CN115187444B - Image tracing information safety protection method and device and electronic equipment - Google Patents
Image tracing information safety protection method and device and electronic equipment Download PDFInfo
- Publication number
- CN115187444B CN115187444B CN202211095280.XA CN202211095280A CN115187444B CN 115187444 B CN115187444 B CN 115187444B CN 202211095280 A CN202211095280 A CN 202211095280A CN 115187444 B CN115187444 B CN 115187444B
- Authority
- CN
- China
- Prior art keywords
- image
- watermark
- tested
- preset
- optimal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/0021—Image watermarking
- G06T1/005—Robust watermarking, e.g. average attack or collusion attack resistant
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
Abstract
The application provides an image tracing information safety protection method, an image tracing information safety protection device and electronic equipment, wherein the method comprises the following steps: acquiring an original image of watermark tracing information to be superimposed and a predetermined superimposed image watermark; adjusting characteristic parameters of the preset superimposed image watermarks according to preset constraint conditions of the characteristic parameters of the image watermarks to obtain an adjusted image watermark set; respectively superposing each image watermark in the adjusted image watermark set with the original image to obtain a watermark image set to be tested; performing intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by using a preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested; and determining an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested. The method can effectively resist the intelligent image information removal attack.
Description
Technical Field
The application relates to the field of information security, in particular to an image tracing information security protection method, an image tracing information security protection device and electronic equipment.
Background
The digital watermark is widely applied to scenes such as multimedia data transmission, distribution, sharing and the like as an effective true and false identification and copyright protection means.
However, existing digital watermarks, particularly image watermarks, suffer from a large number of smart watermark information removal attacks based on image restoration techniques, rendering the vast majority of the existing visible watermark mechanisms ineffective. This has a serious impact on copyright protection and traceability evidence collection of images.
Therefore, it is highly desirable to provide a security protection method that can successfully resist such smart image information removal attacks.
Disclosure of Invention
In view of the above, the present application provides a method, an apparatus and an electronic device for protecting image tracing information.
Specifically, the application is realized by the following technical scheme:
according to a first aspect of an embodiment of the present application, there is provided an image tracing information security protection method, including:
acquiring an original image of watermark tracing information to be superimposed and a predetermined superimposed image watermark;
adjusting characteristic parameters of the preset superimposed image watermarks according to preset constraint conditions of the characteristic parameters of the image watermarks to obtain an adjusted image watermark set; wherein the characteristic parameters comprise color parameters of pixel level, and the characteristic parameter constraint conditions comprise adjustment ranges of the characteristic parameters;
Respectively superposing each image watermark in the adjusted image watermark set with the original image to obtain a watermark image set to be tested;
performing intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by using a preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested;
and determining an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested.
According to a second aspect of embodiments of the present application, there is provided an image tracing information security protection apparatus, including:
the acquisition unit is used for acquiring an original image of the watermark tracing information to be superimposed and the image watermark of the preset superposition;
the adjusting unit is used for adjusting the characteristic parameters of the preset superimposed image watermarks according to the preset constraint conditions of the characteristic parameters of the image watermarks to obtain an adjusted image watermark set; wherein the characteristic parameters comprise color parameters of pixel level, and the characteristic parameter constraint conditions comprise adjustment ranges of the characteristic parameters;
the superposition unit is used for respectively superposing each image watermark in the adjusted image watermark set and the original image to obtain a watermark image set to be tested;
The testing unit is used for performing intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by utilizing a preset intelligent image information removal attack testing model to obtain an output image corresponding to each watermark image to be tested;
the determining unit is used for determining the optimal watermark image and the optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested.
According to a third aspect of embodiments of the present application, there is provided an electronic device comprising a processor and a memory storing machine executable instructions executable by the processor for executing the machine executable instructions to implement the method provided in the first aspect.
Before the original image is subjected to image watermark superposition, color parameter adjustment of pixel level is carried out on the preset image watermark feature parameter constraint condition, an adjusted image watermark set is obtained, each image watermark in the adjusted image watermark set is respectively superposed with the original image to obtain a watermark image set to be tested, a preset intelligent image information removal attack test model is utilized, intelligent watermark removal attack is carried out on each watermark image to be tested in the watermark image set to obtain an output image corresponding to each watermark image to be tested, further, the optimal watermark image and the optimal image watermark are determined according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, the feature parameter adjustment of pixel level is carried out on the preset image watermark according to the set constraint condition, on the basis of modifying the watermark information of the preset image to be tested, comprehensibility and copyright content integrity of the watermark image to be tested can be ensured, intelligent image information removal attack can be effectively carried out, and robustness of the watermark information to be greatly improved.
Drawings
Fig. 1 is a flow chart of an image tracing information security protection method according to an exemplary embodiment of the present application;
fig. 2 is a flow chart of an image tracing information security protection method according to an exemplary embodiment of the present application;
fig. 3 is a schematic structural diagram of an image tracing information security protection device according to an exemplary embodiment of the present application;
fig. 4 is a schematic hardware structure of an electronic device according to an exemplary embodiment of the present application.
Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present application as detailed in the accompanying claims.
The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used in this application and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
In order to better understand the technical solutions provided by the embodiments of the present application and make the above objects, features and advantages of the embodiments of the present application more obvious, the technical solutions in the embodiments of the present application are described in further detail below with reference to the accompanying drawings.
Referring to fig. 1, a flow chart of an image tracing information security protection method provided in an embodiment of the present application, as shown in fig. 1, the image tracing information security protection method may include the following steps:
and step S100, acquiring an original image of the watermark tracing information to be superimposed and the image watermark to be superimposed.
Step S110, according to preset image watermark characteristic parameter constraint conditions, characteristic parameters of the preset superimposed image watermarks are adjusted to obtain an adjusted image watermark set; wherein the characteristic parameter includes a color parameter at a pixel level, and the characteristic parameter constraint includes an adjustment range of the characteristic parameter.
By way of example, the predetermined superimposed image watermark may include, but is not limited to, a picture, a string, etc.
In the embodiment of the application, in order to improve the performance of removing attack resistance of the image watermark in the watermark image, the characteristic parameters of the obtained image watermark with predetermined superposition can be adjusted before the obtained image watermark with the original image is superimposed.
