Watermark area determination method and device, computer equipment and storage medium
By determining and updating the quantized boundary value and selecting suitable watermark embedding areas, the problem of sparse watermark blocks being easily changed in image compression is solved, and the robustness and identification effect of the watermark are improved.
Patent Information
- Application Number
- CN202310512264.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-08
- Publication Date
- 2024-11-15
AI Technical Summary
The sparse watermark block is easily changed or removed during image compression, resulting in the inability to effectively identify the image copyright, affecting the identification effect of the watermark.
By obtaining the spectral components of the image chunking, determining the quantization boundary value, and compressing and updating according to the compression quality, selecting the watermark embedding area to reduce the impact on image compression and improving the robustness of the watermark.
It effectively avoids the impact of image compression on watermarks, enhances the identification effect of watermarks, and improves the storage probability of watermarks in image compression processing.
Smart Images

Figure CN118967417A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of computer technology, and in particular to a method, device, computer equipment and storage medium for determining a watermark region. Background Art
[0002] In order to better identify the copyright of an image, watermarks are often embedded in the image. However, the more watermark blocks are embedded in the image, the more changes may be made to the image content. Therefore, in order to accurately identify the copyright of the image and reduce the changes to the image content, sparse watermark blocks are often embedded in the image, that is, the number of watermark blocks embedded in the image is minimized.
[0003] However, the sparsity of the watermark block makes the watermark in the image more likely to be changed or removed during the transmission process due to image compression, which makes it impossible to use the watermark to detect copyright on the compressed image, affecting the identification effect of the watermark. Therefore, how to reduce the impact of image compression processing on the watermark while ensuring the sparsity of the watermark has become an urgent problem to be solved. Summary of the invention
[0004] The embodiments of the present disclosure at least provide a method, apparatus, computer device and storage medium for determining a watermark area.
[0005] In a first aspect, an embodiment of the present disclosure provides a method for determining a watermark area, comprising:
[0006] Acquire multiple image blocks of the target image, wherein the image blocks are corresponding blocks on the image channel to be embedded with the watermark, and the pixel points in the image blocks have corresponding spectral components;
[0007] For any of the image blocks, determine the quantization boundary values corresponding to the spectral components respectively according to the first spectral values of the spectral components corresponding to the image blocks and the quantization values in the target quantization template; the target quantization template is related to the preset compression quality, and the quantization boundary value is used to represent the difference between the first spectral value and the value obtained after the spectral component is quantized;
[0008] Compressing the image blocks according to the preset compression quality to obtain compression values of the respective spectral components corresponding to the image blocks;
[0009] According to the first spectrum value and the compression value, the quantization boundary value corresponding to each of the spectrum components is updated to obtain an updated quantization boundary value;
[0010] According to the updated quantization boundary value, the total quantization boundary value corresponding to the image block is determined, and according to the total quantization boundary value and a preset area size, a watermark embedding area is determined from the target image.
[0011] In a possible implementation, determining the quantization boundary values corresponding to the spectral components respectively according to the first spectral values of the spectral components corresponding to the image blocks and the quantization values in the target quantization template includes:
[0012] For any of the spectral components, determine a remainder corresponding to the spectral component according to a first spectral value of the spectral component and a quantization value in the target quantization template that matches the spectral component;
[0013] A minimum value is selected from the remainder and the difference between the quantized value and the remainder as the quantization boundary value corresponding to the spectral component.
[0014] In a possible implementation manner, updating the quantization boundary values corresponding to the spectral components respectively according to the first spectral value and the compression value to obtain updated quantization boundary values includes:
[0015] According to the value range of the pixel points in the target image, inverse transform is performed on each of the compression values to obtain a conversion pixel value corresponding to each of the compression values;
[0016] Determine, according to the converted pixel value, second spectrum values corresponding to each of the spectrum components;
[0017] According to the first spectrum value and the second spectrum value respectively corresponding to each of the spectrum components, the quantization boundary value respectively corresponding to each of the spectrum components is updated to obtain an updated quantization boundary value.
[0018] In a possible implementation manner, updating the quantization boundary value corresponding to each of the spectral components according to the first spectral value and the second spectral value corresponding to each of the spectral components to obtain the updated quantization boundary value includes:
[0019] Determining target weighted values corresponding to the respective spectral components according to the first spectral values and the second spectral values corresponding to the respective spectral components;
[0020] Selecting the target quantization value with the largest value from each quantization value in the target quantization template;
[0021] The target weighted value and the target quantized value are used to update the quantized boundary values corresponding to the spectral components to obtain updated quantized boundary values.
[0022] In a possible implementation manner, determining the target weighted value corresponding to each of the spectral components according to the first spectral value and the second spectral value corresponding to each of the spectral components includes:
[0023] Using a preset first weighting value as a target weighting value of a spectral component corresponding to the first spectral value and the second spectral value being consistent; and
[0024] The preset second weighting value is used as a target weighting value of a spectral component corresponding to a difference between the first spectral value and the second spectral value.
[0025] In a possible implementation manner, before determining the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, the method further includes:
[0026] Determining a target threshold value according to the preset compression quality;
[0027] In a case where the updated quantization boundary value corresponding to the spectral component is less than or equal to the target threshold, the updated quantization boundary value is updated to a quantization value in the target quantization template that matches the spectral component.
[0028] In a possible implementation, determining the total quantization boundary value corresponding to the image block according to the updated quantization boundary value includes:
[0029] Sorting the updated quantization boundary values corresponding to the respective spectral components of the image blocks in ascending order to obtain a sorting order of the updated quantization boundary values;
[0030] According to each of the updated quantization boundary values whose sorting order is less than or equal to the preset order, the total quantization boundary value corresponding to the image block is determined.
[0031] In a possible implementation manner, determining the watermark embedding area from the target image according to the total quantized boundary value and the preset area size includes:
[0032] Determining a target size according to the preset area size and the image size of the image block;
[0033] According to the position of each image block in the image channel to be embedded with the watermark, using the target sliding window with the target size, sequentially traverse the total quantized boundary values corresponding to the image blocks, and determine the sum of the boundary values obtained in each traversal;
[0034] Determine each of the target image blocks corresponding to the minimum sum of the boundary values, and use the area corresponding to the target image block in the target image as the watermark embedding area.
[0035] In a possible implementation manner, determining the watermark embedding area from the target image according to the total quantized boundary value and the preset area size includes:
[0036] Determine a target size according to the preset area size and the image size of the image block, and determine a target convolution kernel using the target size;
[0037] According to the position of each of the image blocks in the image channel to be embedded with the watermark, using the target convolution kernel, the target sliding window with the target size and the total quantization boundary value of the image block, each convolution value corresponding to the target image is determined in parallel;
[0038] A watermark embedding area is determined from the target image according to each of the convolution values.
[0039] In a possible implementation, determining the watermark embedding area from the target image according to each of the convolution values includes:
[0040] Sorting the convolution values in ascending order to obtain a sorting order of the convolution values;
[0041] Determine a target order according to the number of targets in a preset area, and select target convolution values whose sorting order is less than or equal to the target order from the convolution values;
[0042] The preset number of watermark embedding areas is determined according to the image blocks corresponding to the target convolution values and the areas to which they belong in the target image.
[0043] In a second aspect, the present disclosure also provides a watermark region determination device, including:
[0044] An acquisition module is used to acquire a plurality of image blocks of a target image, wherein the image blocks are corresponding blocks on the image channel to be embedded with a watermark, and the pixel points in the image blocks have corresponding spectral components;
[0045] A first determination module is used to determine, for any of the image blocks, the quantization boundary values corresponding to the spectral components respectively according to the first spectral values of the spectral components corresponding to the image blocks and the quantization values in the target quantization template; the target quantization template is related to the preset compression quality, and the quantization boundary value is used to represent the difference between the first spectral value and the value obtained after the spectral component is quantized;
[0046] A compression module, used for compressing the image blocks according to the preset compression quality to obtain compression values of the respective spectral components corresponding to the image blocks;
[0047] A first updating module, configured to update the quantization boundary values corresponding to the spectral components according to the first spectral value and the compression value, to obtain updated quantization boundary values;
[0048] The second determination module is used to determine the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, and determine the watermark embedding area from the target image according to the total quantization boundary value and a preset area size.
[0049] In a third aspect, an optional implementation of the present disclosure further provides a computer device, a processor, and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the processor is used to execute the machine-readable instructions stored in the memory, and when the machine-readable instructions are executed by the processor, the machine-readable instructions perform the steps of the above-mentioned first aspect, or any possible implementation of the first aspect.
[0050] In a fourth aspect, an optional implementation of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed, the steps of the above-mentioned first aspect, or any possible implementation of the first aspect are executed.
