CN104902341A - Method for resisting geometric attack of safe digital video watermark - Google Patents
Method for resisting geometric attack of safe digital video watermark Download PDFInfo
- Publication number
- CN104902341A CN104902341A CN201510268912.1A CN201510268912A CN104902341A CN 104902341 A CN104902341 A CN 104902341A CN 201510268912 A CN201510268912 A CN 201510268912A CN 104902341 A CN104902341 A CN 104902341A
- Authority
- CN
- China
- Prior art keywords
- watermark
- digital video
- frame
- watermarking
- video
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/835—Generation of protective data, e.g. certificates
- H04N21/8358—Generation of protective data, e.g. certificates involving watermark
Abstract
The embodiment of the invention discloses a method for resisting geometric attack of a safe digital video watermark. The method comprises the steps as follows: embedding the digital video watermark in average alternating current energy AAE in the process of embedding the digital video watermark; using wavelet decomposition of an excellent space frequency decomposition property, detecting whether synchronization exists frame by frame in the process of extracting the digital video watermark, calculating whether frame number between adjacent synchronizations is correct if the synchronization exists, extracting the watermark if the frame number between adjacent synchronizations is correct, not extracting if the frame number between adjacent synchronizations is not correct. The method for resisting the geometric attack of the safe digital video watermark of the invention could effectively resist attack of geometric deformation by embedding a watermark signal in the AAE, thereby improving robustness of the watermark and protecting copyright of video content.
Description
Technical field
The present invention relates to field of computer technology, especially relate to a kind of method of secure digital video watermark of resist geometric attacks.
Background technology
Now, the piracy of digital video content is more and more easier, also more and more serious.Copy can make traditional analog media (program as TV is play) produce content improve quality on loss, and digital video content can copy many times, and every a copy all can not cause the decline of quality, and marginal cost is almost nil.Popularizing of PC, the Internet, makes unwarranted video sharing more prevalent.Therefore, copyright side can on oneself video content embed digital watermark, to protect copyright rights whatsoever.
Current video watermarking algorithm to such as video compression, add and to make an uproar and the general process such as filtering all has good robustness, but also not ideal for geometric attack performance.The attack of geometric deformation is that malicious attacker utilizes image processing means to carry out the amendment of geometry to the image or video that added watermark, as rotated image, translation, the conversion of flexible equiaffine or random geometry conversion.After such attack, be lose synchronized relation between the image after attack and watermark signal when generation dispute over copyright needs to extract watermark, make both correlations very low, thus cause detecting unsuccessfully.Nearest research shows, even very faint geometric transformation, also can destroy most watermarking project, and therefore, the attack of how effectively to resist geometric deformation is one of the focus and difficult point of current digital watermarking research.
Meanwhile, this type of is attacked often along with the time shaft synchronization attack method for digital video watermarking, as frame deletion, frame insertion etc., simple to operate, but still destroys greatly watermark information.
Existing scheme, can be referred to as geometry inverse transformation method, and after geometric deformation, watermark signal is still present among image, synchronously there occurs change only between watermark signal with original image, causes and detects unsuccessfully.Be not difficult thus to imagine, if we can know the geometric transformation that assailant carries out, so just the image inverse transformation after attack can be returned when watermark detection, thus extract watermark.This kind of algorithm, according to the need of original image, is divided into half-blindness correlation method and blind correlation method.
Semiblind detecting algorithm by original image when detecting watermark, judges experienced geometric transformation according to the corresponding relation between original image and the characteristic point of the image after attacking or makes up with original image the distortion that geometric transformation (shear convergent-divergent etc.) causes.
Blind correlation method is the main flow of the watermarking algorithm of resist geometric attacks, when resisting geometric attack and detecting watermark, does not need original image.Some person proposes a kind of simultaneous techniques based on template, for affine transformation, by the specific template embedded in watermarked process, utilizes the affine transformation that this template prediction suffers in testing process.These templates are made up of frequency peak usually, and the position of frequency peak forms special shape, as square in circle etc., utilize the change prediction affine transformation of template peak position in testing process.
