers to the image channel to be extracted that determines the watermark bit corresponding to the coordinate value as 0. The first vote score refers to the vote score for the watermark bit corresponding to the coordinate value being determined as 1. The second vote score refers to the vote score for the watermark bit corresponding to the coordinate value being determined as 0.

[0064] Specifically, when voting to determine the target watermark bit corresponding to a certain coefficient in the mid-frequency region, if the watermark bit to be determined corresponding to the coordinate value of the coefficient in the channel image to be extracted is 1, then the channel image to be extracted is the first voter; if the watermark bit to be determined corresponding to the coordinate value of the coefficient in the channel image to be extracted is 0, then the channel image to be extracted is the second voter. The confidence scores of all first voters are summed to obtain the first voting score, and the confidence scores of all second voters are summed to obtain the second voting score. If the first voting score is greater than the second voting score, then the target watermark bit corresponding to the coordinate value is determined to be 1; if the first voting score is less than or equal to the second voting score, then the target watermark bit corresponding to the coordinate value is determined to be 0. For example, when voting to determine the target watermark position corresponding to a certain coefficient in the mid-frequency region, if the confidence level of the first channel image to be extracted (R channel image) is 0.3, the watermark position to be determined corresponding to the coordinate value of that coefficient in its matrix is ​​0; the confidence level of the second channel image to be extracted (G channel image) is 0.2, the watermark position to be determined corresponding to the coordinate value of that coefficient in its matrix is ​​0; and the confidence level of the third channel image to be extracted (B channel image) is 0.7, the watermark position to be determined corresponding to the coordinate value of that coefficient in its matrix is ​​1; then the first voting party is the B channel image, the second voting party is the R channel image and the G channel image, the first vote score is 0.7, the second vote score is 0.5, and the target watermark position is 1.

[0065] This invention employs a four-step watermark embedding process: coordinate mapping, coefficient quantization, parity encoding, and matrix update. By precisely binding a single intermediate frequency coefficient to one watermark bit, excessive modification of the DCT coefficients is avoided. This ensures stable watermark embedding while preventing image distortion caused by significant coefficient changes, resulting in stronger reproducibility and more concealed embedding traces. This enhances the watermark's invisibility and resistance to compression and noise. A watermark extraction method is constructed, starting from channel separation and DCT... From transformation and intermediate frequency coefficient extraction to watermark position determination and binary restoration, each step strictly corresponds to the parameters of the embedding process (quantization step size, intermediate frequency region range, coordinate mapping rules), ensuring accurate restoration of watermark information. The extraction process does not rely on the original image to be embedded; only the length of the watermark text is required for extraction. A multi-channel confidence-weighted voting mechanism is used to determine the target watermark position, using the confidence of each channel to be extracted as the voting weight, avoiding misjudgments caused by noise and compression interference during single-channel extraction. The above image watermarking processing scheme combines invisibility, robustness, versatility, and accuracy.

[0066] Figure 3This is a schematic diagram of an image watermarking processing device provided in an embodiment of the present invention. This embodiment is applicable to situations involving watermark processing. The device can be implemented in hardware and / or software and can be configured in an electronic device with corresponding data processing capabilities, such as a server. Figure 3 As shown, the device includes: The first data acquisition module 310 is used to acquire the image to be embedded, watermark text information, and quantization step size; The first channel determination module 320 is used to perform channel separation on the image to be embedded, and determine the first channel image, the second channel image, and the third channel image; the first channel image is the R channel image; the second channel image is the G channel image; and the third channel image is the B channel image. The first matrix acquisition module 330 is used to perform discrete cosine transform on the third channel image to obtain the first DCT coefficient matrix; The watermark encoding acquisition module 340 is used to perform binary conversion on the watermark text information to obtain the watermark encoding information; The second matrix acquisition module 350 is used to embed the watermark encoding information into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix. The fourth channel acquisition module 360 ​​is used to perform inverse discrete cosine transform on the second DCT coefficient matrix to obtain the fourth channel image; The target image acquisition module 370 is used to obtain a target image based on the fourth channel image, the first channel image, and the second channel image.