[0051] For a description of the effects of the above-mentioned watermark area determination device, computer equipment, and computer-readable storage medium, please refer to the description of the above-mentioned watermark area determination method, which will not be repeated here.
[0052] The watermark region determination method, device, computer equipment and storage medium provided by the disclosed embodiment, since the size of the quantization boundary value is used to characterize whether the change of the position of the spectral component in the image is easier to preserve, the quantization boundary value is calculated by each quantization value in the target quantization template and the first spectral value of each spectral component, and the value corresponding to each spectral component that is easier to preserve can be determined; and the quantization change value is then used to select the watermark embedding region, which can improve the probability that the watermark is completely preserved after the watermark is embedded in the selected region. By compressing the image blocks, and then using the compression value of each spectral component and the first spectral value to update the quantization boundary value, not only can the unreasonable quantization boundary value be updated, and the rationality and accuracy of the quantization boundary value corresponding to each spectral component be improved, but also the sensitivity of each spectral component to the compression coefficient can be reduced, and the robustness of the watermark in the subsequently selected watermark embedding region to the image compression processing can be improved. By using the total quantized boundary values corresponding to each image block to select the watermark embedding area, we can select the watermark embedding area where the image compression processing has the least impact on the watermark in the area. By embedding the watermark in the selected watermark embedding area, we can effectively avoid the impact of image compression processing on the watermark and enhance the identification effect of the watermark.
[0053] In order to make the above-mentioned objectives, features and advantages of the present disclosure more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following is a brief introduction to the drawings required for use in the embodiments. The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present disclosure and are used together with the specification to illustrate the technical solutions of the present disclosure. It should be understood that the following drawings only illustrate certain embodiments of the present disclosure and should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can also be obtained based on these drawings without creative work.
[0055] Figure 1 A flow chart of a method for determining a watermark area provided by an embodiment of the present disclosure is shown;
[0056] Figure 2 A schematic diagram showing a target quantization template corresponding to the Y channel when a preset compression quality Q=50 provided by an embodiment of the present disclosure is shown;
[0057] Figure 3 A schematic diagram showing a watermark embedding area selected from a target image provided by an embodiment of the present disclosure;
[0058] Figure 4 A schematic diagram of a watermark region determination device provided by an embodiment of the present disclosure is shown;
[0059] Figure 5 A schematic diagram of the structure of a computer device provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION
[0060] In order to make the purpose, technical scheme and advantages of the embodiments of the present disclosure clearer, the technical scheme in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. The components of the embodiments of the present disclosure generally described and shown here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present disclosure is not intended to limit the scope of the present disclosure claimed for protection, but merely represents the selected embodiments of the present disclosure. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present disclosure.
[0061] In addition, the terms "first", "second", etc. in the description and claims of the embodiments of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein.
[0062] The "multiple or several" mentioned in this article refers to two or more. "And / or" describes the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.
[0063] Research has found that compared to embedding watermark blocks in the entire image, the fewer watermark blocks there are in the image, the less change will be made to the image content. However, the watermark in the sparse watermark block is very easy to be changed by image compression processing, resulting in the watermark of the compressed image being unable to effectively identify the copyright of the image. Therefore, it is very important to avoid the impact of image compression processing on sparse watermark blocks while not changing the watermark embedding algorithm. In this way, how to avoid the impact of image compression processing on sparse watermark blocks has become a problem that needs to be solved urgently.
[0064] Based on the above research, the present disclosure provides a method, device, computer equipment and storage medium for determining a watermark region. Since the size of the quantization boundary value is used to characterize whether the change of the position of the spectral component in the image is easier to preserve, the quantization boundary value is calculated through each quantization value in the target quantization template and the first spectral value of each spectral component, and the value corresponding to each spectral component that is easier to preserve can be determined; and the quantization change value is then used to select the watermark embedding region, which can increase the probability that the watermark is completely preserved after the watermark is embedded in the selected region. By compressing the image blocks, and then using the compression value of each spectral component and the first spectral value to update the quantization boundary value, not only can the unreasonable quantization boundary value be updated, and the rationality and accuracy of the quantization boundary value corresponding to each spectral component be improved, but also the sensitivity of each spectral component to the compression coefficient can be reduced, and the robustness of the watermark in the subsequently selected watermark embedding region to the image compression processing can be improved. By using the total quantized boundary values corresponding to each image block to select the watermark embedding area, we can select the watermark embedding area where the image compression processing has the least impact on the watermark in the area. By embedding the watermark in the selected watermark embedding area, we can effectively avoid the impact of image compression processing on the watermark and enhance the identification effect of the watermark.
[0065] The defects existing in the above solutions are the results obtained by the inventor after practice and careful research. Therefore, the discovery process of the above problems and the solutions proposed by the present disclosure for the above problems below should be the contributions made by the inventor to the present disclosure during the disclosure process.
[0066] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.
[0067] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.
[0068] It should be noted that the specific terms mentioned in the embodiments of the present disclosure include:
[0069] JPEG: Joint Photographic Experts Group, a standard for continuous-tone still image compression;
[0070] RGB: RGB color mode is a color standard in the industry. It obtains various colors by changing the three color channels of red (R), green (G), and blue (B) and superimposing them on each other. RGB represents the colors of the three channels of red, green, and blue.
[0071] YUV is a color encoding method. "Y" represents brightness, that is, grayscale value; "U" and "V" represent chrominance (Chrominance or Chroma), which is used to describe the color and saturation of the image and is used to specify the color of the pixel.
[0072] uint8 type: It is an 8-bit unsigned integer with a value range of 0 to 255.
[0073] To facilitate understanding of the present embodiment, a watermark area determination method disclosed in the embodiment of the present disclosure is first introduced in detail. The execution subject of the watermark area determination method provided in the embodiment of the present disclosure is generally a terminal device or other processing device with certain computing capabilities, wherein the terminal device may be a user equipment (UE), a mobile device, a user terminal, a terminal, a personal digital assistant (PDA), a handheld device, a computer device, etc.; in some possible implementations, the watermark area determination method may be implemented by a processor calling computer-readable instructions stored in a memory.
[0074] The watermark area determination method provided by the embodiment of the present disclosure is described below by taking a computer device as an example of an execution subject.
[0075] like Figure 1 As shown, it is a flow chart of a method for determining a watermark area provided by an embodiment of the present disclosure, which may include the following steps:
[0076] S101: Acquire multiple image blocks of a target image, where the image blocks are corresponding blocks on the image channels to be embedded with watermarks, and pixels in the image blocks have corresponding spectral components.
[0077] Here, since JPEG compression is a common image compression method, the watermark embedding area determined by the watermark area determination method provided in the embodiment of the present disclosure can have good robustness to JPEG compression after embedding the watermark, that is, the watermark area determination method in the embodiment of the present disclosure is a method for JPEG compression, and the watermark area determination method provided in the embodiment of the present disclosure is designed based on the principle of JPEG compression. The traditional JPEG compression process is first introduced below:
[0078] 1. [Lossless] Convert an RGB image of shape 3×H×W to a YUV image; H represents the height of the image and W represents the width of the image; lossless means that the operation of this step will not lose image information.
[0079] 2. [Lossy] Downsample the images of the U and V channels (the purpose is to reduce both H and W to 1 / 2 of the original); lossy means that the operation of this step may lose image information.
[0080] 3. [Lossless] Divide the Y, U and V channel images into blocks respectively, and obtain the image blocks corresponding to the Y, U and V channel images respectively, where the size of each image block is 8×8; the H×W Y channel image is divided into That is, the Y channel image includes 8×8 image blocks; similarly, The U channel (or V channel) of the image is divided into Blocks, that is, the Y channel (or V channel) image includes 8×8 image blocks.
[0081] 4. [Lossless] Perform discrete cosine transform on the blocked image matrices corresponding to the Y, U, and V channels respectively to obtain the spectral values corresponding to the spectral components on each channel, where one spectral component corresponds to one pixel point; the dimension of the image matrix remains unchanged before and after the transformation, that is, the dimension of the matrix of the spectral component corresponding to one image block is also 8×8.
[0082] 5. [Lossy] The matrices corresponding to the Y, U and V channels are quantized using 8×8 quantization templates to obtain the compression values corresponding to the above spectral components of each channel; the compression values of each channel can be expressed by the following formulas 1 to 3:
[0083]
[0084]
[0085]
[0086] Among them, D(x Y ) Q Indicates the corresponding compression value on the Y channel, D(x U ) Q Indicates the corresponding compression value on the U channel, D(x V ) Q Indicates the corresponding compression value on the V channel, Q indicates the preset compression quality, round(·) indicates rounding, and M indicates the quantization template. Indicates the quantization template corresponding to the Y channel, Indicates the quantization template corresponding to the U channel, Denotes the quantization template corresponding to the V channel, D(x Y ) represents the spectrum value on the Y channel, D(x U ) represents the spectrum value on the U channel, D(x V ) represents the spectrum value on the V channel, and ⊙ represents multiplication.