In addition, in order to refine the adjustment granularity of the characteristic parameters of the image watermark, the adjustment of the characteristic parameters of the image watermark is not limited to the adjustment of the whole image watermark, but the adjustment of the color parameters of the image watermark can be performed at the pixel level, i.e. the color parameters of each pixel of the image watermark can be respectively adjusted.
By way of example, color parameters may include, but are not limited to, some or all of color parameters and transparency.
By way of example, color parameters may include, but are not limited to, RGB (Red, green, blue, red Blue Green) values, YUV (a color coding scheme) values, ycbCr (a color coding scheme) values, or the like.
In this embodiment of the present application, an image watermark feature parameter constraint condition, that is, an adjustment range of an image watermark feature parameter may be preset, and in a process of adjusting a feature parameter of an image watermark, the feature parameter of a predetermined superimposed image watermark needs to be adjusted according to the preset image watermark parameter constraint condition, that is, the feature parameter of the predetermined superimposed image watermark is adjusted in a preset feature parameter adjustment range, and a value of the feature parameter of the adjusted image watermark is controlled to be within a preset value range.
For example, regarding transparency adjustment, when adjusting the transparency of any pixel of the image watermark, it is necessary to ensure that the transparency of the pixel after adjustment is within a preset range, for example, greater than or equal to a preset transparency threshold.
For example, for a predetermined superimposed image watermark, a plurality of different adjusted image watermarks may be obtained according to the above-mentioned characteristic parameter adjustment manner. The characteristic parameters of different adjusted image watermarks are different.
Optionally, in an embodiment of the present application, the image watermarks in the adjusted image watermark set may also include a predetermined superimposed image watermark.
And step S120, respectively superposing each image watermark in the adjusted image watermark set with the original image to obtain a watermark image set to be tested.
In this embodiment of the present application, when the adjusted image watermark set is obtained in the manner described in the foregoing embodiment, for any image watermark in the adjusted image watermark set, the image watermark may be superimposed with the original image, so as to obtain a corresponding watermark image to be tested.
And step S130, performing intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by utilizing a preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested.
In the embodiment of the application, in order to ensure the performance of the adjusted image watermark against the removal attack, a preset intelligent image information removal attack test mode can be utilized to perform intelligent watermark removal attack on each watermark image to be tested in the watermark image set to obtain an output image after the intelligent watermark removal attack corresponding to each watermark image to be tested.
The smart image information removal attack test model may be a white-box model, a gray-box model, or a black-box model, for example.
The smart image information removal attack test model may be an open source model or a commercial model, for example.
The intelligent image information removal attack test model provides an input-output interface, wherein the input can comprise a watermark image to be tested, and the output result comprises the watermark image after the intelligent watermark removal attack.
And step 140, determining an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested.
In this embodiment of the present application, when the output image corresponding to each watermark image to be tested is obtained in the manner described in the foregoing embodiment, the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested may be determined, and according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, the watermark image with the best watermark removal attack resistance (referred to herein as the optimal watermark image) and the optimal image watermark may be determined.
Illustratively, the difference between the watermark image to be tested and the corresponding output image of the watermark image to be tested may be characterized by a Peak Signal-to-Noise Ratio (PSNR) and/or a structural similarity (Structural SIMilarity, SSIM) between the watermark image to be tested and the corresponding output image.
Illustratively, the greater the PSNR between the watermark image to be tested and the corresponding output image, the less the difference between the watermark image to be tested and the corresponding output image of the watermark image to be tested.
Illustratively, the greater the structural similarity between the watermark image to be tested and the corresponding output image, the less the difference between the watermark image to be tested and the corresponding output image of the watermark image to be tested.
For example, as the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested is smaller, the change of the image is smaller after the watermark image to be tested is subjected to intelligent watermark removal attack by the intelligent image information removal attack test model, that is, the watermark image to be tested has better watermark removal attack resisting effect.
Accordingly, the optimal watermark image may be the watermark image to be tested that has the smallest difference from the corresponding output image.
It can be seen that, in the process of the method shown in fig. 1, before the original image is subjected to image watermark superposition, color parameter adjustment of pixel level is performed on the predetermined superimposed image watermark according to the preset image watermark feature parameter constraint condition, an adjusted image watermark set is obtained, each image watermark in the adjusted image watermark set is respectively superimposed with the original image, a watermark image set to be tested is obtained, and an attack test model is removed by using preset intelligent image information, intelligent watermark removal attack is performed on each watermark image to be tested in the watermark image set to be tested, so as to obtain output images corresponding to each watermark image to be tested, further, an optimal watermark image and an optimal image watermark are determined according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, the feature parameter adjustment of pixel level is performed on the predetermined superimposed image watermark according to the set constraint condition, and on the basis of modification of the trace source information of the predetermined image watermark is reduced as much as possible, the understandability and the integrity of the trace source information of the predetermined image can be ensured, and the attack of the trace source information of the image information can be effectively removed by intelligent image information, so that the trace source information of the robustness of the image source information can be greatly improved.
In some embodiments, in step S110, adjusting the characteristic parameters of the predetermined superimposed image watermark according to the preset constraint conditions of the characteristic parameters of the image watermark may include:
and adjusting color parameters of part or all pixels of the image watermark to be overlapped according to preset constraint conditions of the characteristic parameters of the image watermark.
For example, in order to improve flexibility of feature parameter adjustment of the image watermark, in the case of performing feature parameter adjustment on the image watermark, color parameters of some or all pixels of the image watermark that are predetermined to be superimposed may be adjusted according to preset constraint conditions of the feature parameter of the image watermark.
In one example, the above feature parameters may further include: some or all of the spatial parameters, morphological parameters, and quantitative parameters.
For example, in order to improve the effect of adjusting the characteristic parameters of the image watermark, the performance of watermark removal attack resistance of the watermark image corresponding to the image watermark after the characteristic parameter adjustment is optimized, and in the case of adjusting the characteristic parameters of the image watermark, the spatial parameters, the morphological parameters and some or all of the number parameters of the image watermark can be adjusted in addition to the adjustment of the pixel-level color parameters of the image watermark.