For the former, in fact, also may not necessarily solve watermark synchronization problem by original image, particularly when image simultaneously Rotation and Zoom or when experiencing print scanned process, moreover if need original image when detection, this can increase carrying cost undoubtedly.
And for the latter, the shortcoming of this method is that template is easy to by filtering; Someone proposes a kind of self-reference method, same watermark is embedded respectively in the position that image is different, the auto-correlation function of first computational prediction watermark in testing process, the repeatedly embedding due to watermark can make result occur multiple peak value, utilizes peak to change and can predict affine transformation.But this watermarking algorithm can resisting rotary extension translation preferably and change overall geometric attack and the attacks of print_scan such as length-width ratio, but can not resist the attack such as filtering and mirror image well.
Summary of the invention
This patent is for the problems referred to above, propose a kind of algorithm of new video digital watermark, utilize the time-space domain geometric properties of video sequence, for the invariant (average AC energy) of geometry deformation, realize the synchronous and orientation correction of the time shaft of video sequence, effectively can resist attack.
In order to solve the problem, the present invention proposes a kind of method of secure digital video watermark of resist geometric attacks, comprising the steps:
In the process that digital video watermarking embeds, digital video watermarking is embedded in average AC energy AAE; And adopt the wavelet decomposition of excellent empty resolution characteristic frequently;
Whether, in the process that digital video watermarking extracts, detect frame by frame and synchronously exist, if fruit exists, whether the frame number calculated between adjacent sync is correct, if correct, extract watermark, if incorrect, do not extract.
Describedly digital video watermarking is embedded into average AC energy AAE and comprises:
Embed synchronous while embedding Meaningful watermarking in the video sequence, and by synchronous and the overlapping embedding of Meaningful watermarking;
Watermark signal is embedded in the AAE sequence of video, revised the alternating component of each frame by watermark signal.
The wavelet decomposition of the excellent empty resolution characteristic frequently of described employing comprises:
After amendment coefficient terminates, by wavelet reconstruction, obtain the picture frame of spatial domain, thus complete the embedding of this frame watermark signal.
Whether described detection frame by frame synchronously exists comprises:
Utilize the average AC energy sequence calculating this video, produce the pseudo random sequence representing watermark bit and synchronization bit according to the password in copyright owner's hand simultaneously, calculate average AC energy sequence and the cross-correlation between pseudo random sequence to detect watermark and synchronously whether to exist.
By implementing the embodiment of the present invention, due to AAE (the average AC energy of image, i.e. average AC energy) remain unchanged under the attack of most of geometric deformation, even if there occurs change in some situation (part as new images is filled), also can keep a kind of approximate linear relationship, affect not quite for detection.And such attack is more rare, because can expose by the sign attacked.Therefore, by watermark signal is embedded in AAE, be highly resistant to the attack of geometric deformation, improve watermark robustness, protection digital video content copyright.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, be briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is the method flow diagram that the digital video watermarking in the embodiment of the present invention embeds;
Fig. 2 is the wavelet coefficient schematic diagram to be modified in the embodiment of the present invention;
Fig. 3 is the method flow diagram that the digital video watermarking in the embodiment of the present invention extracts.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, be clearly and completely described the technical scheme in the embodiment of the present invention, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making the every other embodiment obtained under creative work prerequisite, belong to the scope of protection of the invention.
In embodiment of the present invention process, in the process that digital video watermarking embeds, digital video watermarking is embedded in average AC energy AAE; And adopt the wavelet decomposition of excellent empty resolution characteristic frequently; Whether, in the process that digital video watermarking extracts, detect frame by frame and synchronously exist, if fruit exists, whether the frame number calculated between adjacent sync is correct, if correct, extract watermark, if incorrect, do not extract.
The pixel of the t frame of setting video sequence is f
t(x, y).After geometric transformation, its position can change, but its image intensity value can not change, that is:
f
t(x，y)＝f
t(x′，y′)
Due to
the gross energy of the t frame entire image of video sequence,
for the energy of the DC component of image, wherein K, L are the length of image and wide, then can the average AC energy of this image be:
AAE has after geometric deformation, if the length and width of image are constant, then and the character remained unchanged.And, if image stretches, still can remain unchanged.
Fig. 1 shows the method flow diagram that the digital video watermarking in the embodiment of the present invention embeds, in order to watermark information can be obtained from any frame, the malicious attacks such as frame are gone in opposing simultaneously, embed synchronous while embedding Meaningful watermarking in the video sequence, and will the synchronous and overlapping embedding of Meaningful watermarking.
In practical solution, synchronizing signal is a Bit String preset.Any watermark and the synchronous Bit String all can regarding 1 and 0 composition as, also just can regard the bit sequence of 1 and-1 composition as, with (b
1.b
2... b
i... .) represent.We use spread spectrum mode to carry out these bit strings hiding.Pseudo random sequence generator in Figure 1 is used for generation two pseudo random sequences, one for representing the information bit of watermark, one for representing synchronization bit, the generation key of two pseudo random sequences and generate length can be different.
These two pseudo random sequences are Gauss white noises in statistical significance, consider that malicious attacker can when affecting little on vision signal, the process such as the low-pass filtering of time-axis direction are utilized to remove watermark signal, this programme does not directly adopt white noise, but first carries out low-pass filtering to white noise.Consider that time dimension human visual system is in this characteristic that declines lower than sensitiveness during 1Hz, the cut-off frequency of the low pass filter designed by the embodiment of the present invention is greatly about about 1Hz.
Structure due to the pseudo random sequence of synchronization bit and the pseudo random sequence of watermark bit embed and the process such as extraction similar, therefore unifiedly below represent pseudo random sequence with W (t).For simplifying follow-up coherent detection, requiring to there is zero-mean, i.e. EW (t)=0 when constructing W (t), and the embedment strength of watermark can be controlled by controlling its amplitude peak Am.
Watermark signal W (t) embeds in AAE sequence v (t) of video by the telescopiny of watermark exactly, is namely revised the alternating component of each frame by watermark signal.For this reason, for the picture frame of 128 × 128 sizes, biorthogonal wavelet is used to carry out 7 layers of wavelet decomposition.According to Fig. 1, for certain information bit bi, its watermarked process, is modulated on v (t) by the pseudo random sequence W (t) of watermark, that is: exactly
v
w(t)＝v(t)+b
i*W(t)
After wavelet transformation, when revising AC energy, consider the anti-attack ability (do not select the LL7 comprising DC energy, do not select most high frequency coefficient to resist the attack of low pass spatial filtering simultaneously) of system, the AC coefficient only choosing 2nd ~ 7 layers is modified.The scope of amendment as shown in Figure 2.In figure, dash area is the modifier area chosen.
The embodiment of the present invention chooses wavelet decomposition, is because wavelet decomposition has excellent sky-frequency resolution characteristic.In the region that spatial domain changes greatly, the absolute value of the coefficient of wavelet decomposition of its correspondence is also comparatively large, and the amendment that these aspects carry out is also larger; Otherwise, carry out less amendment in the less region of change.And the research of human-eye visual characteristic is found, human eye is lower than simple region for the sensitivity of complex texture regional change in image.Therefore, this modification is consistent with the visual characteristic of human eye.After amendment coefficient terminates, by wavelet reconstruction, obtain the picture frame of spatial domain, thus complete the embedding of this frame watermark signal.After all frames of video embed, also just complete the watermark embedment to this video.
Fig. 3 shows the method flow diagram that digital video watermarking in the embodiment of the present invention extracts, and the extraction of watermark need not from the 1st frame of video, can from any frame of video.Detector detects synchronous existence first frame by frame, if from current point until terminate all to can't detect synchronously, then prove to there is not watermark in this video.If detect synchronous, then calculate adjacent two synchronous between frame number whether correct.If correct, then extract watermark; Otherwise then may suffer to go frame or increase the attacks such as frame, this part inspection does not measure correct watermark, need to check below two synchronously, the like.Synchronous detect and watermark bit extraction in, all take the method for cross-correlation.