[0087] Based on the above formula 1, the image blocks on the Y channel can be compressed, based on the above formula 2, the image blocks on the U channel can be compressed, and based on the above formula 3, the image blocks on the V channel can be compressed.
[0088] The watermark region determination method provided by the embodiment of the present disclosure is mainly for analyzing the above-mentioned step 5, so as to determine the ideal compression robust region. In the specific implementation, the embodiment of the present disclosure does not specifically limit the channel for watermark embedding, that is, watermark embedding can be performed in any channel of the Y, U and V channels or a combination of multiple channels.
[0089] The target image mentioned above may be an image for which a target area needs to be selected for watermark embedding. The image blocks are blocks corresponding to the image channel to be embedded with the watermark. Taking the Y channel as an example, each image block may be an 8×8 block on the Y channel; taking the U channel as an example, each image block may be an 8×8 block on the U channel; taking the V channel as an example, each image block may be an 8×8 block on the V channel. The spectrum values of the spectrum components corresponding to each pixel point in the image block may be the values in the 8×8 matrix obtained after performing a discrete cosine transform (DCT forDiscrete Cosine Transform, referred to as DCT transform) on the values of each pixel point in the image block on the image channel. For example, the spectrum values corresponding to the image block on the Y channel may be the values in the 8×8 matrix obtained after performing a DCT transform on the values of each pixel point in the image block on the Y channel. An 8×8 block will correspond to 64 spectrum components, and one spectrum component corresponds to one spectrum value. The spectrum value may be, for example, D(x Y )、D(x U ) or D(x V). The first spectrum value is the spectrum value of each spectrum component corresponding to the 8×8 block. The image size of each image block is set to 8×8 because the minimum size of a watermark is 8×8. The spectrum components corresponding to each pixel in the image block are the spectrum components corresponding to the image block. The spectrum values of each spectrum component corresponding to an 8×8 image block are the spectrum values of each spectrum component in the 8×8 spectrum block obtained after discrete cosine transform is performed on the 8×8 image block. That is, the spectrum values of each spectrum component corresponding to an image block are the spectrum values of each spectrum component in the spectrum block corresponding to the image block.
[0090] In a specific implementation, after acquiring the target image, the target image can be divided on the image channel to be embedded with the watermark to obtain a plurality of image blocks corresponding to the target image on the image channel to be embedded with the watermark. At the same time, the first spectrum value of the spectrum component corresponding to each pixel point in the image block can be determined according to the pixel value of each pixel point in the image block on the image channel to be embedded with the watermark.
[0091] S102: For any image block, determine the quantization boundary values corresponding to each spectral component according to the first spectral value of each spectral component corresponding to the image block and the quantization values in the target quantization template; the target quantization template is related to the preset compression quality, and the quantization boundary value is used to characterize the difference between the first spectral value and the value obtained after the spectral component is quantized.
[0092] Here, the target quantization template may be a template corresponding to the channel where the watermark needs to be embedded, and the image channel to be embedded with the watermark is the channel where the watermark needs to be embedded. For example, when the image channel to be embedded with the watermark is the Y channel, the target quantization template may be the quantization template corresponding to the Y channel. When the image channel to be embedded with watermark is the U channel, the target quantization template can be the quantization template corresponding to the U channel. When the image channel to be embedded with watermark is the V channel, the target quantization template can be the quantization template corresponding to the V channel.
[0093] The size of the target quantization template is consistent with the image size of the image block, which is 8×8. The target quantization template can include 64 quantization values, which are used to compress and quantize the spectrum values of the spectrum components. For example, when the first spectrum value 10 corresponding to the spectrum component is compressed and quantized on the Y channel using the quantization value 16, 10 can be substituted into D(x Y ), substitute 16 into the above formula 1 Get the compressed value D(x) of the spectral component Y ) Q is 16.
[0094] The preset compression quality Q is used to characterize the strength of compression of the target image. Specifically, the larger the preset compression quality Q is, the lower the compression strength is. The target quantization templates under different preset compression qualities are different. Figure 2 As shown, it is a schematic diagram of a target quantization template corresponding to the Y channel when a preset compression quality Q=50 provided in an embodiment of the present disclosure. Figure 2 It includes 64 quantized values.
[0095] The quantization boundary value is used to characterize the difference between the first spectrum value and the value obtained after the spectrum component is quantized. The spectrum component is quantized, that is, the first spectrum value of the spectrum component is modified by the quantization value. The size of the quantization boundary value is used to indicate whether the change in the position of the spectrum component in the image is easier to preserve. In the embodiment of the present disclosure, the change in the position of the spectrum component in the image can be specifically the added watermark information. Specifically, it can be combined with Figure 2 The target quantization template on the Y channel is shown to illustrate the quantization boundary value by embedding the watermark on the Y channel:
[0096] Since the compression value is determined by rounding in the JPEG compression quantization process, it can be known that:
[0097] like but Wherein, i and j both belong to the interval [0, 7]. For example, when i, j is (0, 0), it represents the first spectrum value corresponding to the spectrum component of the 0th row and 0th column of the image block, that is, the first spectrum value in the upper left corner.
[0098] by Figure 2 Taking the first quantization value (16) of the upper left corner of the target quantization template as an example, for the spectrum component of the upper left corner of the 8×8 matrix obtained after the image block is transformed by DCT, if the first spectrum value D(x Y ) 0,0 ∈[8, 24), then the quantized value D(x Y ) Q 0,0 It is still 16. Therefore, if a watermark δ is added to the image position corresponding to the spectral component and then after DCT transformation, the first spectral value of the spectral component is still in the interval [8, 24) (that is, D(x Y +δ) 0,0 ∈[8, 24)), then after compression and quantization using the target quantization template, the watermark will be erased, and the following formula 4 can be obtained:
[0099]
[0100] Based on this, we can also know that for the spectrum component in the upper left corner of the 8×8 matrix obtained after the image block is transformed by DCT, if a watermark δ is added at the image position corresponding to the spectrum component and then transformed by DCT, the first spectrum value D(x y +δ) 0,0 →16, then after compression quantization using the target quantization template, the first spectrum value of the spectrum component is still 16, thereby achieving the effect of retaining the added watermark δ, that is, the following formula 5 can be obtained:
[0101]
[0102] Generally speaking, the smaller the change to the first spectrum value corresponding to the spectrum component, the less obvious the added noise will be. Combined with the above analysis of the JPEG compression principle when the preset compression quality Q=50, it can be seen that the watermark embedding area can be determined by finding the image block with the smallest information filtered out by "rounded quantization". The watermark information embedded in such an area is more likely to be retained. Therefore, the concept of quantization boundary value corresponding to the spectrum component is proposed. The first spectrum value of each spectrum component corresponding to the image block and each quantization value in the target quantization template are used to calculate the minimum quantization boundary value corresponding to each spectrum component.
[0103] Taking embedding a watermark on the Y channel as an example, for any spectral component corresponding to any image block, the quantization boundary value corresponding to the spectral component can be determined according to the first spectral value corresponding to the spectral component and the quantization value matching the position of the spectral component in the target quantization template. For example, when the position of the spectral component is (i, j) = (0, 0), the quantization value matching the position of the spectral component in the target quantization template is the quantization value at the 0th row and the 0th column (i.e., the upper left corner).
[0104] In one embodiment, the above S102 may be implemented according to the following steps:
[0105] S102-1: For any spectral component, determine a remainder corresponding to the spectral component according to a first spectral value of the spectral component and a quantization value in a target quantization template that matches the spectral component.
[0106] S102-2: Select a minimum value from the remainder and the difference between the quantized value and the remainder as the quantization boundary value corresponding to the spectral component.
[0107] Here, the quantization value that matches the spectral component in the target quantization template is the quantization value that matches the position of the spectral component in the target quantization template.
[0108] This step is explained by taking the determination of the quantization boundary value of the spectral component corresponding to the image block on the Y channel as an example:
[0109] For the image on the Y channel, after the image is divided into blocks, the size of the matrix corresponding to the image is For any 8×8 image block, the quantization boundary value of each spectral component corresponding to the image block can be determined by the following formula 6:
[0110]
[0111] Among them, B i , j represents the quantization boundary value of the spectral component of the image block corresponding to the i-th row and j-th column, D(x Y ) i , j represents the first spectrum value corresponding to the spectrum component at the i-th row and j-th column, represents the quantization value at the i-th row and j-th column in the target quantization template, where i∈[0,7], j∈[0,7], and min represents the value of and The minimum value among , % means taking the remainder.
[0112] In specific implementation, the first spectrum value of any spectrum component corresponding to the image block and the quantization value matching the spectrum component in the target quantization template can be substituted into the above formula 6 to obtain the quantization boundary value corresponding to the spectrum component.