By way of example, the spatial parameters of the image watermark may include the location where the image watermark is superimposed into the image, such as the location in the image of the center coordinates of the image watermark.
Morphological parameters of the image watermark may include the size and rotation angle of the image watermark, etc.
The number of image watermarks parameter may be predetermined for the number of superimposed image watermarks.
Accordingly, the above-mentioned image watermark feature parameter constraint conditions may further include constraint conditions for some or all of the spatial parameter, the morphological parameter, and the quantity parameter.
For example, constraints on the spatial parameters include that the center coordinates lie within a particular region, etc., such as x max >x>x min And y is max >y>y min 。
Constraints on morphological parameters include that the image size meets upper and lower limits, etc., e.g., u max >u>u min And v max >v>v min . Where u may be the length of the image watermark and v may be the width of the image watermark.
The constraint of the quantity parameter comprises that the number of the image watermarks satisfies n max >n>n min 。
In the embodiment of the present application, when the number of image watermarks to be superimposed is plural, the spatial parameters, the morphological parameters, and the color parameters of each image watermark may be different.
In some embodiments, in step S110, the adjusting the characteristic parameters of the image watermark to be overlapped according to the preset constraint condition of the characteristic parameters of the image watermark to obtain the adjusted image watermark set may include:
According to the preset image watermark characteristic parameter constraint conditions and the preset optimization target, characteristic parameters of the preset superimposed image watermarks are adjusted to obtain an adjusted image watermark set meeting the preset optimization target;
wherein the preset optimization target comprises that the difference between the first watermark image and the second watermark image is smaller than a first threshold value; the first watermark image is a watermark image obtained by superposing the image watermark of the preset superposition with the original image, and the second watermark image is a watermark image obtained by superposing the image watermark of which the characteristic parameters are adjusted with the original image.
For example, in order to ensure the understandability of the image watermark after the characteristic parameter adjustment, an optimization target may be preset for the adjustment of the characteristic parameter of the image watermark, where the optimization target may constrain the overall adjustment effect of the characteristic parameter of the image watermark, so as to avoid excessive differences between watermark images obtained by superimposing the image watermarks before and after the characteristic parameter adjustment on the original image.
Correspondingly, when characteristic parameters of the image watermarks are adjusted, characteristic parameters of the image watermarks which are overlapped in a preset mode can be adjusted according to preset constraint conditions of the characteristic parameters of the image watermarks and preset optimization targets, and an adjusted image watermark set meeting the preset optimization targets is obtained.
For example, for any one of the set of adjusted image watermarks (image watermarks after adjustment of the characteristic parameters) that satisfies the preset optimization target, a difference between a watermark image (herein referred to as a second watermark image) obtained by superimposing the image watermark with the original image and a watermark image (herein referred to as a first watermark image) obtained by superimposing a predetermined superimposed image watermark (image watermark before adjustment of the characteristic parameters) with the original image is smaller than a preset threshold (herein referred to as a first threshold).
Illustratively, the difference between the first watermark image and the second watermark image may be characterized by a peak signal-to-noise ratio and/or a structural similarity between the first watermark image and the second watermark image.
In some embodiments, an intelligent watermark removal attack is performed on each watermark image to be tested in a watermark image set to be tested by using a preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested, including:
for any watermark image to be tested, respectively inputting the watermark image to be tested into N preset intelligent image information removal attack test models, and obtaining an output image after intelligent watermark removal attack corresponding to the watermark image to be tested; wherein N is more than or equal to 1.
For example, N intelligent image information removal attack test models may be preset, for performing an intelligent watermark removal attack test on the watermark image to be tested, so as to test the watermark removal attack resistance of the watermark image to be tested.
For any watermark image to be tested, the watermark image to be tested can be respectively input into N preset intelligent image information removal attack test models, and an output image after intelligent watermark removal attack corresponding to the watermark image to be tested is obtained.
For example, in view of the fact that in the practical application scenario, the model for performing watermark removal attack on the watermark image is not fixed, so, in order to better test the watermark removal attack resistance of the watermark image to be tested, the watermark removal attack resistance of the finally obtained watermark image (such as the optimal watermark image) is optimized, a plurality of intelligent image information removal attack test models can be preset, and the intelligent watermark removal attack test is performed on each watermark image to be tested by using the plurality of intelligent image information removal attack test models.
That is, preferably, N.gtoreq.2.
In some embodiments, the determining the optimal watermark image and the optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested may include:
According to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, determining the watermark image to be tested, which has the smallest difference between the watermark image to be tested and the corresponding output image and has the difference between the watermark image to be tested and the corresponding output image smaller than a second threshold value, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark.
For example, in order to ensure the watermark removal resistance performance of the image watermark after the characteristic parameter adjustment, when the watermark removal resistance performance test is performed on the watermark image to be tested by using the intelligent image information removal attack test model, the optimal image watermark may be determined from the watermark image to be tested, where the difference between the watermark image and the corresponding output image is smaller than a preset threshold (referred to herein as a second threshold).
For example, the watermark image to be tested, which has the smallest difference with the corresponding output image and the difference with the corresponding output image smaller than the second threshold, may be determined as the optimal watermark image, and the image watermark superimposed in the optimal watermark image may be determined as the optimal image watermark.
For example, the second threshold may be less than the first threshold described above.
In one example, the difference between the watermark image to be tested and the corresponding output image being less than the second threshold may be determined by:
under the condition that the preset intelligent image information removal attack test model comprises a plurality of intelligent image information removal attack test models, for any watermark image to be tested, if the difference between the watermark image to be tested and each corresponding output image is smaller than a second threshold value, determining that the difference between the watermark image to be tested and each corresponding output image is smaller than the second threshold value; and performing intelligent watermark removal attack on the watermark image to be tested through different intelligent image information removal attack test models by using different output images corresponding to the watermark image to be tested.
For example, in the case that the preset intelligent image information removal attack test model includes a plurality of intelligent image information removal attack test models, for any watermark image to be tested, the watermark image to be tested may be input into the plurality of intelligent image information removal attack test models, so as to obtain a multi-frame output image corresponding to the watermark image to be tested, and differences between the watermark image to be tested and the corresponding output images are respectively determined.