First the average AC energy sequence v calculating this video is utilized
wt (), produces the pseudo random sequence W (t) representing watermark bit and synchronization bit simultaneously according to the password in copyright owner's hand, calculate v
wt whether the cross-correlation between () with W (t) detects watermark and synchronously to exist, as shown in the formula:
R
v，w(0)＝E[v(t)W(t)]+b
iE[W(t)
2]
＝R
v，w(0)+b
iE[W(t)
2]
Therefore, if video does not have watermarked, then above-mentioned is 0; Otherwise, equal.Therefore we can by arranging threshold value E [W (t)
2] judge the existence of watermark or synchronization bit, and judge 0 or 1 according to sign, thus extract watermark.
There is the resist geometric attacks Video Watermarking Scheme based on human visual system (Human Vision System, HVS).The method first carries out standardization processing to original video object, recording synchronism information, is similar to above-mentioned semiblind detecting.When watermark embedment, according to human visual system, from object video, choose high details area watermarked.In watermark detection with before extracting, standardization processing is carried out to the watermark object video that contains under fire, adjusts the position of some jitter points, to improve the performance of watermark detection further.
To sum up, due to AAE (the average AC energy of image, i.e. average AC energy) remain unchanged under the attack of most of geometric deformation, even if there occurs change in some situation (part as new images is filled), also can keep a kind of approximate linear relationship, affect not quite for detection.And such attack is more rare, because can expose by the sign attacked.Therefore, by watermark signal is embedded in AAE, be highly resistant to the attack of geometric deformation, improve watermark robustness, protection digital video content copyright.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is that the hardware that can carry out instruction relevant by program has come, this program can be stored in a computer-readable recording medium, storage medium can comprise: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), disk or CD etc.
In addition, above the method for the secure digital video watermark of the resist geometric attacks that the embodiment of the present invention provides is described in detail, apply specific case herein to set forth principle of the present invention and execution mode, the explanation of above embodiment just understands method of the present invention and core concept thereof for helping; Meanwhile, for one of ordinary skill in the art, according to thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (4)
1. a method for the secure digital video watermark of resist geometric attacks, is characterized in that, comprise the steps:
In the process that digital video watermarking embeds, digital video watermarking is embedded in average AC energy AAE; And adopt the wavelet decomposition of excellent empty resolution characteristic frequently;
Whether, in the process that digital video watermarking extracts, detect frame by frame and synchronously exist, if fruit exists, whether the frame number calculated between adjacent sync is correct, if correct, extract watermark, if incorrect, do not extract.
2. the method for the secure digital video watermark of resist geometric attacks as claimed in claim 1, is characterized in that, describedly digital video watermarking is embedded into average AC energy AAE and comprises:
Embed synchronous while embedding Meaningful watermarking in the video sequence, and by synchronous and the overlapping embedding of Meaningful watermarking;
Watermark signal is embedded in the AAE sequence of video, revised the alternating component of each frame by watermark signal.
3. the method for the secure digital video watermark of resist geometric attacks as claimed in claim 2, is characterized in that, the wavelet decomposition of the excellent empty resolution characteristic frequently of described employing comprises:
After amendment coefficient terminates, by wavelet reconstruction, obtain the picture frame of spatial domain, thus complete the embedding of this frame watermark signal.