[0113] It can be understood that if it is necessary to determine the quantization boundary value of the spectral component corresponding to the image block on the U channel, then D(x Y ) i,j Replace with D(x U ) i,j , and Replace with If you need to determine the quantization boundary value of the spectral component corresponding to the image block on the V channel, you can replace D(x Y ) i,j Replace with D(x V ) i,j , and Replace with That's it.
[0114] For example, the first spectrum value of the spectrum component at the position (0, 0) corresponding to an image block on the Y channel is 10, and the target quantization template is Figure 2 The template shown in FIG. 1 is used to determine the quantization boundary value B of the spectral component at the position (0, 0) corresponding to the image block. 0,0=min(10%16, 16-10%16)=min(10,6)=6. In other words, if you want to change the first spectrum value from 10 to 16, the minimum step size is 6, and at the same time, the change to the image position corresponding to the first spectrum value is the smallest and the change can be retained after compression and quantization.
[0115] S103: compressing the image blocks according to a preset compression quality to obtain compression values of each spectral component corresponding to the image blocks.
[0116] Here, for RGB images, the pixel values of each pixel point belong to the interval [0, 255], and the target image with the watermark added is also in RGB format. The quantization boundary value is only used to determine the target area. After the target area is determined, the quantization boundary value needs to be used to restore the RGB pixel value of the corresponding position. Therefore, after determining the value of each B, the value of B cannot be directly used to determine the watermark area, because there may be a pixel value restored by using a certain B value that exceeds the interval [0, 255], resulting in an unreasonable determined watermark embedding area. Therefore, in order to improve the rationality of the determined watermark embedding area, it is necessary to update some of the B values that are out of bounds, and in order to filter out the B values that are out of bounds, it is necessary to use the compressed values of each spectral component.
[0117] In specific implementation, after determining the image channel that needs to be embedded with a watermark, the target quantization template under the preset compression quality can be used to compress and quantize each image block on the image channel according to the quantization formula matching the image channel, so as to obtain the compression value of each spectral component corresponding to each image block.
[0118] Optionally, target quantization templates under other compression qualities may be used to compress each image block to obtain compression values of each spectral component corresponding to each image block. Here, other compression qualities may be set based on experience and are not specifically limited in the embodiments of the present disclosure.
[0119] For example, when the image channel to be embedded with a watermark is the Y channel, the following formula 7 may be used to determine the compression value of the spectral component corresponding to each image block on the Y channel:
[0120]
[0121] Where Q represents the preset compression quality, Indicates a compressed value.
[0122] It should be noted that there is no strict execution order between S102 and S103.
[0123] S104: updating the quantization boundary values corresponding to the respective spectral components according to the first spectral value and the compression value to obtain updated quantization boundary values.
[0124] In specific implementation, after determining the compression value of the spectrum component, the RGB pixel value corresponding to the spectrum component can be restored according to the compression value of the spectrum component. Then, according to the RGB pixel value and the first spectrum value of the spectrum component, it is determined whether the quantization boundary value corresponding to the spectrum component is an out-of-bounds boundary value. If so, the quantization boundary value corresponding to the spectrum component can be updated to obtain an updated quantization boundary value. If not, it means that the quantization boundary value corresponding to the spectrum component is reasonable, and it can be directly used to determine the watermark embedding area.
[0125] In one embodiment, the above S104 can be implemented according to the following steps:
[0126] S104-1: Perform inverse transformation on each compression value according to the value range of the pixel points in the target image to obtain a converted pixel value corresponding to each compression value.
[0127] Here, the value range of the pixel points in the target image is [0, 255], and the inverse transform may specifically refer to an inverse discrete cosine transform.
[0128] In a specific implementation, taking the Y channel as an example, the following formula 8 can be used to inversely transform each compression value to clip the pixel value of the compressed and quantized image to [0, 255]:
[0129]
[0130] Among them, x′ Y Indicates the converted pixel value. Indicates the compression value corresponding to the spectral component, D I Represents inverse discrete cosine transform, and clip represents clipping, which is used to limit the converted pixel value to the interval [0, 255].
[0131] S104-2: Determine second spectrum values corresponding to respective spectrum components according to the converted pixel values.
[0132] In a specific implementation, after obtaining the converted pixel values corresponding to the spectral components, the converted pixel values can be used to perform DCT transformation to obtain the second spectral values corresponding to the spectral components when the pixel values of the pixels corresponding to the spectral components are the converted pixel values. It can be understood that the first spectral value is the spectral value when the pixel values of the pixels corresponding to the spectral components are the original pixel values corresponding to the target image.
[0133] S104-3: According to the first spectrum value and the second spectrum value corresponding to each spectrum component, respectively, the quantization boundary value corresponding to each spectrum component is updated to obtain an updated quantization boundary value.
[0134] In specific implementation, for any spectral component, it can be determined whether the first spectral value and the second spectral value corresponding to the spectral component are consistent. If they are consistent, it can be determined that the quantization boundary value corresponding to the spectral component is in a reasonable range and does not need to be updated. On the contrary, if they are inconsistent, it means that the quantization boundary value corresponding to the spectral component has crossed the boundary and needs to be updated. For example, the quantization boundary value can be updated to the second spectral value, or updated to the quantization value that matches the position of the spectral component in the target quantization template.
[0135] In this way, by updating the quantization boundary value, the quantization boundary value that exceeds the value range corresponding to the RGB image after being restored to the pixel value can be updated. The watermark area is determined by using the updated quantization boundary value, which can improve the rationality and reliability of the determined watermark area.
[0136] In one embodiment, the above S104-3 may be implemented according to the following steps:
[0137] S104-3-1: Determine target weighted values corresponding to the respective spectral components according to the first spectral values and the second spectral values corresponding to the respective spectral components.
[0138] Here, one spectrum component may correspond to one target weighting value, and the target weighting value may be used to perform weighted processing on the quantization boundary value corresponding to the spectrum component, thereby realizing the updating of the quantization boundary value.
[0139] In a specific implementation, two weighting values may be preset, namely, a first weighting value and a second weighting value. For example, the first weighting value may be 0, and the second weighting value may be 1. Then, according to whether the first spectrum value and the second spectrum value corresponding to the spectrum component are the same, a target weighting value may be determined for the spectrum component from the first weighting value and the second weighting value.
[0140] Exemplarily, the preset first weighting value can be used as the target weighting value for the spectral components whose corresponding first spectral values and second spectral values are consistent; and the preset second weighting value can be used as the target weighting value for the spectral components whose corresponding first spectral values and second spectral values are different.
[0141] Specifically, the target weighted values corresponding to the various spectral components can be determined according to the following formula 9:
[0142]
[0143] Among them, M D represents the target weighted value, D(x′Y ) represents the second spectrum value, D(x Y ) represents the first spectrum value, else represents the value of the spectrum except D(x′ Y )==D(x Y )
[0144] S104-3-2: Select the target quantization value with the largest value from each quantization value in the target quantization template.
[0145] In specific implementation, the maximum target quantization value can be selected from the various quantization values in the target quantization template corresponding to the image channel to be embedded with the watermark. For example, the target quantization value can be represented by ε. When the image channel to be embedded with the watermark is the Y channel, ε=max , where max means taking the maximum value. Figure 2 Taking the target quantization template shown as an example, the selected target quantization value can be 121.
[0146] S104-3-3: Using the target weighted value and the target quantized value, update the quantized boundary value corresponding to each spectral component to obtain an updated quantized boundary value.
[0147] In specific implementation, the following formula 10 can be used to update the quantization boundary values corresponding to each spectral component, thereby obtaining the updated quantization boundary values corresponding to each spectral component:
[0148] B N =M D ×ε+(1-M D )×B; (Formula 10)
[0149] Among them, B N represents the updated quantization boundary value, ε=max(M Y ), represents the target quantization value, and B represents the quantization boundary value before updating. Based on the above formula 10, it can be known that although the quantization boundary value corresponding to each spectral component can be updated, the quantization boundary value corresponding to the spectral component corresponding to the first spectral value and the second spectral value that are consistent can be kept consistent before and after the update.
[0150] S105: Determine the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, and determine the watermark embedding area from the target image according to the total quantization boundary value and the preset area size.
[0151] Here, the preset area size is a size corresponding to a preset watermark area. The size of the preset area size can be set based on experience and is not specifically limited in the embodiments of the present disclosure.
[0152] One image block corresponds to one total quantization boundary value, and the total quantization boundary value of one image block can be determined according to the updated quantization boundary values of each spectral component corresponding to the image block.
