CN100559856C - A kind of video-frequency identifying method based on wavelet transformation and mixed watermark - Google Patents
A kind of video-frequency identifying method based on wavelet transformation and mixed watermark Download PDFInfo
- Publication number
- CN100559856C CN100559856C CNB2006100363848A CN200610036384A CN100559856C CN 100559856 C CN100559856 C CN 100559856C CN B2006100363848 A CNB2006100363848 A CN B2006100363848A CN 200610036384 A CN200610036384 A CN 200610036384A CN 100559856 C CN100559856 C CN 100559856C
- Authority
- CN
- China
- Prior art keywords
- frame
- video
- subband
- watermark
- authentication
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000009466 transformation Effects 0.000 title claims abstract description 15
- 239000000284 extract Substances 0.000 claims abstract description 23
- 230000008569 process Effects 0.000 claims description 18
- 238000000844 transformation Methods 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 2
- 238000013139 quantization Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 12
- 238000005516 engineering process Methods 0.000 description 7
- 238000007906 compression Methods 0.000 description 6
- 230000006835 compression Effects 0.000 description 6
- 238000000605 extraction Methods 0.000 description 3
- 238000010008 shearing Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000008485 antagonism Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000000205 computational method Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000021615 conjugation Effects 0.000 description 1
- 238000013144 data compression Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Landscapes
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
A kind of digital video authentication method based on wavelet transformation and mixed watermark, the present invention embeds the not watermark of same-action in a different manner at the different sub-band of wavelet transformation, combine effectively authenticating in interframe authentication and the frame, at first will identify the positional information of frame of video in video sequence embeds in the wavelet transformation subband of original video frame in the mode of robust watermark, extract the positional information that watermark obtains frame during authentication, by judging whether to exist unusual positional information, thereby whether the authentication video is under attack on time shaft, the feature with half fragile characteristic of extracting each two field picture then is as watermark, whether be embedded in two subbands of wavelet transformation with sane and half fragile mode, it is credible and in the malice tampering location of two field picture when being insincere to be respectively applied for the judgment frame image during authentication.The inventive method can judge correctly that the processing to video is that malice is distorted or normal Video processing, and can accurately locate the position of distorting in interframe and the frame.
Description
Technical field
The present invention is a kind of based on wavelet transformation and mixed watermark technology, can judge correctly that the processing to video is that malice is distorted or normal Video processing, and can belong to field of multimedia signal processing to the pinpoint video-frequency identifying method in the position of distorting.
Background technology
In recent ten years, digital multimedia is penetrated in the daily life just gradually.In various multimedia messagess, digital video is in occupation of very important position.But along with the development of multimedia treatment technology, the edit-modify of video is become very easy, and with the naked eye be difficult to distinguish the true and false.When relating to information such as politics, military affairs, court evidence when video content, its confidence level (whether being modified as monitoring video, television program etc.) must obtain confirming.Therefore, the video authentication technology becomes important research project of information security field.
Digital signature and digital watermarking are the major technique means of multimedia authentication.Traditional information integrity resist technology based on digital signature comes with some shortcomings when being applied to multimedia messages.For example, too responsive to normal signal processing (as data compression, signal filtering etc.) and interchannel noise (as wireless channel etc.), can't the positioning tampering position etc.Distorting very responsive for malice and the accurate positioning tampering position of energy based on the multimedia authentication specification requirement of digital watermarking, simultaneously, must be robust to normal signal processing or noise jamming, has stronger practicality.But the multimedia authentication Study on Technology based on digital watermarking mostly concentrates on the image authentication at present, and is less relatively to the research of video authentication.Video authentication not only needs to comprise in the frame similar with image authentication and authenticates, and also needs to comprise the interframe authentication.The method of existing video authentication is not overripened, particularly lacks the method that can effectively authentication in interframe authentication and the frame be combined.
The purpose of video authentication is to guarantee the integrality of digital video, and the video of acknowledging receipt is true and reliable, does not distort through malice.It comprises the authentication of two aspects: the interframe authentication can pick out video and whether be attack on time shaft, and indicate attack type (as deleting frame, frame losing, rearrangement frame etc.); Authentication is exactly the authentication to the video frame images content in the frame, is similar to image authentication.If video only be subjected to not changing its content operation (as various video compression, add make an uproar, format conversion etc.), just think that this video is genuine and believable; If the operation of implementing causes the change (as shearing, replacement etc.) of video content, think that then this video has been subjected to malice and has distorted, content is insincere, and further orients the zone of being distorted.