For any watermark image to be tested, if the difference between the watermark image to be tested and each corresponding output image is smaller than a second threshold, it is determined that the difference between the watermark image to be tested and each corresponding output image is smaller than the second threshold.
For example, in the case that the preset smart image information removal attack test model includes a plurality of smart image information removal attack test models, when determining a watermark image to be tested with a minimum difference from a corresponding output image, for any watermark image to be tested, the difference between the watermark image to be tested and the corresponding output image may be determined according to the difference between the watermark image to be tested and each corresponding output image.
For example, for any watermark image to be tested, an average value of the differences between the watermark image to be tested and the corresponding output images may be determined as the difference between the watermark image to be tested and the corresponding output image.
Accordingly, the minimum difference between the output images may be the average value of the differences between the output images.
For example, assuming that the predetermined smart image information removal attack test model includes three smart image information removal attack test models, for the watermark image a to be tested, the corresponding output images thereof are output images A1, A2 and A3, respectively, the difference between the watermark image a to be tested and the output image A1 is C1, the difference between the watermark image a to be tested and the output image A2 is C2, the difference between the watermark image a to be tested and the output image A3 is C3, and the difference between the watermark image a to be tested and the corresponding output image may be an average of C1, C2 and C3.
The average value may be an arithmetic average value or a weighted average value, for example.
For example, in the case that the average value is a weighted average value, the weighting coefficient of the difference between the watermark image to be tested and the corresponding output images may remove the attack performance setting of the attack test model according to the intelligent image information.
Taking the above example as an example, assuming that the three intelligent image information removal attack test models are M1, M2, and M3, respectively, the weighting coefficients of C1, C2, and C3 may be set according to the attack performance of M1, M2, and M3.
Illustratively, the more powerful the attack, the greater the weighting coefficient of the smart image information removal attack test model.
Illustratively, the attack performance of the smart image information removal attack test model may be determined according to the attack test result of the smart image information removal attack test model on the test set.
For example, for any intelligent image information removal attack test model, each test watermark image in the test set may be input to the intelligent image information removal attack test model, to obtain an output image corresponding to each test watermark image, and an arithmetic average value of differences between each test watermark image and the corresponding output image is counted, and the arithmetic average value is determined as a test difference value of the intelligent image information removal attack test model. Furthermore, the attack performance of each intelligent image information removal attack test model can be determined according to the test difference values corresponding to different intelligent image information removal attack test models.
Illustratively, the larger the corresponding test difference value, the stronger the attack performance of the intelligent image information removal attack test model.
In one implementation manner, determining the optimal watermark image and the optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested may further include:
if no watermark image to be tested, the difference between the watermark image and each corresponding output image of which is smaller than the second threshold value, is present and the iteration number does not reach the maximum iteration number, generating a new adjusted image watermark set according to a preset image watermark optimization algorithm and the corresponding output image of each watermark image to be tested, and determining the optimal watermark image and the optimal image watermark of a new round.
For example, it is considered that in an actual scene, there may be no watermark image to be tested, which has a difference from the corresponding output image smaller than the second threshold, in the watermark image to be tested obtained after the characteristic parameter adjustment is performed on the image watermark in the above manner. Therefore, in order to optimize the finally determined watermark removal attack resistance performance of the optimal watermark image, the adjusted characteristic parameter value determined when the characteristic parameter adjustment is performed on the image watermark can be iteratively optimized.
Correspondingly, for any iteration process, if no watermark image to be tested, of which the differences between the watermark image to be tested and the corresponding output images are smaller than the second threshold value, exists after the characteristic parameter adjustment is carried out on the image watermark according to the characteristic parameter value determined in the iteration process, whether the current iteration number reaches the preset maximum iteration number can be determined.
If the iteration number does not reach the maximum iteration number, generating a new adjusted image watermark set according to a preset image watermark optimization algorithm and output images corresponding to each watermark image to be tested, and determining the optimal watermark image and the optimal image watermark of a new round, namely, entering a new round of iteration (adding 1 to the iteration number).
Respectively superposing each image watermark in the new adjusted image watermark set with the original image to obtain a new watermark image set to be tested; and performing intelligent watermark removal attack on each watermark image to be tested in the new watermark image set to be tested by using a preset intelligent image information removal attack test model to obtain output images corresponding to each watermark image to be tested, determining whether the watermark image to be tested, of which the difference between the watermark image to be tested and the corresponding output image is smaller than a second threshold value, exists according to the difference between the new watermark image to be tested and the output image corresponding to the watermark image to be tested, if so, determining the watermark image to be tested, of which the difference between the watermark image to be tested and the corresponding output image is minimum, and the difference between the watermark image to be tested and the corresponding output image is smaller than the second threshold value, determining the watermark image superimposed in the optimal watermark image as the optimal watermark image.
If there is no watermark image to be tested, the difference between the watermark image and each corresponding output image of which is smaller than the second threshold value, and the iteration number reaches the maximum iteration number, determining the watermark image to be tested, of which the most intelligent image information removal attack test model attack fails, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark.
Correspondingly, for any watermark image to be tested, whether the watermark image to be tested fails the attack of the corresponding intelligent image information removal attack test model or not can be determined according to the difference between the watermark image to be tested and the corresponding output images, and the intelligent image information removal attack test model which fails the attack of the watermark image to be tested is counted among a plurality of preset intelligent image information removal attack test models.
For any watermark image to be tested and any intelligent image information removal attack test model, if the difference between the output image obtained after the watermark image to be tested is subjected to the removal attack test model by using the intelligent image information removal attack test model and the watermark image to be tested is smaller than the second threshold value, determining that the watermark image to be tested causes the intelligent image information removal attack test model to attack and fail (namely, the intelligent image information removal attack test model fails to attack the watermark image to be tested).
In order to enable those skilled in the art to better understand the technical solutions provided by the embodiments of the present application, the technical solutions provided by the embodiments of the present application are described below with reference to specific examples.