[0153] Exemplarily, for any image block, the total quantized boundary value corresponding to the image block can be determined based on the sum of the updated quantized boundary values of the various spectral components corresponding to the image block. Then, the number of image blocks that a watermark area needs to include can be determined based on the preset area size, and then the target image block can be determined from the image blocks based on the number and the total quantized boundary value of each image block, and then the area corresponding to the target image block in the target image is used as the watermark embedding area. Finally, the watermark can be embedded in the determined watermark area to obtain a watermark image. In this way, the robustness of the watermark in the watermark embedding area to image compression processing (especially JPEG compression) can be improved.
[0154] In one embodiment, considering that the value type of the pixel value of the RGB image is uint8 type with a value range of 0 to 255, for the smaller updated quantization boundary value (ie, B N ) The image area corresponding to the selected watermark may be removed when the image is saved even if the watermark information is added. Therefore, the updated quantization boundary value needs to be further updated to filter the smaller updated quantization boundary value, thereby improving the accuracy of the selected watermark embedding area.
[0155] Specifically, before executing “determining the total quantization boundary value corresponding to the image block according to the updated quantization boundary value”, a target threshold for filtering the updated quantization boundary value may be determined according to a preset compression quality. Exemplarily, the target threshold may be determined according to the following formula 11:
[0156] ε B =(200-2×Q)×0.05; (Formula 11)
[0157] Among them, ε B represents the target threshold, and Q represents the preset compression quality.
[0158] Then, the target threshold value may be used to filter and update each updated quantization boundary value, thereby obtaining the final updated quantization boundary value corresponding to each spectral component.
[0159] For example, still taking the image channel to be embedded with a watermark as the Y channel as an example, the following formula 12 can be used to update each updated quantization boundary value:
[0160]
[0161] in, represents the final updated quantization boundary value corresponding to the spectral component of the i-th row and j-th column, represents the updated quantization boundary value corresponding to the spectral component of the i-th row and j-th column, Indicates the quantization value at the i-th row and j-th column in the target quantization template corresponding to the Y channel, else indicates that Other than this.
[0162] Based on the above formula twelve, it can be known that when the updated quantization boundary value corresponding to the spectral component is less than or equal to the target threshold, the updated quantization boundary value can be updated to the quantization value that matches the spectral component in the target quantization template. Similarly, when the updated quantization boundary value corresponding to the spectral component is greater than the target threshold, the updated quantization boundary value corresponding to the spectral component can be directly used as the final updated quantization boundary value corresponding to the spectral component.
[0163] In this way, the target threshold can be used to achieve the smaller updated quantization boundary value (i.e., B N ) is updated, which effectively avoids the problem that the watermark cannot be saved after the image area corresponding to the updated quantization boundary value is used as the watermark embedding area and the watermark is added.
[0164] Further, after obtaining the final updated quantization boundary value corresponding to each spectral component, that is, after obtaining the B corresponding to each spectral component N′ When the B corresponding to each spectral component is used N′ , calculate the total quantized boundary value corresponding to each image block.
[0165] In one embodiment, the step of “determining the total quantization boundary value” in S105 above may be implemented according to the following steps:
[0166] Step 1: Sort the updated quantization boundary values corresponding to the respective spectral components of the image blocks in order from small to large, and obtain the sorting order of the updated quantization boundary values.
[0167] Step 2: Determine the total quantization boundary value corresponding to the image block according to each updated quantization boundary value whose sorting order is less than or equal to the preset order.
[0168] Here, since the spectral component with a larger quantization boundary value has a greater impact on the overall image content of the target image, in order to avoid the problem that the excessively large quantization boundary value dominates the finally determined watermark embedding area, causing the determined watermark embedding area to be unreasonable, the updated quantization boundary value of the spectral component corresponding to the image block can be filtered first to improve the rationality of the determined watermark embedding area. The preset order can be determined based on the total number of spectral components corresponding to the image block and the preset retention ratio. For example. When the preset retention ratio is 50% and the total number of spectral components corresponding to the image block is 64, the preset order can be determined to be 32.
[0169] For example, for any image block, when determining the total quantization boundary value corresponding to the image block, the B corresponding to each spectral component corresponding to the image block can be first calculated in ascending order. N′ Sort and get each B N′ The minimum sorting order can be 1 and the maximum sorting order can be 64. Then, the B with a sorting order less than or equal to 32 can be sorted. N′ The sum is performed to obtain the total quantized boundary value corresponding to the image block.
[0170] Specifically, the total quantization boundary value corresponding to the image block can be determined according to the following formula 13:
[0171]
[0172] Among them, B m,n Indicates the total quantized boundary value corresponding to the image block in the mth row and nth column. For example, the image channel to be embedded with watermark is the Y channel. H is the image height of the target image, and W is the image width of the target image. For example, the image channel to be embedded with watermark is U channel or V channel. Used to filter out B whose sorting order is greater than the preset order N′ .
[0173] In specific implementation, the B of each spectral component corresponding to each image block is N′ Substituting into the above formula 13, the total quantization boundary value of the image block can be determined.
[0174] Optionally, when the image channel to be embedded with watermark includes multiple channels, the above formula 13 Corresponding updates can be made to obtain the total quantization boundary value under multiple image channels. However, when the image channel to be embedded with a watermark is the U channel or the V channel, based on the above-mentioned JPEG compression principle, it can be known that for the U channel and the V channel, the corresponding H and W are reduced to 1 / 2 of the original, so the number of spectral components on the U channel and the V channel is equal to 1 / 4 of the number of spectral components on the Y channel. The updated quantization boundary value corresponding to a spectral component on the image channel is an element in the quantization boundary value matrix corresponding to the image channel. In order to make the number of elements in the quantization boundary value matrix on the U channel and the V channel consistent with the number of elements in the quantization boundary value matrix on the Y channel during calculation, the Kronecker product can be used to expand the number of elements in the quantization boundary value matrix on the U channel and the V channel to 4 times. Exemplarily, the Kronecker product of the 2×2 unit matrix and the quantization boundary value matrix on the U (or V) channel can be used as the corresponding expanded quantization boundary value matrix on the U (or V) channel. Among them, an element in the quantization boundary value matrix on the U (or V) channel represents an updated quantization boundary value corresponding to a spectral component on the U (or V) channel, and the number of elements in the expanded quantization boundary value matrix is increased by 4 times compared with the number of elements in the quantization boundary value matrix before expansion.
[0175] After obtaining the extended quantization boundary value matrices corresponding to the U channel and the V channel, the total quantization boundary value can be determined according to the elements of the extended quantization boundary value matrices corresponding to the U channel and the V channel, and the elements of the quantization boundary value matrix corresponding to the Y channel:
[0176] For example, when the image channels to be embedded with watermarks include Y channel, U channel and V channel, the above formula 13 Can be updated to in, The image block of the Y channel corresponds to the final updated quantization boundary value of the spectral component of the i-th row and j-th column, The image block of the U channel corresponds to the final updated quantization boundary value of the spectral component of the i-th row and j-th column, represents the final updated quantization boundary value corresponding to the spectral component of the i-th row and j-th column of the image block of the V channel, and · represents multiplication. For another example, when the image channels to be watermarked include the Y channel and the V channel, the above formula 13 Can be updated to
[0177] After determining the total quantization boundary value of each image block, the preset region size can be used as the size of the watermark embedding region to determine the various to-be-confirmed embedding regions with the preset region size that can be formed by using the various image blocks. The size of each to-be-confirmed embedding region is the preset region size, and the image blocks included in different to-be-confirmed embedding regions are different, and the image blocks belonging to the same to-be-confirmed embedding region are continuous in position. Then, the to-be-confirmed embedding region with the smallest total quantization boundary value can be selected from the to-be-confirmed embedding regions as the watermark embedding region.
[0178] Specifically, after determining the total quantization boundary value of each image block, the following formula 14 for solving the coordinates of the upper left corner of the image block in the watermark embedding area can be constructed:
[0179]
[0180] Among them, (x * ,y * ) represents the coordinate of the upper left corner of the determined watermark embedding area, argmin is used to calculate the minimum value of the function, W represents the width of the target image, H represents the height of the target image, w represents the width of the area to be embedded with the watermark in the preset area size, h represents the height in the preset area size, that is, w×h is the preset area size.
[0181] In one embodiment, the step of "determining a watermark embedding area" in S105 above may be implemented according to the following steps:
[0182] S1: Determine a target size according to a preset region size and an image size of the image blocks.
[0183] Here, after obtaining the objective function corresponding to the above formula 14, the objective function can be solved using steps S1 to S3 to determine the watermark embedding area. The image size of the image block can be 8×8, and the target size can be the size of the sliding window used to solve the above objective function.
[0184] For example, still taking the embedding of watermark in the Y channel as an example, according to the preset region size h×w and the image size of the image block 8×8, the target size can be determined to be
[0185] S2: According to the position of each image block in the image channel to be embedded with the watermark, the total quantized boundary values corresponding to the image blocks are traversed in turn using a target sliding window with a target size to determine the sum of the boundary values obtained in each traversal.