[0067] This invention significantly reduces the computational load required for DCT transformation and embedding operations by introducing image scaling and single-channel processing strategies, making the watermark embedding process more efficient. By selecting the blue channel, which is less sensitive to the human eye, for embedding, it ensures that the introduction of the watermark will not have a significant impact on the subjective visual quality of the image. Embedding the watermark in the mid-frequency region enables the watermark to resist a certain degree of common attacks such as compression and filtering. It can effectively improve the efficiency of watermark processing, take into account the invisibility and robustness of watermark processing, and meet the needs of real-time or large-scale image processing, such as batch watermarking of cloud photo albums and real-time video stream embedding.

[0068] Optionally, the second matrix acquisition module 350 includes: The first coefficient acquisition unit is used to acquire at least one coefficient and the corresponding coordinate value in the intermediate frequency region of the first DCT coefficient matrix. The watermark position determination unit is used to perform position mapping between coordinate values ​​and watermark encoding information, and to determine the watermark position to be embedded corresponding to the coefficient from the watermark encoding information. The initial quantization value determination unit is used to quantize the coefficients based on the quantization step size using the coefficient quantization formula to determine the initial quantization value corresponding to the coefficients. The target quantization value determination unit is used to process the initial quantization value according to the parity coding rule, based on the watermark bit to be embedded corresponding to the coefficient and the quantization step size, to obtain the target quantization value; The second matrix acquisition unit is used to update the first DCT coefficient matrix based on at least one target quantization value to obtain the second DCT coefficient matrix.

[0069] Optionally, the first channel determination module 320 includes: The first image determining unit is used to reduce the image to be embedded according to a first ratio to obtain a first image; The first image separation unit is used to perform channel separation on the first image to obtain a first channel image, a second channel image, and a third channel image.

[0070] Optionally, the target image acquisition module 370 includes: The second image determination unit is used to merge the fourth channel image, the first channel image, and the second channel image to obtain the second image. The target image acquisition unit is used to enlarge the second image according to a second ratio to obtain the target image; the ratio of the first ratio and the second ratio are reciprocals of each other.

[0071] Optionally, the target quantization value determination unit includes: The quotient determination subunit is used to determine the quotient based on the quantization value and the quantization step size; The first target quantization value determination subunit is used to determine the target quantization value based on the quantization value and quantization step size by using the parity coding rule: if the watermark bit to be embedded corresponding to the coefficient is 1 and the quotient is even, or the watermark bit to be embedded corresponding to the coefficient is 0 and the quotient is odd. The second target quantization value determination subunit is used to determine the quantization value as the target quantization value by using the parity coding rule: if the watermark bit to be embedded corresponding to the coefficient is 1 and the quotient is odd, or the watermark bit to be embedded corresponding to the coefficient is 0 and the quotient is even.

[0072] Optionally, the device further includes a watermark update module, used to: determine a watermark bit depth threshold based on the region size of the intermediate frequency region and the image size of the third channel image; and update the watermark encoding information if the number of bits in the watermark encoding information is greater than the watermark bit depth threshold.

[0073] Optionally, the device may also include: a watermark extraction module; The watermark extraction module includes: The watermark length acquisition unit is used to obtain the length of the watermark text to be extracted. The channel to be extracted determination unit is used to perform channel separation on the target image to obtain the channel to be extracted image; the channel to be extracted image includes a first channel to be extracted image, a second channel to be extracted image and a third channel to be extracted image; the first channel to be extracted image is an R channel image; the second channel to be extracted image is a G channel image; the third channel to be extracted image is a B channel image; The watermark encoding bit length determination unit is used to determine the watermark encoding bit length based on the watermark text length; The extractable matrix acquisition unit is used to perform discrete cosine transform on the channel image to be extracted to obtain the DCT coefficient matrix to be extracted. The coefficient acquisition unit is used to acquire at least one coefficient to be extracted and the corresponding coordinate value of the coefficient to be extracted in the mid-frequency region of the DCT coefficient matrix to be extracted, based on the number of watermark codes. The watermark position determination unit is used to determine the watermark position corresponding to the coordinate value based on the quotient of the coefficient to be extracted and the quantization step size. The target watermark position determination unit is used to determine the target watermark position corresponding to the coefficient to be extracted based on the confidence level of the channel image to be extracted and the watermark position to be determined. The watermark text determination unit is used to perform binary conversion on at least one target watermark bit to obtain the watermark text information to be extracted.

[0074] Optionally, the target watermark bit determination unit is specifically used for: determining the watermark bit to be determined corresponding to the coordinate value in the channel image to be extracted; if the watermark bit to be determined is 1, then the channel image to be extracted is determined as the first voting party; if the watermark bit to be determined is 0, then the channel image to be extracted is determined as the second voting party; summing the confidence scores of the first voting party to obtain the first voting score, and summing the confidence scores of the second voting party to obtain the second voting score; if the first voting score is greater than the second voting score, then the target watermark bit corresponding to the coordinate value is determined to be 1; if the first voting score is less than or equal to the second voting score, then the target watermark bit corresponding to the coordinate value is determined to be 0.

[0075] The image watermarking processing apparatus provided in the embodiments of the present invention can execute the image watermarking processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0076] According to embodiments of the present invention, the present invention also provides an electronic device, a readable storage medium, and a computer program product.

[0077] Figure 4A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0078] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0079] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0080] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as image watermarking processing methods.

[0081] In some embodiments, the image watermarking processing method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the image watermarking processing method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the image watermarking processing method by any other suitable means (e.g., by means of firmware).

[0082] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0083] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0084] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0085] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0086] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0087] A computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product within the cloud computing service system to address the shortcomings of traditional physical hosts and virtual private servers, such as high management difficulty and weak business scalability.

[0088] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0089] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. An image watermarking processing method, characterized in that, The method includes: Obtain the image to be embedded, the watermark text information, and the quantization step size; The image to be embedded is subjected to channel separation to determine a first channel image, a second channel image, and a third channel image; the first channel image is an R channel image; the second channel image is a G channel image; and the third channel image is a B channel image. Perform a discrete cosine transform on the third channel image to obtain the first DCT coefficient matrix; The watermark text information is converted into binary to obtain the watermark encoding information; The watermark encoding information is embedded into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix; Perform inverse discrete cosine transform on the second DCT coefficient matrix to obtain the fourth channel image; The target image is obtained based on the fourth channel image, the first channel image, and the second channel image.

2. The method according to claim 1, characterized in that, The step of embedding the watermark encoding information into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix includes: For the mid-frequency region of the first DCT coefficient matrix, obtain at least one coefficient in the mid-frequency region and the coordinate value corresponding to the coefficient; The coordinate values ​​and the watermark encoding information are mapped to positions, and the watermark bit to be embedded corresponding to the coefficient is determined from the watermark encoding information; The coefficients are quantized using a coefficient quantization formula based on the quantization step size to determine the initial quantization value corresponding to the coefficients. By using parity coding rules, based on the watermark bit to be embedded corresponding to the coefficient and the quantization step size, the initial quantization value is processed to obtain the target quantization value; The first DCT coefficient matrix is ​​updated based on at least one target quantization value to obtain the second DCT coefficient matrix.

3. The method according to claim 1, characterized in that, The step of performing channel separation on the image to be embedded to determine the first channel image, the second channel image, and the third channel image includes: The image to be embedded is reduced in size according to a first ratio to obtain a first image; The first image is subjected to channel separation to obtain a first channel image, a second channel image, and a third channel image.

4. The method according to claim 3, characterized in that, The process of obtaining the target image based on the fourth channel image, the first channel image, and the second channel image includes: The fourth channel image, the first channel image, and the second channel image are merged to obtain the second image; The second image is enlarged according to the second ratio to obtain the target image; the ratio of the first ratio and the second ratio are reciprocals of each other.