In this embodiment, aiming at the actual requirement of the security protection of the image tracing information, an image tracing information security protection method is provided, and the color parameters of the image tracing information are adjusted at the pixel level fine granularity, and a plurality of dimensions such as the space parameters, the form parameters, the quantity parameters and the like of the image watermark are considered, so that the robust image watermark resisting watermark removing attack is generated and optimized under the specific constraint condition. In the generating and optimizing process, at least one intelligent image information removal attack algorithm or model input/output interface can be combined, an iteratively generated watermark image is continuously input under a specific constraint condition, an intelligent watermark removal output result of the watermark image is obtained, and modification and optimization of the watermark image are carried out according to the fed-back output result.
The implementation flow of the image tracing information security protection method provided in this embodiment is described below.
As shown in fig. 2, the implementation flow of the image tracing information security protection method provided in this embodiment may include the following steps:
S1: and acquiring the original image of the watermark tracing information to be superimposed and the image watermark of the preset superposition.
The image watermark may be, for example, a picture, a string, etc.
S2: and obtaining characteristic parameter constraint conditions and optimization targets of the predetermined image watermark.
By way of example, the characteristic parameters may include spatial parameters, pixel level color parameters, morphological parameters, quantity parameters, and the like.
By way of example, the spatial parameters may include location information of the image watermark superimposed into the image, such as the center coordinates (x, y) of the image watermark; color parameters may include RGB values for each pixel in the image watermark and its transparency (r, g, b, α); the morphological parameters may include the size (u, v) (length and width) and rotation angle θ of the image watermark; the number parameter may comprise a predetermined number n of superimposed image watermarks.
For example, constraints on the spatial parameters include that the center coordinates lie within a particular region, etc., such as x max >x>x min And y is max >y>y min 。
Constraints on morphological parameters include that the image size meets upper and lower limits, etc., e.g., u max >u>u min And v max >v>v min . Where u may be the length of the image watermark and v may be the width of the image watermark.
The constraint of the quantity parameter comprises that the number of the image watermarks satisfies n max >n>n min 。
The image watermark optimization target includes, for example, that a difference between a watermark image obtained by superimposing the image watermark with the characteristic parameter adjusted and the original image (i.e., the second watermark image) and a watermark image obtained by superimposing the image watermark with the original image, which are predetermined to be superimposed, is smaller than a preset threshold (e.g., the first threshold).
For example, the difference between the two (i.e., the difference between the first watermark image and the second watermark image) is less than ε in norm.
Illustratively, ε is a preset value; the norms may include a 1-norm, a 2-norm, or an infinite norm.
S3: randomly generating an image watermark set (namely the adjusted image watermark set) meeting constraint conditions, and overlapping the image watermark set with an original image one by one to generate a watermark image set to be detected.
Illustratively, each adjusted image watermark is obtained by modifying a predetermined superimposed image watermark with randomly adjusted characteristic parameters.
S4: inputting each watermark image to be tested in the watermark image set to be tested into N (N is more than or equal to 1) intelligent image information removal attack test models one by one, and obtaining an output image set after intelligent watermark removal attack. If the optimization ending condition is met, outputting an optimal watermark image in the watermark image set and an image watermark corresponding to the optimal watermark image; otherwise, go to step S5.
The smart image information removal attack test model may be a white-box model, a gray-box model, or a black-box model, for example.
The smart image information removal attack test model may be an open source model or a commercial model, for example.
The intelligent image information removal attack test model provides an input-output interface, wherein the input can comprise a watermark image to be tested, and the output result comprises the watermark image after the intelligent watermark removal attack.
For example, the optimization ending condition may include that the optimization process reaches the maximum iteration number, or/and any watermark image exists in the generated watermark image set to be tested to meet the preset requirement.
For example, the preset requirements may include: any watermark image exists in the generated watermark image set to be tested, so that differences between output images of the N intelligent image information removal attack test models and input watermark images (namely watermark images to be tested) are smaller than a preset threshold (such as the second threshold).
It should be noted that, in the embodiment of the present application, in order to optimize the determined watermark removal resistance performance of the optimal watermark image as much as possible, if any watermark image in the generated watermark image set to be tested meets the preset requirement, if the iteration number does not reach the maximum iteration number, a new adjusted image watermark set may still be generated according to the preset image watermark optimization algorithm and the output image corresponding to each watermark image to be tested, and a new round of optimal watermark image and optimal image watermark determination may be performed until the iteration number reaches the maximum iteration number, and the final optimal watermark image and optimal image watermark may be determined.
S5: and calculating characteristic parameter adjustment values of the image watermarks in the set to be detected according to a preset image watermark optimization algorithm and the output image set, regenerating the watermark image set to be detected, and entering S4.
The preset image watermark optimization algorithm includes a particle swarm algorithm, a genetic algorithm, a differential evolution algorithm, a Bayesian optimization algorithm, a simulated annealing algorithm, or the like.
For example, the particle swarm algorithm may generate an optimized image watermark by iteratively updating the position and velocity of each particle within a predetermined number of populations.
The particles may be, for example, a combined representation of some or all of the characteristic parameters of the image watermark.
For example, the genetic algorithm may iteratively modify the generated image watermark continually by three main steps of selection, crossover, and mutation.
The methods provided herein are described above. The apparatus provided in this application is described below:
referring to fig. 3, a schematic structural diagram of an image tracing information security protection apparatus provided in an embodiment of the present application, as shown in fig. 3, the image tracing information security protection apparatus may include:
an obtaining unit 310, configured to obtain an original image of the watermark tracing information to be superimposed and a watermark of the image to be superimposed;
An adjusting unit 320, configured to adjust the characteristic parameters of the predetermined superimposed image watermark according to a preset constraint condition of the characteristic parameters of the image watermark, so as to obtain an adjusted image watermark set; wherein the characteristic parameters comprise color parameters of pixel level, and the characteristic parameter constraint conditions comprise adjustment ranges of the characteristic parameters;
a superimposing unit 330, configured to superimpose each image watermark in the adjusted image watermark set with the original image, so as to obtain a watermark image set to be tested;
the testing unit 340 is configured to perform an intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by using a preset intelligent image information removal attack testing model, so as to obtain an output image corresponding to each watermark image to be tested;
the determining unit 350 is configured to determine an optimal watermark image and an optimal image watermark according to a difference between the watermark image to be tested and an output image corresponding to the watermark image to be tested.