[0186] Here, we still take embedding watermark in the Y channel as an example. We can divide an image block of the Y channel into a point. In this way, we can get the image size corresponding to the image of the Y channel as That is, an image of the Y channel can include points, one of which is an 8×8 image block.
[0187] In specific implementation, according to the position of each image block in the target image, a target size of The sliding window on the Y channel The sliding window is traversed in turn in the divided image, and then the image blocks covered by each traversal can be determined, and the total quantized boundary values corresponding to each covered image block (for example, B N′ ) and determines the sum of the boundary values obtained in this traversal.
[0188] For example, when h and w are both 16, and H and W are both 256, the size of the image of the Y channel after block division is 32×32, that is, an image of the Y channel includes 32×32 image blocks, and the target size is 2×2. In the first traversal, the image blocks covered by the sliding window include four, with coordinates of (0, 0), (0, 1), (1, 0), and (1, 1), respectively. Therefore, the B of these four image blocks can be used. N′ Then, in the second traversal, the image blocks covered by the sliding window also include four, with coordinates (0, 1), (0, 2), (1, 1) and (1, 2), respectively. Therefore, the B values of these four image blocks can be used. N′ The sum of the boundary values corresponding to the second traversal is obtained, and so on. Through continuous sliding traversal, the sum of the boundary values corresponding to each traversal can be obtained. At the same time, the traversal of each image block in the 32×32 image can also be completed.
[0189] The objective function is solved by traversing the target sliding window with the target size. The time complexity of the whole solution process is O(H′W′K 2 ), where O represents the time complexity,
[0190] S3: Determine each target image block corresponding to the minimum sum of the boundary values, and use the area corresponding to the target image block in the target image as the watermark embedding area.
[0191] For example, using After the target window has traversed all the image blocks on the Y channel, the sum of the boundary values corresponding to each traversal can be obtained. Then, the minimum sum of boundary values can be determined from the sum of all boundary values. The image blocks used to obtain the minimum sum of boundary values are used as target image blocks, and finally the area corresponding to the target image blocks in the target image can be used as the watermark embedding area. As for the above objective function, the coordinates of the target image block with the smallest row and column coordinates in each target image block (that is, the coordinates of the target image block with the relative position in the upper left corner among multiple target image blocks) can be determined as the result of the objective function, that is, the coordinates of the target image block with the smallest row and column coordinates can be used as the (x * ,y * Then, we get (x * ,y * ) can be used according to (x * ,y * ) determine the corresponding coordinate position in the target image, and determine the watermark embedding area corresponding to the target image on the image channel to be embedded with the watermark according to the preset area size and the coordinate position.
[0192] In another embodiment, in order to further improve the speed of solving the objective function, the objective function may be solved by parallel convolution operations. Specifically, the step of "determining the watermark embedding area" in the above S105 may also be implemented according to the following steps:
[0193] P1: Determine the target size based on the preset region size and the image size of the image block, and use the target size to determine the target convolution kernel.
[0194] Here, the target size is the size of each target sliding window used in the convolution calculation process. The number of target convolution kernels can be equal to the total number of the sum of the above-mentioned boundary values (in this case, one target sliding window corresponds to one convolution kernel), or it can be equal to the number of image blocks (in this case, one image block corresponds to one target convolution kernel). The size of each target convolution kernel is consistent, each convolution operation corresponds to a target sliding window, and each target sliding window corresponds to a target convolution kernel, which can be the target convolution kernel of the upper left corner of each image block included in the target sliding window. It can be understood that since the size of each target convolution kernel is consistent, the target convolution kernel of the target sliding window can also be determined using the target convolution kernel of any image block included in the target sliding window.
[0195] Exemplarily, the target convolution kernel can be determined using the following formula 15:
[0196]
[0197] Among them, kernel i , j It represents the convolution kernel corresponding to the image block whose upper left corner coordinate is the i-th row and j-th column, and K is the target size.
[0198] Taking the Y channel as an example, according to the preset region size h×w and the image size of the image block 8×8, the target size can be determined as According to the target size and the above formula 15, it can be determined that the convolution kernel corresponding to each target sliding window is
[0199] Optionally, the size of the target convolution kernel may be pre-set. When performing convolution calculation, the pre-set target convolution kernel may be directly used to perform convolution calculation, thereby obtaining a convolution value.
[0200] P2: According to the position of each image block in the image channel to be embedded with watermark, the target convolution kernel, the target sliding window with the target size and the total quantization boundary value of the image block are used to determine the convolution values corresponding to the target image in parallel.
[0201] Exemplarily, the convolution value is a value obtained by performing convolution calculation on the total quantization boundary values of each image block included in the target sliding window using the target convolution kernel corresponding to the target sliding window.
[0202] Since the convolution operation has been highly optimized and can be calculated in parallel, after obtaining the target convolution kernel, the number of target sliding windows to be used and the image blocks corresponding to each target sliding window can be determined according to the position of each image block in the image channel to be embedded with the watermark and the target size of the target sliding window. Then, the convolution values obtained by performing convolution calculations on each target sliding window can be determined by parallel calculation.
[0203] The number of target sliding windows may be equal to the total number of the sums of the boundary values, and the number of convolution values is also equal to the total number.
[0204] P3: Determine the watermark embedding area from the target image based on each convolution value.
[0205] In specific implementation, the minimum convolution value can be selected from each convolution value to determine each target image block included in the target sliding window corresponding to the convolution value, and then the coordinates of the target image block with the smallest row coordinates and column coordinates in each target image block (that is, the coordinates of the target image block with the relative position located in the upper left corner in multiple target image blocks) can be determined as the result of the objective function, that is, the coordinates of the target image block with the smallest row coordinates and column coordinates can be used as the (x * ,y* ). After that, the position mapping method can be used to determine the (x * ,y * ) in the target image, and determine the watermark embedding area corresponding to the target image on the image channel to be embedded with the watermark according to the preset area size and the coordinate position.
[0206] In this way, by means of parallel computing, the speed of determining the convolution value can be increased, thereby increasing the speed of solving the above objective function.
[0207] In another embodiment, if multiple watermark embedding regions need to be selected from the target image, the above P3 can be implemented according to the following steps to obtain multiple watermark embedding regions:
[0208] P3-1: Sort the convolution values in ascending order to obtain the sorting order of the convolution values.
[0209] P3-2: Determine the target order according to the preset number of regions, and select the target convolution values whose sorting order is less than or equal to the target order from the convolution values.
[0210] Here, the preset area number is the number of watermark embedding areas that need to be selected in advance. The target order is the order indicated by the preset area number. For example, if the preset area number is 3, the target order is 3. The preset area number can be greater than 1.
[0211] With the target order being α, target convolution values whose sorting order is less than or equal to α can be screened out from the convolution values. The number of target convolution values is equal to the number of preset regions.
[0212] P3-3: Determine a preset number of watermark embedding areas according to the image blocks corresponding to each target convolution value and the area to which it belongs in the target image.
[0213] Exemplarily, for each target convolution value, the coordinates of the image block with the smallest row coordinates and column coordinates in the image block corresponding to the target convolution value can be determined as the result of the objective function. In this way, multiple results corresponding to the objective function can be obtained. Then, for each result, the coordinate position corresponding to the result in the target image can be determined by position mapping, and the watermark embedding area corresponding to the target image on the image channel to be embedded with the watermark can be determined according to the preset area size and the coordinate position. The watermark embedding area is the area in the target image to which each image block corresponding to the target convolution value belongs in the target image.
[0214] In this way, by sorting the convolution values to screen the watermark embedding areas, it is possible to screen out multiple watermark embedding areas with relatively small corresponding convolution values in the target area. The smaller the convolution value, the more likely the watermark embedded in the watermark embedding area corresponding to the convolution value will be completely preserved after JPEG compression, ultimately improving the robustness of the watermark to image compression processing.
[0215] like Figure 3 As shown, it is a schematic diagram of a watermark embedding area selected from a target image provided by an embodiment of the present disclosure, wherein the watermark embedding areas selected in the target image include 3 (i.e., areas 1 to 3), each of which corresponds to the Y channel, and the size of each watermark embedding area is 128×128. The three watermark embedding areas are areas where the convolution values are located in the first 3 smallest areas. Figure 3 Also included are multiple image blocks showing the total quantization boundary values located in the first 1024, and one image block corresponds to Figure 3 A dot in the image, each watermark embedding area includes multiple image blocks.