5. The method according to claim 2, characterized in that, The step of processing the initial quantization value using parity coding rules, based on the watermark bit to be embedded corresponding to the coefficient and the quantization step size, to obtain the target quantization value includes: The quotient is determined based on the initial quantization value and the quantization step size; According to the parity coding rule, if the watermark bit to be embedded corresponding to the coefficient is 1 and the quotient is even, or the watermark bit to be embedded corresponding to the coefficient is 0 and the quotient is odd, then the target quantization value is determined based on the initial quantization value and the quantization step size. According to the parity coding rule, if the watermark bit to be embedded corresponding to the coefficient is 1 and the quotient is odd, or the watermark bit to be embedded corresponding to the coefficient is 0 and the quotient is even, then the initial quantization value is determined as the target quantization value.

6. The method according to claim 1, characterized in that, The method further includes: Based on the region size of the intermediate frequency region and the image size of the third channel image, the watermark bit depth threshold is determined; If the number of bits in the watermark encoding information is greater than the watermark bit threshold, then the watermark encoding information is updated.

7. The method according to claim 1, characterized in that, The method further includes: Get the length of the watermark text to be extracted; The target image is subjected to channel separation to obtain an image of the channel to be extracted; the image of the channel to be extracted includes a first image of the channel to be extracted, a second image of the channel to be extracted, and a third image of the channel to be extracted; the first image of the channel to be extracted is an R channel image; the second image of the channel to be extracted is a G channel image; and the third image of the channel to be extracted is a B channel image. Based on the length of the watermark text, determine the number of bits in the watermark encoding; Perform a discrete cosine transform on the image of the channel to be extracted to obtain the DCT coefficient matrix to be extracted; Based on the watermark encoding bits, at least one coefficient to be extracted and the coordinate value corresponding to the coefficient to be extracted in the mid-frequency region of the DCT coefficient matrix to be extracted are obtained. Based on the quotient of the coefficient to be extracted and the quantization step size, the watermark position to be determined corresponding to the coordinate value is determined; Based on the confidence level of the channel image to be extracted and the watermark position to be determined, the target watermark position corresponding to the coefficient to be extracted is determined; At least one target watermark bit is converted into binary to obtain the watermark text information of the watermark to be extracted.

8. The method according to claim 7, characterized in that, Based on the confidence level of the image to be extracted and the watermark position to be determined, the target watermark position corresponding to the coordinate value is determined, including: Determine the watermark position to be determined corresponding to the coordinate value in the channel image to be extracted; If the watermark bit to be determined is 1, then the channel image to be extracted is determined as the first voting party; If the watermark bit to be determined is 0, then the channel image to be extracted is determined as the second voting party; The confidence scores of the first voting party are summed to obtain the first voting score, and the confidence scores of the second voting party are summed to obtain the second voting score; If the first voting score is greater than the second voting score, then the target watermark bit corresponding to the coordinate value is determined to be 1; If the first voting score is less than or equal to the second voting score, then the target watermark bit corresponding to the coordinate value is determined to be 0.

9. An image watermarking processing device, characterized in that, The device includes: The first data acquisition module is used to acquire the image to be embedded, watermark text information, and quantization step size; The first channel determination module is used to perform channel separation on the image to be embedded, and determine a first channel image, a second channel image, and a third channel image; the first channel image is an R channel image; the second channel image is a G channel image; and the third channel image is a B channel image. The first matrix acquisition module is used to perform discrete cosine transform on the third channel image to obtain the first DCT coefficient matrix; The watermark encoding acquisition module is used to perform binary conversion on the watermark text information to obtain watermark encoding information; The second matrix acquisition module is used to embed the watermark encoding information into the mid-frequency region of the first DCT coefficient matrix to obtain the second DCT coefficient matrix. The fourth channel acquisition module is used to perform inverse discrete cosine transform on the second DCT coefficient matrix to obtain the fourth channel image; The target image acquisition module is used to obtain a target image based on the fourth channel image, the first channel image, and the second channel image.

10. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to perform the image watermarking processing method according to any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the image watermarking processing method according to any one of claims 1-8.

12. A computer program product comprising a computer program that, when executed by a processor, implements the image watermarking processing method according to any one of claims 1-8.

Images