Summary of the invention
The objective of the invention is to propose a kind of video-frequency identifying method, both can determine whether to be subjected between frame of video malice and distort, also can determine whether to be subjected in the frame of video malice and distort, can accurately locate the position that malice is distorted simultaneously based on mixed watermark.
Method of the present invention is divided into watermark embedding and two processes of video authentication, utilizes the interframe authenticating water-mark to determine whether video is subjected to the malicious attack on the time shaft during video authentication, as frame deletion, frame insertion and frame exchange etc.; Utilization judges correctly that based on the watermark of the feature of two field picture the processing to two field picture is that malice is distorted or normal Video processing, simultaneously the position of distorting is accurately located, concrete steps are: telescopiny is as follows: 1) luminance component with every frame in the video sequence carries out 3 grades of two-dimensional wavelet transformations, obtains 10 subband { LH
i, HL
i, HH
i, LL
3, i=1-3}, shown in Figure of description (1), the low pass subband after wherein the third level is decomposed is designated as the LL3 subband, and low high pass subband is designated as the LH3 subband, and high low pass subband, is designated as the HL3 subband; 2) preliminary treatment of interframe authenticating water-mark and embedding; 3) generation and the embedding of authenticating water-mark in the frame; 4) wavelet inverse transformation obtains watermarked frame of video.The video authentication process does not need original video, and step is as follows: 1) luminance component with every frame in the video sequence carries out 3 grades of two-dimensional wavelet transformations, obtains 10 subband { LH
i, HL
i, HH
i, LL
3, i=1-3} is shown in Figure of description (1); 2) interframe authentication: extract the interframe authenticating water-mark, judge whether video is subjected to interframe and distorts; 3) authentication in the frame: extract authenticating water-mark in the frame and calculate the feature of two field picture to be certified, judge whether video is subjected to distorting in the frame, if be subjected to distorting in the frame, further positioning tampering then.
Being specifically described as of the inventive method:
1. the preliminary treatment of interframe authenticating water-mark and embedding
The present invention with frame number as the interframe authenticating water-mark.Because frame number is the sign of frame of video on time shaft, represents the watermark of frame number no matter to be subjected to normal process or malicious attack, all should be able to correct recovery come out in the test side, so the interframe authenticating water-mark should have very strong robustness.Processing method of the present invention is as follows: 1) with after the frame segmentation, represent frame number (0-255) with 8 bit-binary streams, generate watermark Wt; 2) with BCH (61,8) to its coding, obtain the watermark W ' t behind the 61 bit Bose-Chaudhuri-Hocquenghem Codes; 3) BCH code with 61 bits repeats 3 times, obtains 61 * 3=183 bit watermark W " t; 4) 183 bit informations are interweaved obtain time shaft watermark W " ' t, interleaving process is by cipher controlled; 5) with watermark W " ' bit among the t embeds the LH3 subband equably.
The embedding grammar of watermark bit is as follows:
S=64 wherein, X is a watermark bit, A (i), A ' (i) are respectively LH3 subband wavelet coefficient before and after embedding.
2. the generation and the embedding grammar of authenticating water-mark in the frame
Authentication adopts the feature of two field picture as watermark in the frame of the present invention.Feature Extraction and watermark embed process are as follows: 1) coefficient in the LL3 subband is mapped on the 0-255 level gray scale, obtains low frequency sub-band figure; 2) extract the edge of low frequency sub-band figure and carry out binaryzation; 3) amplitude of 49 normalized Zernike squares of calculating on binary image is designated as ZMM; 4) be 16 with each ZMM uniform quantization, and the highest 4 significance bits of getting each square value are as watermark Ws; 5) carry out error correction coding and generate Ws '; 6) Ws ' is embedded in the LL3 subband; 7) 4 watermarks of extracting from each ZMM among the Ws are embedded in the 2x2 piece in the HL3 subband, shown in Figure of description 2.