[0216] Based on the above embodiments, since the size of the quantization boundary value is used to characterize whether the change of the position of the spectral component in the image is easier to preserve, the quantization boundary value is calculated through each quantization value in the target quantization template and the first spectrum value of each spectral component, and the value corresponding to each spectral component that is easier to preserve can be determined; and the quantization change value is used to select the watermark embedding area, which can improve the probability that the watermark is completely preserved after the watermark is embedded in the selected area. By compressing the image blocks, and then using the compression value of each spectral component and the first spectrum value to update the quantization boundary value, not only can the unreasonable quantization boundary value be updated, and the rationality and accuracy of the quantization boundary value corresponding to each spectral component can be improved, but also the sensitivity of each spectral component to the compression coefficient can be reduced, and the robustness of the watermark in the subsequently selected watermark embedding area to the image compression processing can be improved. The total quantization boundary value corresponding to each image block is used to select the watermark embedding area, and the watermark embedding area with the least effect of image compression processing on the watermark in the area can be selected. By embedding the watermark in the selected watermark embedding area, the influence of image compression processing on the watermark can be effectively avoided, and the identification effect of the watermark can be enhanced.
[0217] Exemplarily, still taking the preset area size of 128×128 as an example, the test data set is 1000 test images with a resolution of 1200×1200, and the watermark area determination method provided by the embodiment of the present disclosure and the random area positioning method are used to select 3 watermark embedding areas in the test image. The watermark embedding algorithm and extraction algorithm select the HiDDen algorithm and the MBRS algorithm. The bit error (Bit Error) of watermark extraction is tested under the compression strength of the compression quality Q∈[40,80]. Among them, the smaller the Bit Error, the better the watermark embedding effect. As shown in Table 1 below, the mean value of the Bit Error of the three watermark embedding areas corresponding to the two watermark area positioning methods are calculated, as well as the improvement effect:
[0218]
[0219] Table 1
[0220] Based on the above Table 1, it can be seen that the watermark embedded in the watermark embedding area determined by the watermark area determination method provided in the embodiment of the present disclosure can have better robustness to image compression processing than the watermark embedded in the watermark embedding area determined by the random area positioning method.
[0221] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.
[0222] Based on the same inventive concept, a watermark area determination device corresponding to the watermark area determination method is also provided in the embodiment of the present disclosure. Since the principle of solving the problem by the device in the embodiment of the present disclosure is similar to the above-mentioned watermark area determination method in the embodiment of the present disclosure, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.
[0223] like Figure 4 FIG. 1 is a schematic diagram of a watermark region determination device provided by an embodiment of the present disclosure, comprising:
[0224] The acquisition module 401 is used to acquire a plurality of image blocks of a target image, wherein the image blocks are corresponding blocks on the image channel to be embedded with a watermark, and the pixel points in the image blocks have corresponding spectral components;
[0225] A first determination module 402 is used to determine, for any of the image blocks, the quantization boundary values corresponding to the spectral components respectively according to the first spectral values of the spectral components corresponding to the image blocks and the quantization values in the target quantization template; the target quantization template is related to the preset compression quality, and the quantization boundary value is used to represent the difference between the first spectral value and the value obtained after the spectral component is quantized;
[0226] A compression module 403, configured to compress the image blocks according to the preset compression quality to obtain compression values of the respective spectral components corresponding to the image blocks;
[0227] A first updating module 404, configured to update the quantization boundary values corresponding to the spectral components according to the first spectral value and the compression value, to obtain updated quantization boundary values;
[0228] The second determination module 405 is used to determine the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, and determine the watermark embedding area from the target image according to the total quantization boundary value and a preset area size.
[0229] In a possible implementation manner, the first determination module 402, when determining the quantization boundary values corresponding to the respective spectral components according to the first spectral values of the respective spectral components corresponding to the image blocks and the respective quantization values in the target quantization template, is configured to:
[0230] For any of the spectral components, determine a remainder corresponding to the spectral component according to a first spectral value of the spectral component and a quantization value in the target quantization template that matches the spectral component;
[0231] A minimum value is selected from the remainder and the difference between the quantized value and the remainder as the quantization boundary value corresponding to the spectral component.
[0232] In a possible implementation manner, the first updating module 404, when updating the quantization boundary values corresponding to the spectral components respectively according to the first spectral value and the compression value to obtain the updated quantization boundary values, is configured to:
[0233] According to the value range of the pixel points in the target image, inverse transform is performed on each of the compression values to obtain a conversion pixel value corresponding to each of the compression values;
[0234] Determine, according to the converted pixel value, second spectrum values corresponding to each of the spectrum components;
[0235] According to the first spectrum value and the second spectrum value respectively corresponding to each of the spectrum components, the quantization boundary value respectively corresponding to each of the spectrum components is updated to obtain an updated quantization boundary value.
[0236] In a possible implementation manner, the first updating module 404, when updating the quantization boundary value corresponding to each of the spectral components according to the first spectral value and the second spectral value corresponding to each of the spectral components to obtain the updated quantization boundary value, is configured to:
[0237] Determining target weighted values corresponding to the respective spectral components according to the first spectral values and the second spectral values corresponding to the respective spectral components;
[0238] Selecting the target quantization value with the largest value from each quantization value in the target quantization template;
[0239] The target weighted value and the target quantized value are used to update the quantized boundary values corresponding to the spectral components to obtain updated quantized boundary values.
[0240] In a possible implementation manner, the first updating module 404, when determining the target weighted value corresponding to each of the spectral components according to the first spectral value and the second spectral value corresponding to each of the spectral components, is configured to:
[0241] Using a preset first weighting value as a target weighting value of a spectral component corresponding to the first spectral value and the second spectral value being consistent; and
[0242] The preset second weighting value is used as a target weighting value of a spectral component corresponding to a difference between the first spectral value and the second spectral value.
[0243] In a possible implementation, the apparatus further includes a second updating module 406, which, before determining the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, is configured to:
[0244] Determining a target threshold value according to the preset compression quality;
[0245] In a case where the updated quantization boundary value corresponding to the spectral component is less than or equal to the target threshold, the updated quantization boundary value is updated to a quantization value in the target quantization template that matches the spectral component.
[0246] In a possible implementation manner, the second determination module 405, when determining the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, is configured to:
[0247] Sorting the updated quantization boundary values corresponding to the respective spectral components of the image blocks in ascending order to obtain a sorting order of the updated quantization boundary values;
[0248] According to each of the updated quantization boundary values whose sorting order is less than or equal to the preset order, the total quantization boundary value corresponding to the image block is determined.
[0249] In a possible implementation manner, the second determination module 405, when determining the watermark embedding area from the target image according to the total quantized boundary value and the preset area size, is configured to:
[0250] Determining a target size according to the preset area size and the image size of the image block;
[0251] According to the position of each image block in the image channel to be embedded with the watermark, using the target sliding window with the target size, sequentially traverse the total quantized boundary values corresponding to the image blocks, and determine the sum of the boundary values obtained in each traversal;
[0252] Determine each of the target image blocks corresponding to the minimum sum of the boundary values, and use the area corresponding to the target image block in the target image as the watermark embedding area.
[0253] In a possible implementation manner, the second determination module 405, when determining the watermark embedding area from the target image according to the total quantized boundary value and the preset area size, is configured to:
[0254] Determine a target size according to the preset area size and the image size of the image block, and determine a target convolution kernel using the target size;
[0255] According to the position of each of the image blocks in the image channel to be embedded with the watermark, using the target convolution kernel, the target sliding window with the target size and the total quantization boundary value of the image block, each convolution value corresponding to the target image is determined in parallel;
[0256] A watermark embedding area is determined from the target image according to each of the convolution values.
[0257] In a possible implementation manner, the second determination module 405, when determining the watermark embedding area from the target image according to each of the convolution values, is configured to:
[0258] Sorting the convolution values in ascending order to obtain a sorting order of the convolution values;
[0259] Determine a target order according to the preset number of regions, and select target convolution values whose sorting order is less than or equal to the target order from the convolution values;
[0260] The preset number of watermark embedding areas is determined according to the image blocks corresponding to the target convolution values and the areas to which they belong in the target image.
[0261] For descriptions of the processing flow of each module in the device and the interaction flow between each module, reference may be made to the relevant descriptions in the above method embodiment, which will not be described in detail here.
[0262] Based on the same technical concept, the embodiment of the present application also provides a computer device. Figure 5 FIG. 1 is a schematic diagram of a computer device provided in an embodiment of the present application, including:
[0263] Processor 51, memory 52 and bus 53. The memory 52 stores machine-readable instructions executable by the processor 51, and the processor 51 is used to execute the machine-readable instructions stored in the memory 52. When the machine-readable instructions are executed by the processor 51, the processor 51 executes the following steps: SS101: Acquire multiple image blocks of the target image, where the image blocks are corresponding blocks on the image channels to be embedded with watermarks, and the pixels in the image blocks have corresponding spectral components; S102: For any image block, determine the quantization boundaries corresponding to each spectral component according to the first spectral values of each spectral component corresponding to the image block and each quantization value in the target quantization template. value; the target quantization template is related to the preset compression quality, and the quantization boundary value is used to characterize the difference between the first spectrum value and the value obtained after the spectrum component is quantized; S103: compress the image blocks according to the preset compression quality to obtain the compression values of each spectrum component corresponding to the image blocks; S104: according to the first spectrum value and the compression value, update the quantization boundary values corresponding to each spectrum component respectively to obtain the updated quantization boundary values and S105: according to the updated quantization boundary values, determine the total quantization boundary values corresponding to the image blocks, and determine the watermark embedding area from the target image according to the total quantization boundary values and the preset area size.
[0264] The above-mentioned memory 52 includes a memory 521 and an external memory 522; the memory 521 here is also called an internal memory, which is used to temporarily store the calculation data in the processor 51, as well as the data exchanged with the external memory 522 such as a hard disk. The processor 51 exchanges data with the external memory 522 through the memory 521. When the computer device is running, the processor 51 and the memory 52 communicate through the bus 53, so that the processor 51 executes the execution instructions mentioned in the above-mentioned method embodiment.
[0265] The present disclosure also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the steps of the watermark region determination method described in the above method embodiment are executed. The storage medium can be a volatile or non-volatile computer-readable storage medium.
[0266] The computer program product of the watermark area determination method provided in the embodiments of the present disclosure includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the steps of the watermark area determination method described in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.
[0267] The computer program product may be implemented in hardware, software or a combination thereof. In one optional embodiment, the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (SDK) and the like.
[0268] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the several embodiments provided in the present disclosure, it should be understood that the disclosed device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of the device or unit can be electrical, mechanical or other forms.
[0269] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0270] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
[0271] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present disclosure, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.
[0272] If the technical solution of this application involves personal information, the product using the technical solution of this application has clearly informed the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using the technical solution of this application has obtained the individual's separate consent before processing the sensitive personal information, and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, clear and prominent signs are set to inform that the personal information collection scope has been entered and personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that he or she agrees to the collection of his or her personal information; or on the device that processes personal information, the personal information processing rules are notified by obvious signs / information, and the individual's authorization is obtained through pop-up information or by asking the individual to upload his or her personal information; among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.
[0273] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present disclosure, which are used to illustrate the technical solutions of the present disclosure, rather than to limit them. The protection scope of the present disclosure is not limited thereto. Although the present disclosure is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed in the present disclosure, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.
Claims
1. A method for determining a watermark region, characterized in that: include: Acquire multiple image blocks of the target image, wherein the image blocks are corresponding blocks on the image channel to be embedded with the watermark, and the pixel points in the image blocks have corresponding spectral components; For any of the image blocks, determining the quantization boundary values corresponding to the respective spectral components according to the first spectral values of the respective spectral components corresponding to the image block and the respective quantization values in the target quantization template; The target quantization template is related to the preset compression quality, and the quantization boundary value is used to represent the difference between the first spectrum value and the value obtained after the spectrum component is quantized; Compressing the image blocks according to the preset compression quality to obtain compression values of the respective spectral components corresponding to the image blocks; According to the first spectrum value and the compression value, the quantization boundary value corresponding to each of the spectrum components is updated to obtain an updated quantization boundary value; According to the updated quantization boundary value, the total quantization boundary value corresponding to the image block is determined, and according to the total quantization boundary value and a preset area size, the watermark embedding area is determined from the target image.
2. The method according to claim 1, characterized in that The step of determining the quantization boundary values corresponding to the spectral components according to the first spectral values of the spectral components corresponding to the image blocks and the quantization values in the target quantization template comprises: For any of the spectral components, determine a remainder corresponding to the spectral component according to a first spectral value of the spectral component and a quantization value in the target quantization template that matches the spectral component; A minimum value is selected from the remainder and the difference between the quantized value and the remainder as the quantization boundary value corresponding to the spectral component.
3. The method according to claim 1, characterized in that The updating of the quantization boundary values corresponding to the spectral components according to the first spectral value and the compression value to obtain updated quantization boundary values includes: According to the value range of the pixel points in the target image, inverse transform is performed on each of the compression values to obtain a conversion pixel value corresponding to each of the compression values; Determine, according to the converted pixel value, second spectrum values corresponding to each of the spectrum components; According to the first spectrum value and the second spectrum value respectively corresponding to each of the spectrum components, the quantization boundary value respectively corresponding to each of the spectrum components is updated to obtain an updated quantization boundary value.
4. The method according to claim 3, characterized in that The updating of the quantization boundary value corresponding to each of the spectral components according to the first spectral value and the second spectral value corresponding to each of the spectral components to obtain an updated quantization boundary value includes: Determining target weighted values corresponding to the respective spectral components according to the first spectral values and the second spectral values corresponding to the respective spectral components; Selecting the target quantization value with the largest value from each quantization value in the target quantization template; The target weighted value and the target quantized value are used to update the quantized boundary values corresponding to the spectral components to obtain updated quantized boundary values.
5. The method according to claim 4, characterized in that The determining, according to the first spectrum value and the second spectrum value corresponding to each spectrum component, the target weighted value corresponding to each spectrum component, comprises: Using a preset first weighting value as a target weighting value of a spectral component corresponding to the first spectral value and the second spectral value being consistent; and The preset second weighting value is used as a target weighting value of a spectral component corresponding to a difference between the first spectral value and the second spectral value.
6. The method according to claim 1, characterized in that Before determining the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, the method further includes: Determining a target threshold value according to the preset compression quality; In a case where the updated quantization boundary value corresponding to the spectral component is less than or equal to the target threshold, the updated quantization boundary value is updated to a quantization value in the target quantization template that matches the spectral component.
7. The method according to claim 1, characterized in that The step of determining the total quantization boundary value corresponding to the image block according to the updated quantization boundary value includes: Sorting the updated quantization boundary values corresponding to the respective spectral components of the image blocks in ascending order to obtain a sorting order of the updated quantization boundary values; According to each of the updated quantization boundary values whose sorting order is less than or equal to the preset order, the total quantization boundary value corresponding to the image block is determined.
8. The method according to claim 1, characterized in that The step of determining a watermark embedding area from the target image according to the total quantized boundary value and the preset area size includes: Determining a target size according to the preset area size and the image size of the image block; According to the position of each image block in the image channel to be embedded with the watermark, using the target sliding window with the target size, sequentially traverse the total quantized boundary values corresponding to the image blocks, and determine the sum of the boundary values obtained in each traversal; Determine each of the target image blocks corresponding to the minimum sum of the boundary values, and use the area corresponding to the target image block in the target image as the watermark embedding area.
9. The method according to claim 1, characterized in that: The step of determining a watermark embedding area from the target image according to the total quantized boundary value and the preset area size includes: Determine a target size according to the preset area size and the image size of the image block, and determine a target convolution kernel using the target size; According to the position of each of the image blocks in the image channel to be embedded with the watermark, using the target convolution kernel, the target sliding window with the target size and the total quantization boundary value of the image block, each convolution value corresponding to the target image is determined in parallel; A watermark embedding area is determined from the target image according to each of the convolution values.
10. The method according to claim 9, characterized in that Determining a watermark embedding area from the target image according to each of the convolution values comprises: Sorting the convolution values in ascending order to obtain a sorting order of the convolution values; Determine a target order according to the preset number of regions, and select target convolution values whose sorting order is less than or equal to the target order from the convolution values; The preset number of watermark embedding areas is determined according to the image blocks corresponding to the target convolution values and the areas to which they belong in the target image.
11. A watermark area determination device, characterized in that: include: An acquisition module is used to acquire a plurality of image blocks of a target image, wherein the image blocks are corresponding blocks on the image channel to be embedded with a watermark, and the pixel points in the image blocks have corresponding spectral components; A first determination module is used to determine, for any of the image blocks, the quantization boundary values corresponding to the respective spectral components according to the first spectral values of the respective spectral components corresponding to the image block and the respective quantization values in the target quantization template; The target quantization template is related to the preset compression quality, and the quantization boundary value is used to represent the difference between the first spectrum value and the value obtained after the spectrum component is quantized; A compression module, used for compressing the image blocks according to the preset compression quality to obtain compression values of each of the spectral components corresponding to the image blocks; A first updating module, configured to update the quantization boundary values corresponding to the spectral components according to the first spectral value and the compression value, to obtain updated quantization boundary values; The second determination module is used to determine the total quantization boundary value corresponding to the image block according to the updated quantization boundary value, and determine the watermark embedding area from the target image according to the total quantization boundary value and a preset area size.
12. A computer device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the processor is used to execute the machine-readable instructions stored in the memory. When the machine-readable instructions are executed by the processor, the processor executes the steps of the watermark area determination method as described in any one of claims 1 to 10.
13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program. When the computer program is executed by a computer device, the computer device executes the steps of the watermark area determination method according to any one of claims 1 to 10.