In some embodiments, the adjusting unit 320 adjusts the characteristic parameters of the predetermined superimposed image watermark according to a preset constraint condition of the characteristic parameters of the image watermark, including:
And adjusting color parameters of part or all pixels of the image watermark subjected to predetermined superposition according to preset constraint conditions of characteristic parameters of the image watermark.
In some embodiments, the characteristic parameters further comprise: some or all of the spatial parameters, morphological parameters, and quantitative parameters.
In some embodiments, the adjusting unit 320 adjusts the characteristic parameters of the predetermined superimposed image watermark according to a preset constraint condition of the characteristic parameters of the image watermark, to obtain an adjusted image watermark set, including:
according to a preset image watermark characteristic parameter constraint condition and a preset optimization target, characteristic parameters of the preset superimposed image watermarks are adjusted to obtain an adjusted image watermark set meeting the preset optimization target;
wherein the preset optimization objective comprises that the difference between the first watermark image and the second watermark image is smaller than a first threshold value; the first watermark image is a watermark image obtained by superposing the image watermark of the preset superposition with the original image, and the second watermark image is a watermark image obtained by superposing the image watermark of which the characteristic parameters are adjusted with the original image.
In some embodiments, the testing unit 340 performs an intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by using a preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested, including:
for any watermark image to be tested, respectively inputting the watermark image to be tested into N preset intelligent image information removal attack test models, and obtaining an output image after intelligent watermark removal attack corresponding to the watermark image to be tested; wherein N is more than or equal to 1.
In some embodiments, the determining unit 350 determines the optimal watermark image and the optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, including:
according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, determining the watermark image to be tested, which has the smallest difference between the watermark image to be tested and the corresponding output image and has the difference between the watermark image to be tested and the corresponding output image smaller than a second threshold value, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark.
In some embodiments, the testing unit 340 determines that the difference between the watermark image to be tested and the corresponding output image is less than a second threshold by:
under the condition that the preset intelligent image information removal attack test model comprises a plurality of intelligent image information removal attack test models, for any watermark image to be tested, if the difference between the watermark image to be tested and each corresponding output image is smaller than a second threshold value, determining that the difference between the watermark image to be tested and each corresponding output image is smaller than the second threshold value; and performing intelligent watermark removal attack on the watermark image to be tested through different intelligent image information removal attack test models by using different output images corresponding to the watermark image to be tested.
In some embodiments, the determining unit 350 determines the optimal watermark image and the optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, and further includes:
if no watermark image to be tested, the difference between the watermark image and each corresponding output image of which is smaller than the second threshold value, is present and the iteration number does not reach the maximum iteration number, generating a new adjusted image watermark set according to a preset image watermark optimization algorithm and the corresponding output image of each watermark image to be tested, and determining the optimal watermark image and the optimal image watermark of a new round.
In some embodiments, the determining unit 350 determines the optimal watermark image and the optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, and further includes:
if no watermark image to be tested, the difference between the watermark image and each corresponding output image of which is smaller than a second threshold value, exists, and the iteration number reaches the maximum iteration number, determining the watermark image to be tested, of which the most intelligent image information removal attack test model fails, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark;
and if the difference between the output image obtained after the intelligent image information removal attack test model is used for removing the attack test model from the watermark image to be tested and the watermark image to be tested is smaller than the second threshold value, determining that the watermark image to be tested causes the intelligent image information removal attack test model to fail the attack.
The embodiment of the application provides electronic equipment, which comprises a processor and a memory, wherein the memory stores machine executable instructions capable of being executed by the processor, and the processor is used for executing the machine executable instructions to realize the image tracing information safety protection method.
Fig. 4 is a schematic hardware structure of an electronic device according to an embodiment of the present application. The electronic device may include a processor 401, a memory 402 storing machine-executable instructions. The processor 401 and the memory 402 may communicate via a system bus 403. Also, the processor 401 may perform the image trace information security protection method described above by reading and executing machine executable instructions in the memory 402 corresponding to the image trace information security protection logic.
The memory 402 referred to herein may be any electronic, magnetic, optical, or other physical storage device that may contain or store information, such as executable instructions, data, or the like. For example, a machine-readable storage medium may be: RAM (Radom Access Memory, random access memory), volatile memory, non-volatile memory, flash memory, a storage drive (e.g., hard drive), a solid state drive, any type of storage disk (e.g., optical disk, dvd, etc.), or a similar storage medium, or a combination thereof.
In some embodiments, a storage medium, such as memory 402 in fig. 4, is also provided, where the storage medium is a machine-readable storage medium having stored therein machine-executable instructions that, when executed by a processor, implement the image trace information security protection method described above. For example, the storage medium may be ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing description of the preferred embodiments of the present invention is not intended to limit the invention to the precise form disclosed, and any modifications, equivalents, improvements and alternatives falling within the spirit and principles of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. The image tracing information safety protection method is characterized by comprising the following steps of:
acquiring an original image of watermark tracing information to be superimposed and a predetermined superimposed image watermark; the image watermark is a visible watermark;
adjusting characteristic parameters of the preset superimposed image watermarks according to preset constraint conditions of the characteristic parameters of the image watermarks to obtain an adjusted image watermark set; wherein the characteristic parameters comprise color parameters of pixel level, and the characteristic parameter constraint conditions comprise adjustment ranges of the characteristic parameters;
respectively superposing each image watermark in the adjusted image watermark set with the original image to obtain a watermark image set to be tested;
performing intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by using a preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested;
determining an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and an output image corresponding to the watermark image to be tested;
the adjusting the characteristic parameters of the image watermarks of the preset superposition according to the preset constraint conditions of the characteristic parameters of the image watermarks to obtain an adjusted image watermark set, which comprises the following steps:
According to a preset image watermark characteristic parameter constraint condition and a preset optimization target, characteristic parameters of the preset superimposed image watermarks are adjusted to obtain an adjusted image watermark set meeting the preset optimization target;
wherein the preset optimization objective comprises that the difference between the first watermark image and the second watermark image is smaller than a first threshold value; the first watermark image is a watermark image obtained by superposing the image watermark of the preset superposition with the original image, and the second watermark image is a watermark image obtained by superposing the image watermark of which the characteristic parameters are adjusted with the original image;
the determining the optimal watermark image and the optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested comprises the following steps:
according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, determining the watermark image to be tested, which has the smallest difference between the watermark image to be tested and the corresponding output image and has the difference between the watermark image to be tested and the corresponding output image smaller than a second threshold value, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark.