The Zernike square is proposed in 1934 by Zernike the earliest, and its computational methods are as follows:
For digital picture f (x, y), n rank number of repetition is that the Zernike square of m is defined as:
Wherein when the Zernike of computed image square, the origin of coordinates is located at the center of image, and pixel value is mapped in the unit circle, i.e. x
2+ y
2≤ 1.Drop on the outer pixel of unit circle and ignore, and have
(x, the y) vector distance of pixel, θ are vector ρ and the anticlockwise angle of x axle for initial point arrives.V
Nm *(ρ θ) is V
Nm(ρ, conjugation θ).V
Nm(ρ θ) is defined as follows:
V
nm(ρ，θ)＝R
nm(ρ)exp(jmθ) (3)
Wherein n is a nonnegative number, m be integer and satisfy simultaneously n-|m| be even number and | m|≤n, R
Nm(ρ) be the radius multinomial, the mathematical notation formula is:
And R is arranged
N ,-m(ρ)=R
Nm(ρ).
The amplitude ZMM of Zernike square is | A
Nm|, method for normalizing is | A
Nm| divided by the number of white pixel point in the binary map.ZMM has the robustness to JPEG, noise and rotation, but it is again the description feature of picture shape simultaneously, when picture material is replaced, will cause the bigger variation of ZMM, therefore has half fragile characteristic.The present invention is quantified as 16 with each ZMM, get the highest 4 as watermark Ws.
Ws is carried out BCH (63,7) coding, obtain W ' s, the present invention is called and distinguishes watermark.To distinguish that watermark evenly embeds the LL3 subband.Formula (1) is used in the embedding of watermark bit, S=48 wherein, and X is a watermark bit, A (i), A ' (i) are respectively LL3 subband wavelet coefficient before and after embedding.
Simultaneously 4 watermarks of extracting from each ZMM among the Ws are embedded in the 2x2 piece in the HL3 subband, this watermark is called the location watermark in the present invention.A wavelet coefficient values of HL3 subband is relevant with the piece of 8x8 in the real image behind 3 grades of wavelet transformations, and the piece of HL3 subband 2x2 is relevant with 16 * 16 of real images.The present invention is embedded into high 4 forms with the piece of 2x2 of each ZMM in the HL3 subband successively, and 16 * 16 piece is distorted the value that may change a square in real image so.Can determine the position that malice is distorted according to the position of the square that has changed in when authentication.Formula (1) is used in the embedding of watermark bit, S=32 wherein, and X is a watermark bit, A (i), A ' (i) are respectively HL3 subband wavelet coefficient before and after embedding.
3. interframe authentication
The interframe verification process is as follows: 1) extract authenticating water-mark between 183 bit frame from the LH3 subband; 2) with identical key when interweaving reciprocal cross being carried out in the watermark that extracts knits; 3) watermark information after with the code length (61 bit) of BCH code reciprocal cross being knitted carries out segmentation, determines the 61 bit-binary sequences that extract according to the law of large numbers, carries out the frame number that BCH decoding obtains the expression of corresponding 8 bit-binary stream.Determine that according to the relation between each frame frame number the time shaft that is subjected to attacks classification and position.
The extracting method of watermark bit is:
Wherein A ' is the LH3 sub-band coefficients (i), the watermark bit of X ' for extracting, S=64.
4. authentication in the frame
The process of authentication is as follows in the frame: 1) extract the LL3 sub-band coefficients; 2) the LL3 sub-band coefficients is mapped on the 0-255 level gray scale, obtains low frequency sub-band figure; 3) extract the edge of low frequency sub-band figure and carry out binaryzation; 4) amplitude of 49 normalized Zernike squares of calculating on binary image, note is ZMM_A; 5) extract the watermark information of distinguishing in the LL3 subband, decipher and be converted into the amplitude of 49 Zernike squares, note is ZMM_O; 6) calculate the quadratic sum of the difference of 49 corresponding ZMM_A and ZMM_O, and judge whether to have surpassed preset threshold T1,, think that then this two field picture suffered malicious attack, carried out next step if surpass threshold value; 7) extract location watermark information in the HL3 subband, be converted into the amplitude of 49 Zernike squares, note is ZMM_L; 8) calculate each corresponding ZMM_A and ZMM_L difference square, and judge whether to have surpassed preset threshold T2, if surpass threshold value, think that then position that this square embeds has been subjected to malice and has distorted.According to a large amount of experiments, set T
1=0.1, T
2=0.02.