4. the method for the secure digital video watermark of the resist geometric attacks as described in any one of claims 1 to 3, is characterized in that, whether described detection frame by frame synchronously exists comprises:
Utilize the average AC energy sequence calculating this video, produce the pseudo random sequence representing watermark bit and synchronization bit according to the password in copyright owner's hand simultaneously, calculate average AC energy sequence and the cross-correlation between pseudo random sequence to detect watermark and synchronously whether to exist.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510268912.1A CN104902341A (en) | 2015-05-22 | 2015-05-22 | Method for resisting geometric attack of safe digital video watermark |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510268912.1A CN104902341A (en) | 2015-05-22 | 2015-05-22 | Method for resisting geometric attack of safe digital video watermark |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104902341A true CN104902341A (en) | 2015-09-09 |
Family
ID=54034691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510268912.1A Pending CN104902341A (en) | 2015-05-22 | 2015-05-22 | Method for resisting geometric attack of safe digital video watermark |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104902341A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101489133A (en) * | 2009-01-16 | 2009-07-22 | 华中科技大学 | Geometric attack resisting real-time video watermarking method |
| US20100146282A1 (en) * | 2006-09-29 | 2010-06-10 | Isao Echizen | Dynamic image content tamper detecting device and system |
| CN102892048A (en) * | 2012-09-18 | 2013-01-23 | 天津大学 | Video watermark anti-counterfeiting method capable of resisting geometric attacks |
| CN103237271A (en) * | 2013-04-02 | 2013-08-07 | 北京中视广信科技有限公司 | Method and system for embedding and detecting large-capacity video digital watermarks |
-
2015
- 2015-05-22 CN CN201510268912.1A patent/CN104902341A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100146282A1 (en) * | 2006-09-29 | 2010-06-10 | Isao Echizen | Dynamic image content tamper detecting device and system |
| CN101489133A (en) * | 2009-01-16 | 2009-07-22 | 华中科技大学 | Geometric attack resisting real-time video watermarking method |
| CN102892048A (en) * | 2012-09-18 | 2013-01-23 | 天津大学 | Video watermark anti-counterfeiting method capable of resisting geometric attacks |
| CN103237271A (en) * | 2013-04-02 | 2013-08-07 | 北京中视广信科技有限公司 | Method and system for embedding and detecting large-capacity video digital watermarks |
Non-Patent Citations (2)
| Title |
|---|
| 刘九芬，F王振武，黄达人: "抗几何攻击的小波变换域图像水印算法", 《浙江大学学报 工学版》 * |
| 赵耀: "基于小波变换的抵抗几何攻击的", 《中国科学E辑 信息科学》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Generic lossless visible watermarking—a new approach | |
| Kutter et al. | Digital watermarking of color images using amplitude modulation | |
| CN100586181C (en) | Embedding and detecting method and system for image data watermark information | |
| CN102147912B (en) | Adaptive difference expansion-based reversible image watermarking method | |
| CN102903075B (en) | Robust watermarking method based on image feature point global correction | |
| CN111932432B (en) | Blind watermark implanting method, blind watermark detecting method, blind watermark implanting device, blind watermark detecting equipment and storage medium | |
| Niu et al. | Video watermarking resistance to rotation, scaling, and translation | |
| CN102024244B (en) | Method and device for embedding and detecting watermarks based on image characteristic region | |
| Zamani et al. | Knots of substitution techniques of audio steganography | |
| CN109064376A (en) | The anti-screen of DFT area image based on Harris-Laplace and SIFT description takes the photograph watermarking algorithm | |
| Kejariwal | Watermarking | |
| Piva et al. | Improving DFT watermarking robustness through optimum detection and synchronization | |
| CN104902341A (en) | Method for resisting geometric attack of safe digital video watermark | |
| Dong et al. | Frequency domain digital watermark algorithm implemented in spatial domain based on correlation coefficient and quadratic DCT transform | |
| CN114630130A (en) | Face-changing video tracing method and system based on deep learning | |
| Lin et al. | Fragile watermarking-based authentication scheme for demosaicked images | |
| Voigt et al. | High-capacity reversible watermarking for 2D vector data | |
| KR100945726B1 (en) | Apparatus for watermarking based on local auto-correlation function and method therefor | |
| Xuesong et al. | A dual digital audio watermarking algorithm based on LWT | |
| Behravan et al. | Introducing a new method of image reconstruction against crop attack using sudoku watermarking algorithm | |
| CN102497602A (en) | Digital watermarking processing method and device | |
| Joshi et al. | Efficient dual domain watermarking scheme for secure images | |
| Kadu et al. | Secure data hiding using robust firefly algorithm | |
| Viswanatham et al. | A hybrid digital watermarking algorithm for color images based on DWT and DCT | |
| Wang et al. | A large-size image watermarking algorithm based on pseudo-random |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150909 |