2. The method according to claim 1, wherein said adjusting the characteristic parameters of the predetermined superimposed image watermark according to the preset image watermark characteristic parameter constraint conditions comprises:
and adjusting color parameters of part or all pixels of the image watermark subjected to predetermined superposition according to preset constraint conditions of characteristic parameters of the image watermark.
3. The method according to claim 1 or 2, wherein the characteristic parameters further comprise: some or all of the spatial parameters, morphological parameters, and quantitative parameters.
4. The method according to claim 1, wherein the performing the intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by using the preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested includes:
for any watermark image to be tested, respectively inputting the watermark image to be tested into N preset intelligent image information removal attack test models, and obtaining an output image after intelligent watermark removal attack corresponding to the watermark image to be tested; wherein N is more than or equal to 1.
5. The method according to claim 1, wherein the difference between the watermark image to be tested and the corresponding output image is less than a second threshold value is determined by:
Under the condition that the preset intelligent image information removal attack test model comprises a plurality of intelligent image information removal attack test models, for any watermark image to be tested, if the difference between the watermark image to be tested and each corresponding output image is smaller than a second threshold value, determining that the difference between the watermark image to be tested and each corresponding output image is smaller than the second threshold value; and performing intelligent watermark removal attack on the watermark image to be tested through different intelligent image information removal attack test models by using different output images corresponding to the watermark image to be tested.
6. The method of claim 5, wherein determining the optimal watermark image and the optimal image watermark based on the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, further comprises:
if no watermark image to be tested, the difference between the watermark image and each corresponding output image of which is smaller than the second threshold value, is present and the iteration number does not reach the maximum iteration number, generating a new adjusted image watermark set according to a preset image watermark optimization algorithm and the corresponding output image of each watermark image to be tested, and determining the optimal watermark image and the optimal image watermark of a new round.
7. The method of claim 5, wherein determining the optimal watermark image and the optimal image watermark based on the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, further comprises:
if no watermark image to be tested, the difference between the watermark image and each corresponding output image of which is smaller than a second threshold value, exists, and the iteration number reaches the maximum iteration number, determining the watermark image to be tested, of which the most intelligent image information removal attack test model fails, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark;
and if the difference between the output image obtained after the intelligent image information removal attack test model is used for removing the attack test model from the watermark image to be tested and the watermark image to be tested is smaller than the second threshold value, determining that the watermark image to be tested causes the intelligent image information removal attack test model to fail the attack.
8. An image traceability information safety device, which is characterized by comprising:
The acquisition unit is used for acquiring an original image of the watermark tracing information to be superimposed and the image watermark of the preset superposition; the image watermark is a visible watermark;
the adjusting unit is used for adjusting the characteristic parameters of the preset superimposed image watermarks according to the preset constraint conditions of the characteristic parameters of the image watermarks to obtain an adjusted image watermark set; wherein the characteristic parameters comprise color parameters of pixel level, and the characteristic parameter constraint conditions comprise adjustment ranges of the characteristic parameters;
the superposition unit is used for respectively superposing each image watermark in the adjusted image watermark set and the original image to obtain a watermark image set to be tested;
the testing unit is used for performing intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by utilizing a preset intelligent image information removal attack testing model to obtain an output image corresponding to each watermark image to be tested;
the determining unit is used for determining an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested;
the adjusting unit adjusts the characteristic parameters of the image watermarks of the preset stack according to the constraint condition of the characteristic parameters of the image watermarks of the preset stack to obtain an adjusted image watermark set, and the adjusting unit comprises the following steps:
According to a preset image watermark characteristic parameter constraint condition and a preset optimization target, characteristic parameters of the preset superimposed image watermarks are adjusted to obtain an adjusted image watermark set meeting the preset optimization target;
wherein the preset optimization objective comprises that the difference between the first watermark image and the second watermark image is smaller than a first threshold value; the first watermark image is a watermark image obtained by superposing the image watermark of the preset superposition with the original image, and the second watermark image is a watermark image obtained by superposing the image watermark of which the characteristic parameters are adjusted with the original image;
the determining unit determines an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and an output image corresponding to the watermark image to be tested, and the determining unit comprises:
according to the difference between the watermark image to be tested and the output image corresponding to the watermark image to be tested, determining the watermark image to be tested, which has the smallest difference between the watermark image to be tested and the corresponding output image and has the difference between the watermark image to be tested and the corresponding output image smaller than a second threshold value, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark.
9. The apparatus according to claim 8, wherein the adjusting unit adjusts the characteristic parameters of the predetermined superimposed image watermark according to a preset constraint condition of the characteristic parameters of the image watermark, including:
according to preset image watermark characteristic parameter constraint conditions, color parameters of partial or all pixels of the preset superimposed image watermark are adjusted;
wherein the characteristic parameters further include: some or all of the spatial parameters, morphological parameters, and quantitative parameters;
the test unit performs intelligent watermark removal attack on each watermark image to be tested in the watermark image set to be tested by using a preset intelligent image information removal attack test model to obtain an output image corresponding to each watermark image to be tested, and the method comprises the following steps:
for any watermark image to be tested, respectively inputting the watermark image to be tested into N preset intelligent image information removal attack test models, and obtaining an output image after intelligent watermark removal attack corresponding to the watermark image to be tested; wherein N is more than or equal to 1;
wherein the test unit determines that the difference between the watermark image to be tested and the corresponding output image is less than a second threshold by:
Under the condition that the preset intelligent image information removal attack test model comprises a plurality of intelligent image information removal attack test models, for any watermark image to be tested, if the difference between the watermark image to be tested and each corresponding output image is smaller than a second threshold value, determining that the difference between the watermark image to be tested and each corresponding output image is smaller than the second threshold value; different output images corresponding to the watermark image to be tested are obtained by performing intelligent watermark removal attack on the watermark image to be tested through different intelligent image information removal attack test models;
the determining unit determines an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and an output image corresponding to the watermark image to be tested, and the determining unit further comprises:
if no watermark image to be tested, the difference between which and each corresponding output image is smaller than a second threshold value, is not present and the iteration number does not reach the maximum iteration number, generating a new adjusted image watermark set according to a preset image watermark optimization algorithm and each corresponding output image of the watermark image to be tested, and determining a new round of optimal watermark image and an optimal image watermark;
The determining unit determines an optimal watermark image and an optimal image watermark according to the difference between the watermark image to be tested and an output image corresponding to the watermark image to be tested, and the determining unit further comprises:
if no watermark image to be tested, the difference between the watermark image and each corresponding output image of which is smaller than a second threshold value, exists, and the iteration number reaches the maximum iteration number, determining the watermark image to be tested, of which the most intelligent image information removal attack test model fails, as an optimal watermark image, and determining the superimposed image watermark in the optimal watermark image as an optimal image watermark;
and if the difference between the output image obtained after the intelligent image information removal attack test model is used for removing the attack test model from the watermark image to be tested and the watermark image to be tested is smaller than the second threshold value, determining that the watermark image to be tested causes the intelligent image information removal attack test model to fail the attack.