Distinguish the extraction use formula (5) of watermark bit, wherein A ' is the LL3 sub-band coefficients (i), the watermark bit of X ' for extracting, S=48.
Formula (5) is used in the extraction of location watermark bit, and wherein A ' is the HL3 sub-band coefficients (i), the watermark bit of X ' for extracting, S=32.
The present invention relatively has following advantage with existing video authentication technology based on watermark:
1) adopt mixed watermark to realize authentication in interframe authentication and the frame.
2) interframe authentication adopts frame number as watermark, and carries out preliminary treatment and make it have stronger robustness.
3) feature that authenticates two field picture itself in the frame embeds as watermark, and utilizes the distance between feature to authenticate, and makes authentication result more reliable.Feature watermark embeds in two ways, a kind ofly is used to distinguish the attack classification that is subjected in the frame, a kind ofly is used for tampered position in the locating frame.
Description of drawings
Fig. 1 carries out 3 grades of 2-d wavelets to the luminance component of video frame images in the video-frequency identifying method that proposes of the present invention to decompose the subband schematic diagram that obtains;
Fig. 2 is the embedding grammar schematic diagram of location watermark in the video-frequency identifying method that proposes of the present invention;
Fig. 3 is the authentication method that proposes of the present invention frame of video figure and authentication result after watermarked in Salesman;
Fig. 4 is the authentication method that proposes of the present invention frame of video figure and authentication result after watermarked in Mobile;
Fig. 5 is the authentication method that proposes of the present invention frame of video figure and authentication result after watermarked in Paris;
Fig. 6 is the authentication method that proposes of the present invention frame of video figure and authentication result after watermarked in Stockholm Pan.
Embodiment
The present invention adopts the mixed watermark method, the positional information of representing video frame images on time shaft and the characteristic information of representative frame image content are embedded in the video frame images as watermark, when what video was carried out is that normal signal is when handling, think during authentication that video is believable, and when video is subjected to malice and distorts, can judge that not only video is incredible, but also position that can positioning tampering.
Introduce the practical example of video-frequency identifying method of the present invention below.
Statement is now finished the process that watermark embeds and detects with method of the present invention.Here provided and used salesman, pairs, the experimental result of Stockholm Pan and mobile cycle tests (size is 352x288).Have four aspects: invisibility; The accurate location that malice is distorted; The robustness of interframe authenticating water-mark and the fragility that interframe is distorted; Distinguish the robustness of watermark to normal Video processing such as MPEG-2 compression, noise.
Fig. 3 is the authentication method that proposes of the present invention figure and authentication result after watermarked in salesman, wherein, (a) be Salesman primitive frame figure, (b) be frame figure PSNR=41.97dB after watermarked, the frame that has embedded watermark does not as seen from the figure visually have special evident difference with primitive frame, satisfy the requirement of invisibility, (c) for erasing among (b) figure the forgery figure that obtains behind the box on hand, naked eyes be can't see, but when authentication, be considered to two field picture has been carried out artificial distorting, be incredible, can provide simultaneously the position of distorting shown in figure (d), wherein black part is divided the expression tampered position.
Fig. 4 is the authentication method that proposes of the present invention figure and authentication result after watermarked in mobile, wherein, (a) be Mobile primitive frame figure, (b) be frame figure PSNR=42.35dB after watermarked, the frame that has embedded watermark does not as seen from the figure visually have special evident difference with primitive frame, satisfy the requirement of invisibility, (c) for erasing the forgery figure that obtains after the January on the wall calendar among (b) figure, naked eyes be can't see, but when authentication, be considered to two field picture has been carried out artificial distorting, be incredible, can provide simultaneously the position of distorting shown in figure (d), wherein black part is divided the expression tampered position.
Fig. 5 is the authentication method that proposes of the present invention figure and authentication result after watermarked in paris, wherein, (a) be Paris primitive frame figure, (b) be frame figure PSNR=42.16dB after watermarked, the frame that has embedded watermark does not as seen from the figure visually have special evident difference with primitive frame, satisfy the requirement of invisibility, (c) for adding the forgery figure that obtains behind the hemisphere on the bookshelf in (b) figure, naked eyes be can't see, but when authentication, be considered to two field picture has been carried out artificial distorting, be incredible, can provide simultaneously the position of distorting shown in figure (d), wherein black part is divided the expression tampered position.
Fig. 6 is the authentication method that proposes of the present invention figure and authentication result after watermarked in Stockholm Pan, wherein, (a) be Stockholm Pan primitive frame figure, (b) be frame figure PSNR=41.79dB after watermarked, the frame that has embedded watermark does not as seen from the figure visually have special evident difference with primitive frame, satisfy the requirement of invisibility, (c) the forgery figure for obtaining behind the building in the middle of erasing among (b) figure, naked eyes be can't see, but when authentication, be considered to two field picture has been carried out artificial distorting, be incredible, can provide simultaneously the position of distorting shown in figure (d), wherein black part is divided the expression tampered position.
In order to detect the robustness of interframe authenticating water-mark, we have done the malice shearing, add and make an uproar sequence, and (what intensity represented to add is that scope is at [5 * S
n, 5 * S
n] random noise) and MPEG-2 compression (Mb/s presentation code code check) three kinds of operations.Table 1 has provided interframe watermark antagonism MPEG-2 compression and has added the robustness of making an uproar.The result shows, the size of shearing when malice is no more than 120x120, and the MPEG-2 encoder bit rate is not less than the value in the table 1, and when noise intensity was no more than value in the table 1, the interframe authenticating water-mark can correctly be detected.
| Video sequence | MPEG compresses (Mb/s) | The PSNR value (dB) of compressed video | Add (the S that makes an uproar n) | The PSNR value (dB) that adds the video of making an uproar |
| salesman | 1.25 | 36.51 | 10 | 27.11 |
| mobile | 3 | 31.23 | 10 | 27.11 |
| pairs | 2.5 | 37.29 | 10 | 27.09 |
| stockholmPan | 1.75 | 36.15 | 10 | 27.10 |
Table 1 interframe authenticating water-mark is to the MPEG compression and add the robustness of making an uproar
In order to detect the fragility that the interframe authenticating water-mark is attacked time shaft, table 2 has been listed the situation that frame number detects under several different interframe are attacked.To first kind of situation, owing to lost the 4-8 frame, the 3rd frame is afterwards followed by the 9th frame; To second kind of situation,,, represent that this frame does not belong to video sequence so a strange frame number 99 has appearred in the 3rd frame back owing in the middle of the 3rd frame and the 4th frame, inserted the frame of video of a non-video sequence; To the third situation,,, follow the 22nd frame then so be exactly the 35-39 frame after the 20th frame because the 35-39 frame is inserted between the 20th and 21 frames.From above-named situation as can be seen, no matter be discrete or continuous frame losing, delete frame, add frame, reset frame, the frame number that detects all has unusual appearance, that is to say that the interframe authenticating water-mark attacks interframe fragility is arranged, and can be used for classification and the position that the judgement time axle attacks.
| Lose the 4-8 frame | In 3,4 frames, insert an arbitrary frame | The 35-39 frame is inserted before 21 frames | |
| Detected |
1，2，3，9， 10… | 1，2，3，99，4，5，6，7… | 1，2，3，…，20，35， 36，…，39，21，22… |
Frame number detection case under several different interframe of table 2 are attacked
Whether when distinguishing that watermark has good robustness, utilizing half fragility of feature can finely must distinguish in the frame that institute is subject to processing is that malice is distorted.Table 3 has provided distinguishes the robustness of watermark to normal process.When the bit rate of MPEG compression is higher than the value in the table or adds the intensity of making an uproar when being lower than value in the table, can correctly extract and distinguish watermark, and what think that video is subjected to is normal Video processing.
| Video sequence | Salesman | Mobile | Pairs | Stockholm Pan |
| MPEG(Mb/s) | 1.5 | 5 | 2.5 | 2.5 |
| PSNR(dB) | 36.53 | 34.09 | 35.61 | 36.26 |
| Noise(S) | 7 | 7 | 7 | 7 |
| PSNR(dB) | 30.95 | 30.95 | 30.94 | 30.94 |
Distinguish the robustness of watermark in the authenticating water-mark in table 3 frame to normal process
Claims (6)
1, a kind of video-frequency identifying method based on wavelet transformation and mixed watermark is characterized in that this method is divided into watermark embedding and two processes of video authentication:
Telescopiny is as follows: 1) luminance component with every frame in the video sequence carries out 3 grades of two-dimensional wavelet transformations, obtains the low pass subband after the third level is decomposed, and is designated as the LL3 subband, and low high pass subband is designated as the LH3 subband, and high low pass subband, is designated as the HL3 subband; 2) preliminary treatment of interframe authenticating water-mark and embedding; 3) generation and the embedding of authenticating water-mark in the frame; 4) wavelet inverse transformation obtains watermarked frame of video;
The preliminary treatment and the embedding grammar of above-mentioned interframe authenticating water-mark are: 1) with the frame number of each frame as interframe authenticating water-mark Wt; 2) it is carried out Bose-Chaudhuri-Hocquenghem Code, obtain W ' t; 3) W ' t is repeated 3 times, obtains W " t; 4) W " the t back that interweaves evenly embeds in the LH3 subband;
The generation and the embedding grammar of authenticating water-mark are as follows in the above-mentioned frame: 1) coefficient in the LL3 subband is mapped on the 0-255 level gray scale, obtains low frequency sub-band figure; 2) extract the edge of low frequency sub-band figure and carry out binaryzation; 3) amplitude of 49 Zernike squares of calculating on binary image is designated as ZMM; 4) be 16 with each ZMM uniform quantization, and get the highest 4 significance bits, carry out error correction coding and generate Ws ' as watermark Ws; 5) Ws ' is embedded in the LL3 subband; 6) 4 watermarks of extracting from each ZMM among the Ws are embedded in the 2x2 piece in the HL3 subband;
The video authentication process does not need original video, and step is as follows: 1) luminance component with every frame in the video sequence carries out 3 grades of two-dimensional wavelet transformations, obtains the LL3 subband with telescopiny, LH3 subband and HL3 subband; 2) interframe authentication: extract the interframe authenticating water-mark, judge whether video is subjected to interframe and distorts; 3) the interior authentication of frame: extract authenticating water-mark in the frame, judge whether video is subjected to distorting in the frame, if be subjected to distorting in the frame, then further positioning tampering.
2, video-frequency identifying method according to claim 1 is characterized in that the process of interframe authentication is as follows: 1) extract the interframe authenticating water-mark from the LH3 subband; 2) reciprocal cross being carried out in the watermark that extracts knits; 3) watermark information after with the code length of BCH code reciprocal cross being knitted carries out segmentation, determine the bit-binary sequence that extracts according to the law of large numbers, carry out BCH decoding and obtain the frame number that corresponding 8 bit-binary stream is represented, determine the time shaft attack classification and the position that are subjected to according to the relation between each frame frame number.
3, video-frequency identifying method according to claim 1 is characterized in that the process of authentication in the frame is as follows: 1) extract the LL3 sub-band coefficients; 2) the LL3 sub-band coefficients is mapped on the 0-255 level gray scale, obtains low frequency sub-band figure; 3) extract the edge of low frequency sub-band figure and carry out binaryzation; 4) amplitude of 49 Zernike squares of calculating on binary image, note is made ZMM_A; 5) extract watermark information in the LL3 subband, decipher and be converted into the amplitude of 49 Zernike squares, note is made ZMM_O; 6) calculate the quadratic sum of the difference of 49 corresponding ZMM_A and ZMM_O, and judge whether to have surpassed preset threshold T1,, think that then this two field picture suffered malicious attack, carried out next step if surpass threshold value; 7) extract the watermark information that embeds in each piece of 2 * 2 in the HL3 subband, be converted into the amplitude of 49 Zernike squares, note is made ZMM_L; 8) calculate each corresponding ZMM_A and ZMM_L difference square, and judge whether to have surpassed preset threshold T2, if surpass threshold value, think that then position that this square embeds has been subjected to malice and has distorted.
4, video-frequency identifying method according to claim 1 is characterized in that W " process that interweaves of t is by cipher controlled.
5,, it is characterized in that in the process of interframe authentication that process that reciprocal cross knits is carried out in the watermark that extracts adopts key identical when interweaving according to claim 2 or 4 described video-frequency identifying methods.
6, video-frequency identifying method according to claim 3 is characterized in that two threshold value T that the interior process that authenticates of frame sets
1=0.1, T
2=0.02.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100363848A CN100559856C (en) | 2006-07-07 | 2006-07-07 | A kind of video-frequency identifying method based on wavelet transformation and mixed watermark |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100363848A CN100559856C (en) | 2006-07-07 | 2006-07-07 | A kind of video-frequency identifying method based on wavelet transformation and mixed watermark |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1889672A CN1889672A (en) | 2007-01-03 |
| CN100559856C true CN100559856C (en) | 2009-11-11 |
Family
ID=37578930
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100363848A Expired - Fee Related CN100559856C (en) | 2006-07-07 | 2006-07-07 | A kind of video-frequency identifying method based on wavelet transformation and mixed watermark |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100559856C (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102208096B (en) * | 2011-05-26 | 2012-11-28 | 西安理工大学 | Image tamper detection and tamper localization method based on discrete wavelet transformation |
| CN102932649A (en) * | 2011-08-08 | 2013-02-13 | 华为软件技术有限公司 | Video decoding quality detection method and device of set top box |
| CN108092736B (en) * | 2017-12-26 | 2019-12-10 | 广州海格通信集团股份有限公司 | Low detection probability waveform method based on wavelet transformation |
-
2006
- 2006-07-07 CN CNB2006100363848A patent/CN100559856C/en not_active Expired - Fee Related
Non-Patent Citations (2)
| Title |
|---|
| 一种小波变换域的自适应视频水印算法. 刘红梅，黄继武，肖自美.电子学报，第29卷第12期. 2001 |
| 一种小波变换域的自适应视频水印算法. 刘红梅，黄继武，肖自美.电子学报，第29卷第12期. 2001 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1889672A (en) | 2007-01-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101042769B (en) | Active mode digital image content identification method based on wavelet and DCT dual domain | |
| Bartolini et al. | Image authentication techniques for surveillance applications | |
| Voloshynovskiy et al. | Multibit digital watermarking robust against local nonlinear geometrical distortions | |
| Chang et al. | A secure fragile watermarking scheme based on chaos-and-hamming code | |
| CN101419702B (en) | Semi-fragile watermarking method for color image tamper localization and recovery | |
| CN102968756A (en) | Color image watermark embedding and extraction method based on fractional order hyper-chaos and discrete wavelet transform-singular value decomposition (DWT-SVD) | |
| CN101582158B (en) | Method for embedding and authenticating watermark of digital image | |
| CN101303725A (en) | Method for generating and authenticating frailty watermark based on error correction encoding | |
| CN102682418B (en) | Method for embedding and extracting multiple zero watermarks of digital image | |
| CN100399353C (en) | Electronic stamp certification method based on image features | |
| CN102096894A (en) | Image fragile watermarking algorithm capable of realizing accurate positioning of tampered region | |
| CN101901470A (en) | Image-tampering detection and recovery method based on energy-domain semi-fragile watermarking | |
| CN101835049A (en) | Generating and authenticating method of self-embedded digital watermark of JPEG (Joint Photographic Experts Group) image | |
| Mohan et al. | A new algorithm for data hiding in images using contourlet transform | |
| CN101567958B (en) | Semi-fragile digital watermaking system based on non-redundant Contourlet conversion | |
| CN102521821A (en) | Automatic identification and tampered area positioning method in digital image | |
| CN100559856C (en) | A kind of video-frequency identifying method based on wavelet transformation and mixed watermark | |
| CN103559676A (en) | Method for preventing color image blind watermarking from being printed and scanned based on DCT coefficient statistical property | |
| CN1315091C (en) | Digital image recognising method based on characteristics | |
| CN101272485B (en) | Digital watermarking stacking and detecting method for digital movie copyright protection | |
| Hu et al. | A histogram based watermarking algorithm robust to geometric distortions | |
| Zhang et al. | Blind dual watermarking scheme using stucki kernel and spiht for image self-recovery | |
| CN103179406A (en) | Digital video watermark method using kinoform | |
| Zhu et al. | Multimedia authentication and watermarking | |
| Mathivadhani et al. | Digital watermarking and information hiding using wavelets, SLSB and Visual cryptography method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20091111 Termination date: 20100707 |