10. An electronic device comprising a processor and a memory, the memory storing machine executable instructions executable by the processor for executing the machine executable instructions to implement the method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211095280.XA CN115187444B (en) | 2022-09-08 | 2022-09-08 | Image tracing information safety protection method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211095280.XA CN115187444B (en) | 2022-09-08 | 2022-09-08 | Image tracing information safety protection method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115187444A CN115187444A (en) | 2022-10-14 |
| CN115187444B true CN115187444B (en) | 2023-05-26 |
Family
ID=83523142
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211095280.XA Active CN115187444B (en) | 2022-09-08 | 2022-09-08 | Image tracing information safety protection method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115187444B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104766263A (en) * | 2014-12-20 | 2015-07-08 | 辽宁师范大学 | Color image watermark embedding and detecting method based on quaternion Legendre moment correction |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4042887B2 (en) * | 1999-08-04 | 2008-02-06 | Ｋｄｄｉ株式会社 | Video watermarking device |
| US7616776B2 (en) * | 2005-04-26 | 2009-11-10 | Verance Corproation | Methods and apparatus for enhancing the robustness of watermark extraction from digital host content |
| CN1306453C (en) * | 2003-12-30 | 2007-03-21 | 中国人民解放军第一军医大学 | Image water print embedding and detecting method of strong antiimpact performance in discrete conversion region |
| CN103854249A (en) * | 2013-12-28 | 2014-06-11 | 辽宁师范大学 | Digital image watermarking method based on local index torque characteristic |
| CN104281993A (en) * | 2014-07-29 | 2015-01-14 | 山东科技大学 | Rotary attack resisting digital watermarking method based on visual encrypting and feature point matching |
| CN108648132B (en) * | 2018-04-16 | 2020-08-14 | 深圳市联软科技股份有限公司 | Method, system, terminal and medium for generating watermark according to image |
| CN110599388B (en) * | 2019-08-26 | 2021-10-15 | 华中科技大学 | Blind robust digital watermark embedding and detecting method based on positioning point assistance |
| CN113496449A (en) * | 2020-03-20 | 2021-10-12 | 阿里巴巴集团控股有限公司 | Data processing method and device, electronic equipment and storage equipment |
-
2022
- 2022-09-08 CN CN202211095280.XA patent/CN115187444B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104766263A (en) * | 2014-12-20 | 2015-07-08 | 辽宁师范大学 | Color image watermark embedding and detecting method based on quaternion Legendre moment correction |
Non-Patent Citations (2)
| Title |
|---|
| Optimal periodic watermarking schedule for replay attack detection in cyber–physical systems;Chongrong Fang et al.;《Automatica》;第112卷;全文 * |
| 一种可抗几何攻击的二值文本图像数字水印算法;李京兵;;海南大学学报(自然科学版)(第02期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115187444A (en) | 2022-10-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gou et al. | Noise features for image tampering detection and steganalysis | |
| Gragnaniello et al. | Analysis of adversarial attacks against CNN-based image forgery detectors | |
| CN110232650B (en) | Color image watermark embedding method, detection method and system | |
| JP4398943B2 (en) | Digital watermark detection apparatus, digital watermark detection method, and digital watermark detection program | |
| CN104637068B (en) | Frame of video and video pictures occlusion detection method and device | |
| CN110390643B (en) | License plate enhancement method and device and electronic equipment | |
| Taspinar et al. | PRNU-based camera attribution from multiple seam-carved images | |
| CN105828081B (en) | encoding method and encoding device | |
| CN112700363B (en) | Self-adaptive visual watermark embedding method and device based on region selection | |
| Moghaddam et al. | A robust color image watermarking technique using modified imperialist competitive algorithm | |
| US10909649B2 (en) | Method and apparatus for removing hidden data based on autoregressive generative model | |
| CN110727928B (en) | 3D video copyright comprehensive protection method based on deep reinforcement learning optimization | |
| Mehrish et al. | Joint spatial and discrete cosine transform domain-based counter forensics for adaptive contrast enhancement | |
| Juarez-Sandoval et al. | Digital image ownership authentication via camouflaged unseen-visible watermarking | |
| CN115187444B (en) | Image tracing information safety protection method and device and electronic equipment | |
| Qi et al. | An adaptive visible watermark embedding method based on region selection | |
| CN108171643B (en) | Robust reversible image watermarking method based on histogram shift and application | |
| CN107292803B (en) | Reversible robust digital image watermarking method according with human eye observation rule | |
| CN114881838B (en) | Bidirectional face data protection method, system and equipment for deep forgery | |
| CN113628091B (en) | Safety information extraction method and device for electronic display screen content reproduction scene | |
| CN114359009B (en) | Watermark embedding method, watermark embedding network construction method, system and storage medium for robust image based on visual perception | |
| Nataraj et al. | Holistic image manipulation detection using pixel co-occurrence matrices | |
| CN112019700B (en) | Method for preventing secret-carrying image from being detected, intelligent terminal and storage medium | |
| CN114245116A (en) | Color image steganography method based on edge enhancement and channel correlation algorithm | |
| Kurniawan et al. | Hiding secret data by using modulo function in quad difference expansion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |