CN102307320A - Piracy tracing watermarking method applicable to streaming media environment - Google Patents
Piracy tracing watermarking method applicable to streaming media environment Download PDFInfo
- Publication number
- CN102307320A CN102307320A CN201110229609A CN201110229609A CN102307320A CN 102307320 A CN102307320 A CN 102307320A CN 201110229609 A CN201110229609 A CN 201110229609A CN 201110229609 A CN201110229609 A CN 201110229609A CN 102307320 A CN102307320 A CN 102307320A
- Authority
- CN
- China
- Prior art keywords
- watermark
- video
- motion vector
- frame
- watermark information
- 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.)
- Granted
Links
Images
Abstract
The invention relates to a piracy tracing watermarking method applicable to streaming media environment, and belongs to the multimedia information safety field. The method comprises the following steps: firstly segmenting video before the watermark embedding procedure; according to content of streaming video, selecting macroblocks meeting the conditions to be embedded with watermarks during the watermark embedding process; and determining the macroblock which is best suited for being embedding with watermark information by adopting a quicksort algorithm. In the piracy tracing watermarking method, based on difference between horizontal direction and vertical direction of motion vector of the macroblock, appropriate motion components are adaptively selected to be embedded with 2-bit watermark information, and the NAL (network abstraction layer) head of a B-frame is modified after the watermark embedding procedure so that the watermark can cope with the situations such as frame loss and the like resulting from network transmission. By adopting the piracy tracing watermarking method, the watermark information can be embedded into the streaming video in real time so as to achieve the purpose of piracy tracing.
Description
Technical field
The present invention relates to a kind of piracy tracking water mark method that is applicable to the Streaming Media environment, belong to the multi-media information security field.
Background technology
The integration of three networks is meant that telecommunications network, computer network and cable television network three macroreticulars are through technological transformation; Comprise all kinds of comprehensive multimedia services such as voice, data, image through the IP network carrying; It shows as on the network layer and interconnects, and interpenetrates on the operation layer and intersects.Streaming Media is meant the media formats that the mode of employing stream transmission is play on the internet.The IPTV that IP based network is realized is professional, and its essential information adopts the real-time streaming mode to transmit exactly.Video provider through after the processed compressed, is placed on the transmission of carrying out segmentation on the webserver with video file, and client computer need not download to this locality with whole video file, just can view and admire immediately and watch.IPTV business under the integration of three networks can provide the rich video resource for the user.After yet these video resources were downloaded, often by the random copy propagation of malicious user, this made the interests of businessman receive great loss.User's identity information as watermark, is embedded in the streaming media video, can realize piracy tracking, can protect the interests of businessman to a certain extent.
H.264 video coding is a new digital video coding standard of being developed jointly by ITU-T and ISO/IEC.H.264 not only have high compression ratio, but also have high-quality image, Just because of this, through H.264 the compression video data, in network transmission process needed bandwidth still less, also more economically.Therefore, the H.264 video watermark under the research Streaming Media environment has positive meaning.
Traditional video watermark technology; Based on the video watermark technology of dct transform, all be embed digital watermark in transform domain for example, the complexity of watermarking algorithm is higher; Generally can not satisfy the requirement of the real-time of Streaming Media, can not directly apply in the Streaming Media.
Owing to only need the part of decoded video coding, it is less that watermark embeds required time ratio based on the H.264 video watermark of motion vector, and operand seldom can satisfy under the Streaming Media environment real-time requirement to digital watermarking in the watermark embed process.But existing H.264 video watermark technology based on motion vector is applied to carry out piracy tracking under the later Streaming Media environment of the integration of three networks, also has the following problem:
1. the piracy tracking watermark under the Streaming Media generally can not embed video with user profile in advance; And need be according to user's request; Real-time embeds streaming media video with user profile, i.e. the time that watermark embeds can not be long, otherwise can influence the appreciation effect of user to video.
2. in the Streaming Media environment, video is decoded again and encoded is very common operation, and H.264 video coding partly provides multiple algorithm for estimating in estimation, for example rhombus therapy DS and hexagon method HEXBS etc.Use that these algorithms draw about the macroblock partitions of frame of video and the value of motion vector, many times all there are differences, and with motion vector as the watermark carrier embed watermark after because change has taken place carrier itself, the watermark information that causes embedding is wherein lost.
3. Streaming Media provider generally can provide the video of various modes to supply the user to select, and after the integration of three networks, the video content that various network provides also is very different in order to satisfy different user's requests.For HD video, the motion vector of revising frame of video by conventional method tends to cause some detail section distortion; And for those low image quality videos, number of macroblocks that often can embed watermark is fewer.
4. the video under the Streaming Media environment is to take the network mode transmission; In the transmission course because the unsteadiness of network; Some frame in the video possibly be dropped, and if contain watermark information in these frame of video, just can not extract watermark information again after being dropped so.
5. in order to indicate user's specifying information, realize piracy tracking, the capacity of watermark must be enough big.Present motion vector watermark all can only embed the watermark information of 1bit, and in frame of video, the motion vector that is fit to embed watermark is conditional, and the watermark capacity of general water mark method does not reach requirement.
Summary of the invention
Technical problem:
The present invention proposes a kind of piracy tracking water mark method that is applicable to the Streaming Media environment; Solved traditional H.264 video digital watermark because watermark embeds overlong time; Can not be according to the selection watermark embedded location of video content adaptive; Be not suitable for reasons such as Network Transmission environment, and can not directly be used for the H.264 problem of streaming media video.
Technical scheme:
Method of the present invention is carried out segment processing to video earlier before watermark embeds, then according to the streaming media video content, choose suitable macro block embed watermark, and the motion vector of each macro block embeds the watermark information of 2bit.
Parameter-definition:
If certain macro block of 1 B frame contains a plurality of motion vectors, then call the maximum motion vector of motion amplitude the motion vector of this macro block.
2, the motion vector of macro block uses vector representation to be: (V
x, V
y), V wherein
xBe horizontal component, V
yIt is vertical component.
3, the motion vector magnitude formula:
=
.
4, HD video is meant the video that 1280 * 720 resolution are above, and low image quality video is meant the video of resolution below 480 * 360, and all the other are ordinary video.
The concrete grammar flow process:
1 user profile preliminary treatment:
The first step: at key Y
1Control under, the User Identity ID of m position is carried out scramble handles and obtain ID ';
Second step: ID ' spread spectrum is become the binary bits information W of 2m position, and W is as the final watermark information that embeds video.
The video resource preliminary treatment:
The first step: with t minute be the time interval, needs are carried out the video that watermark embeds carry out segmentation;
Second step: first section with video is transferred to the user;
The 3rd step: remaining all segmentation videos are carried out the watermark embedding operation simultaneously.
Watermark embeds concrete steps:
The first step: the video to segmentation carries out the NAL decoding, i.e. network layer decoding, and the VCL part of acquisition video promptly obtains the video coding layer segment.
Second step: the VCL to video partly carries out the entropy decoding, obtains the motion vector of all B frames.
The 3rd step: utilize key Y
2Select among each GOP, promptly in the picture group, need the B frame of embed watermark, detail is following: choose the binary number of a 3n position arbitrarily, utilize key Y
2It is encrypted the binary string L that obtains the 3n position, be divided into the n group to the binary string L that generates then, generate position sequence (L
1, L
2..., L
i..., L
n), wherein n is the number of GOP unit in the video-frequency band, L
1, L
2..., L
i..., L
nAll be 3 bits, be used for representing the relative position of GOP B frame, promptly with respect to the position of first B frame among this GOP; According to L
iValue, select the L of i GOP unit
iIndividual B frame embed watermark.
The 4th step: for HD video; Select the motion vector magnitude of the macro block of all 16 * 16,16 * 8 and 8 * 16 sizes in this B frame to judge; Then select the motion amplitude of the macro block of 16 * 16,16 * 8,8 * 16 and 8 * 8 sizes to judge for ordinary video and low image quality video, if the motion vector magnitude of current macro is greater than embedding threshold value T
1, then the amplitude with the numbering of this macro block and motion vector is designated as that (wherein Mc is the numbering of macro block for Mc, form f), and deposit array in, and f is the amplitude of motion vector.
The 5th step: based on the size of motion vector magnitude, utilize quick sorting algorithm that all macro blocks are carried out ordering from big to small, and m macro block embed watermark before selecting.
The 6th step: by the order of arranging embed watermark information successively, telescopiny is shown in figure (1) with the macro block of selecting.Detail is following:
Suppose that current macro arranges the i position, at first relatively the horizontal component V of its motion vector
xWith vertical component V
yThe size of value.If || V
x|-| V
y|| be less than or equal to difference threshold T
2, then, promptly by formula 1 revise respectively at the horizontal direction of motion vector and the watermark information of vertical direction embedding 1bit | v|, with the 2i-1 position embedding V of W
xEmbed V with the 2i position
yWherein v representes V
xOr V
y, the watermark information value of λ for embedding, mod is a modulo operation.
Annotate: if v
is a negative, then final v=-| v|; Otherwise v=| v|.
If || V
x|-| V
y|| greater than difference threshold T
2, then revising according to formula 2 | v| is embedded into the bigger component of absolute value with 2i-1 position and the 2i position of W; Wherein v representes the component that absolute value is bigger, promptly | and v|=max{|V
x|, | V
y|, the watermark information value of λ for embedding, mod is a modulo operation.
Annotate: if v
is a negative, then final v=-| v|; Otherwise v=| v|.
The 7th step: after watermark information embeds the B frame, revise B frame place NAL head, the value that makes Nal_ref_idc is 3.The Nal_ref_idc value is 0 to 3, and it is important more to be worth big more representative, if LOF takes place in transmission course, then the Nal_ref_idc value is that 3 frame must be retransmitted.
4 watermark extracting concrete steps:
The flow process of watermark extracting is shown in figure (2), and concrete steps are following:
The first step: according to key Y
2Position sequence (the L that generates
1, L
2..., L
i..., L
n) select in the video and to embed the video B frame that watermark is arranged.
Second step:, select in its B frame all motion vectors greater than embedding threshold value T for HD video
116 * 16,16 * 8 and 8 * 16 the size macro block; For ordinary video and low image quality video, then select in its B frame all motion vectors greater than embedding threshold value T
116 * 16,16 * 8,8 * 16 and 8 * 8 the size macro block.
The 3rd step: to the macro block selected size by motion vector magnitude, ordering from big to small, m macro block extraction watermark before getting: if || V
x|-| V
y|| smaller or equal to difference threshold T
2, the watermark information that then extracts 1bit from the horizontal direction and the vertical direction of this macroblock motion vector respectively, wherein (| V
x| * 4) value of mod2 is and embeds V
xThe watermark information value of direction, (| V
y| * 4) value of mod2 is and embeds V
yThe watermark information value of direction; If || V
x|-| V
y|| greater than difference threshold T
2, then on the bigger component of absolute value, extract the watermark information of 2bit, (| v| * 4) mod4
Value be the watermark information value,
| v| represents the bigger component of absolute value, promptly | and v|=max{|V
x|, | V
y|.
The 4th step: the watermark information W that takes out is carried out anti-spread spectrum, obtain ID '.
The 5th step: utilize key Y
1The random ID ' that is inverted obtains final User Identity ID.
Beneficial effect
1, video watermark process real-time of the present invention is better, and the complexity of algorithm is lower, and the code stream increase of watermark embedding video rear video is littler, is in particular in:
(1) video resource that needs embed watermark is carried out segment processing, parallel embed watermark, watermark embeds more efficient, and required time is still less.
(2) combine actual conditions, first section of video embed watermark not when watermark embeds, is transferred to the user with first section of video, makes the imperceptible watermark of user embed the time-delay that is brought.
(3) embed watermark in all video B frames not has only a B frame to contain watermark information among each GOP, need the frame of video of embed watermark to be controlled at reasonable range, and the increase meeting of code stream is fewer, and the time overhead of embed watermark reduces.
2, video watermark process invisibility of the present invention is better, and is littler to the influence of video quality, is in particular in:
(1) embed watermark in not doing the B frame of reference frame only.The B frame can not cause error diffusion owing to do not do reference frame.The error diffusion in the frame has all only been considered in traditional motion vector watermark, and does not have the error diffusion between considered frame.
(2) consider the horizontal direction of motion vector and the difference of vertical direction, taked new watermark to embed strategy, littler for the influence of low image quality video quality, do not allow to produce blocking effect.
(3) according to the content of video, select corresponding macro block embed watermark, and the macro block that meets the demands is sorted by motion amplitude size, obtain optimum watermark and embed macro block.
3, video watermark process watermark capacity of the present invention is bigger, and each motion vector embeds the 2bit watermark information, and general motion vector algorithms can only embed the information of 1bit.
4, video watermark process of the present invention adapts to the Streaming Media environment more, can tackle the generation of situation such as the frame of video of bringing owing to Network Transmission abandons.
Description of drawings
Fig. 1 embeds flow chart for watermark
Fig. 2 is the watermark extracting flow chart
Embodiment
Describe for ease, we have following application example at supposition:
User Bob plays a film of 100 minutes to the Alice of businessman request, after the Alice of businessman receives the playing request of user Bob, at first will examine the identity information of user Bob.If user Bob is illegal user, promptly do not pay and watch the user of this film, the Alice of businessman is with the playing request of refusing user's Bob; If user Bob is a validated user; Be the user that charges paid is watched this film; The Alice of businessman will utilize watermark embedding method provided by the invention, embed this film to the identity information of user Bob as watermark, and wherein the identity information of Bob is the binary-coded serial numbers of 32bit.When the Alice of businessman finds this film by piracy, he will utilize watermark extracting method provided by the invention, from this film, extract the watermark that contains user profile, thereby find the bootlegger of this film, realize piracy tracking.Embed threshold value T in this example
1Get 6, difference threshold T
2Get 3.
Then its embodiment is:
1 at first carries out preliminary treatment to user profile, and detail is following:
The first step: at key Y
1Control under the identify label ID of user Bob is carried out scramble handle and obtain ID '.
Concrete disorder method is following: establish the ID ' that array A is used to deposit the M position, array B is an interim array of depositing intermediate object program, and the length of array A and array B all is M, and the M value is 32.Choose the minimum positive integer P of prime number each other, make following scramble with M: 1) B (i)=A ((i * P) mod M) i=0,1,2 ..., M-1
2)A=B
Repeat top two step Y
1Inferior, can obtain ID '.
Second step: at last ID ' spread spectrum is become 64 binary bits information W.Binary bits information W is as the final watermark that embeds video.
Concrete spectrum spreading method is following: at first utilize formula ID ' (j)=(1) ID '
ID'
(j)Carry out bipolarityization, j=1 wherein, 2 ..., 32, ID ' (j) represents the j position of ID '.S={s by formula then
j| s
j=m
i, i=jmod2,0≤j≤63} carry out figure place to ID ' and expand, and obtain 64 S sequence, generate pseudo random sequence p, by formula K={K by pseudorandom number generator again
j| K
j=p
j* s
j, 0≤j≤63} formation sequence K becomes 0 with owning-1 among the sequence K at last, promptly obtains final watermarking information W.
2 carry out preliminary treatment to video resource then, and detail is following:
The first step: before to the film embed watermark, earlier film is carried out segment processing.Be the time interval with 5 minutes at first, be divided into 20 sections video segments with identical size with film.
Second step: first video segment of film is transferred to the user, and first fragment is embed watermark not.
The 3rd step: 19 sections videos to remaining, carry out the watermark embedding operation simultaneously.
3 watermarks embed:
The first step: the video to segmentation carries out the NAL decoding, obtains the VCL part of video.
Second step: the VCL to video partly carries out the entropy decoding, obtains the motion vector of all B frames.
The 3rd step: utilize key Y
2Select among each GOP, promptly in the picture group, need the B frame of embed watermark, detail is following: choose the binary number of a 3n position arbitrarily, utilize key Y
2It is encrypted the binary string L that obtains the 3n position, be divided into the n group to the binary string L that generates then, generate position sequence (L
1, L
2..., L
i..., L
n), wherein n is the number of GOP unit in the video-frequency band, L
1, L
2..., L
i..., L
nAll be 3 bits, be used for representing the relative position of GOP B frame, promptly with respect to the position of first B frame among this GOP; According to L
iValue, select the L of i GOP unit
iIndividual B frame embed watermark.
The 4th step: if this video is a HD video; Then select the motion vector magnitude of the macro block of all 16 * 16,16 * 8 and 8 * 16 sizes in its B frame to judge; If not; Then select the motion amplitude of the macro block of 16 * 16,16 * 8,8 * 16 and 8 * 8 sizes to judge, if the motion vector magnitude of current macro is greater than embedding threshold value T
1, then the amplitude with the numbering of this macro block and motion vector is designated as that (wherein Mc is the numbering of macro block for Mc, form f), and deposit array in, and f is the amplitude of motion vector.
The 5th step: based on the size of motion vector magnitude, utilize quick sorting algorithm that all macro blocks are carried out ordering from big to small, and select preceding 32 macro block embed watermarks.
The 6th step: with the macro block of selecting by the order of arranging embed watermark information successively.Detail is following:
Suppose that current macro arranges the i position, at first relatively the horizontal component V of its motion vector
xWith vertical component V
yThe size of value.If || V
x|-| V
y|| be less than or equal to difference threshold T
2, then, promptly by formula 1 revise respectively at the horizontal direction of motion vector and the watermark information of vertical direction embedding 1bit | v|, with the 2i-1 position embedding V of W
xEmbed V with the 2i position
yWherein v representes V
xOr V
y, the watermark information value of λ for embedding, mod is a modulo operation.
Annotate: if v
is a negative, then final v=-| v|; Otherwise v=| v|.
If || V
x|-| V
y|| greater than difference threshold T
2, then revising according to formula 2 | v| is embedded into the bigger component of absolute value with 2i-1 position and the 2i position of W; Wherein v representes the component that absolute value is bigger, promptly | and v|=max{|V
x|, | V
y|, the watermark information value of λ for embedding, mod is a modulo operation.
Annotate: if v
is a negative, then final v=-| v|; Otherwise v=| v|.
The 7th step: after watermark information embeds the B frame, revise B frame place NAL head, the value that makes Nal_ref_idc is 3, when promptly this frame is lost, forces its re-transmission.
4 watermark extracting:
The first step: according to key Y
2Position sequence (the L that generates
1, L
2..., L
i..., L
n) select in the video and to embed the video B frame that watermark is arranged.
Second step:, then select in its B frame all motion vectors greater than embedding threshold value T if this video is a HD video
116 * 16,16 * 8 and 8 * 16 the size macro block; If not, then select in its B frame all motion vectors greater than embedding threshold value T
116 * 16,16 * 8,8 * 16 and 8 * 8 the size macro block.
The 3rd step: to the macro block the selected size by motion vector magnitude, ordering is from big to small got preceding 32 macro blocks and is extracted watermark, and detail is following:
1) if || V
x|-| V
y|| smaller or equal to difference threshold T
2, then respectively from the horizontal direction V of this macroblock motion vector
xExtract the watermark information of 1bit with vertical direction Vy, wherein (| V
x| * 4) value of mod2 is and embeds V
xThe watermark information value of direction, (| V
y| * 4) value of mod2 is and embeds V
yThe watermark information value of direction.
2) if || V
x|-| V
y|| greater than difference threshold T
2, and V
xAbsolute value bigger, V in the horizontal direction then
xExtract the watermark information of 2bit, if V
yAbsolute value bigger, then at vertical direction V
yExtract the watermark information of 2bit, wherein (| V
x| * 4) value of mod4 is and embeds V
xThe watermark information value of direction, (| V
y| * 4) value of mod4 is and embeds V
yThe watermark information value of direction.
The 4th step: the watermark information W that takes out is carried out anti-spread spectrum, obtain ID ', detail is following: make all 0 bits among the watermark information W into-1, obtain sequence K.Then with sequence K S={S by formula
j| S
j=p
j* k
j, the anti-spread spectrum of 0≤j≤63} obtains the S sequence, gets preceding 32 or back 32 of S sequence, all with wherein-1 changes 1,1 into changes 0 into, can obtain the scramble serial ID '.
The 5th step: utilize key Y
1The random ID ' that is inverted obtains final User Identity ID, and detail is following: utilize formula B ((i * P) mod M)=A (i) i=0,1,2 ..., M-1; A=B carries out Y
1Inferior being inverted disorderly obtains final User Identity ID (wherein array A is used to store ID ', and array B is used to store final result, and M is 32).
Claims (7)
1. a piracy tracking water mark method that is applicable to the Streaming Media environment is characterized in that, comprises following four steps:
The first step: user profile preliminary treatment: under the control of key, User Identity ID is carried out scramble and spread spectrum successively, generate the final watermark information that embeds video;
Second step: video resource preliminary treatment: the video that needs is carried out the watermark embedding carries out segment processing;
The 3rd step: watermark embeds: with user profile as watermark, according to the content of video, the adaptive motion vector that is embedded into video B frame;
The 4th step: watermark extracting: select the B frame that embedding has watermark information according to key, from motion vector, extract watermark information then.
2. piracy tracking water mark method as claimed in claim 1 is characterized in that, described user profile preliminary treatment specifically may further comprise the steps:
The first step: at key Y
1Control under, the User Identity ID of m position is carried out scramble handles and obtain ID ';
Second step: ID ' spread spectrum is become the binary bits information W of 2m position, and W is as the final watermark information that embeds video.
3. piracy tracking water mark method as claimed in claim 1 is characterized in that, described video resource preliminary treatment specifically may further comprise the steps:
The first step: with t minute be the time interval, needs are carried out the video that watermark embeds carry out segmentation;
Second step: first section with video is transferred to the user;
The 3rd step: remaining all segmentation videos are carried out the watermark embedding operation simultaneously.
4. piracy tracking water mark method as claimed in claim 1 is characterized in that, described watermark embeds, and specifically may further comprise the steps:
The first step: the video to segmentation carries out the NAL decoding, i.e. network layer decoding, and the VCL part of acquisition video promptly obtains the video coding layer segment;
Second step: the VCL to video partly carries out the entropy decoding, obtains the motion vector of all B frames;
The 3rd step: utilize key Y
2Select among each GOP, promptly in the picture group, need the B frame of embed watermark;
The 4th step: for HD video, select the motion vector magnitude of the macro block of all 16 * 16,16 * 8 and 8 * 16 sizes in this B frame to judge, if the motion vector magnitude of current macro is greater than embedding threshold value T
1, then deposit the numbering of this macro block and the amplitude of motion vector in array; For ordinary video and low image quality video, then select the motion vector magnitude of the macro block of all 16 * 16,16 * 8,8 * 16 and 8 * 8 sizes in this B frame to judge, if the motion vector magnitude of current macro is greater than embedding threshold value T
1, then deposit the numbering of this macro block and the amplitude of motion vector in array;
The 5th step: according to the size of motion vector magnitude, utilize quick sorting algorithm that all macro blocks of electing are carried out ordering from big to small, and m macro block embed watermark before selecting;
The 6th step: with the macro block of selecting by the order of arranging embed watermark information successively: the horizontal component V of the motion vector of macro block relatively
xWith vertical component V
yThe size of value, if || V
x|-| V
y|| be less than or equal to difference threshold T
2, then respectively at the horizontal direction of motion vector and the watermark information of vertical direction embedding 1bit; If || V
x|-| V
y|| greater than difference threshold T
2, then only on the bigger component of absolute value, embed the watermark information of 2bit;
The 7th step: after watermark information embeds the B frame, revise B frame place NAL head, the value that makes Nal_ref_idc is 3, when promptly this frame is lost, forces its re-transmission.
5. piracy tracking water mark method as claimed in claim 4 is characterized in that, the described key Y that utilizes
2Selecting needs the B of embed watermark frame among each GOP, specifically be meant: choose the binary number of a 3n position arbitrarily, utilize key Y
2It is encrypted the binary string L that obtains the 3n position, be divided into the n group to the binary string L that generates then, generate position sequence (L
1, L
2..., L
i..., L
n), wherein n is the number of GOP unit in the video-frequency band, L
1, L
2..., L
i..., L
nAll be 3 bits, be used for representing the relative position of GOP B frame, promptly with respect to the position of first B frame among this GOP; According to L
iValue, select the L of i GOP unit
iIndividual B frame embed watermark.
6. piracy tracking water mark method as claimed in claim 4 is characterized in that, described with the macro block of selecting by the order of arranging embed watermark information successively, specifically be meant: establish current macro and arrange the i position, if || V
x|-| V
y|| be less than or equal to difference threshold T
2, then, promptly by formula 1 revise respectively at the horizontal direction of motion vector and the watermark information of vertical direction embedding 1bit | v|, with the 2i-1 position embedding V of W
xEmbed V with the 2i position
yWherein v representes V
xOr V
y, the watermark information value of λ for embedding, mod is a modulo operation;
If || V
x|-| V
y|| greater than difference threshold T
2, then revising according to formula 2 | v| is embedded into the bigger component of absolute value with 2i-1 position and the 2i position of W; Wherein v representes the component that absolute value is bigger, promptly | and v|=max{|V
x|, | V
y|, the watermark information value of λ for embedding, mod is a modulo operation
7. piracy tracking water mark method as claimed in claim 1 is characterized in that, described watermark extracting specifically may further comprise the steps:
The first step: according to key Y
2Position sequence (the L that generates
1, L
2..., L
i..., L
n) select in the video and to embed the video B frame that watermark is arranged;
Second step:, select in its B frame all motion vectors greater than embedding threshold value T for HD video
116 * 16,16 * 8 and 8 * 16 the size macro block; For ordinary video and low image quality video, then select in its B frame all motion vectors greater than embedding threshold value T
116 * 16,16 * 8,8 * 16 and 8 * 8 the size macro block;
The 3rd step: to the macro block selected size by motion vector magnitude, ordering from big to small, m macro block extraction watermark before getting: if || V
x|-| V
y|| smaller or equal to difference threshold T
2, the watermark information that then extracts 1bit from the horizontal direction and the vertical direction of this macroblock motion vector respectively, wherein (| V
x| * 4) value of mod2 is and embeds V
xThe watermark information value of direction, (| V
y| * 4) value of mod2 is and embeds V
yThe watermark information value of direction; If || V
x|-| V
y|| greater than difference threshold T
2, then on the bigger component of absolute value, extract the watermark information of 2bit, (| v| * 4) mod4
Value be the watermark information value,
| v| represents the bigger component of absolute value, promptly | and v|=max{|V
x|, | V
y|;
The 4th step: the watermark information W that takes out is carried out anti-spread spectrum, obtain ID ';
The 5th step: utilize key Y
1The random ID ' that is inverted obtains final User Identity ID.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110229609 CN102307320B (en) | 2011-08-11 | 2011-08-11 | Piracy tracing watermarking method applicable to streaming media environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110229609 CN102307320B (en) | 2011-08-11 | 2011-08-11 | Piracy tracing watermarking method applicable to streaming media environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102307320A true CN102307320A (en) | 2012-01-04 |
| CN102307320B CN102307320B (en) | 2013-07-10 |
Family
ID=45381116
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201110229609 Active CN102307320B (en) | 2011-08-11 | 2011-08-11 | Piracy tracing watermarking method applicable to streaming media environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102307320B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103379101A (en) * | 2012-04-20 | 2013-10-30 | 腾讯科技（深圳）有限公司 | Watermark generation method, client side and server |
| CN104598782A (en) * | 2014-12-04 | 2015-05-06 | 广东欧珀移动通信有限公司 | Data packaging and analysis method and device |
| CN106331744A (en) * | 2015-07-03 | 2017-01-11 | 上海尚恩华科网络科技股份有限公司 | Video interactive method based on audio watermarking |
| CN108521613A (en) * | 2018-05-15 | 2018-09-11 | 中南民族大学 | A kind of compressed domain video watermark insertion and extracting method based on moving target |
| CN108781302A (en) * | 2016-02-25 | 2018-11-09 | 思科技术公司 | Frame for the embedding data in encoded video |
| RU2688202C1 (en) * | 2018-07-03 | 2019-05-21 | Закрытое акционерное общество "Перспективный мониторинг" | Method for hidden marking of digital television signal data stream |
| CN109922066A (en) * | 2019-03-11 | 2019-06-21 | 江苏大学 | Dynamic watermark insertion and detection method in a kind of communication network based on time slot feature |
| CN109982157A (en) * | 2019-04-01 | 2019-07-05 | 北京火眼目测科技有限公司 | The method, system and storage medium of implicit fingerprint are embedded in digital video |
| CN110062291A (en) * | 2019-04-29 | 2019-07-26 | 王子孟 | A kind of digital watermarking addition and extracting method, apparatus and system |
| US10972807B2 (en) | 2018-04-06 | 2021-04-06 | Deluxe One Llc | Dynamic watermarking of digital media content at point of transmission |
| CN112991742A (en) * | 2021-04-21 | 2021-06-18 | 四川见山科技有限责任公司 | Visual simulation method and system for real-time traffic data |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1719898A (en) * | 2005-05-25 | 2006-01-11 | 中山大学 | Method of protecting MPEG-2 video data |
| CN101056392A (en) * | 2007-05-14 | 2007-10-17 | 华为技术有限公司 | A digital water mark embedding method and its device |
| CN101841700A (en) * | 2010-05-18 | 2010-09-22 | 宁波大学 | H.264 compression bit stream oriented video watermark method |
-
2011
- 2011-08-11 CN CN 201110229609 patent/CN102307320B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1719898A (en) * | 2005-05-25 | 2006-01-11 | 中山大学 | Method of protecting MPEG-2 video data |
| CN101056392A (en) * | 2007-05-14 | 2007-10-17 | 华为技术有限公司 | A digital water mark embedding method and its device |
| CN101841700A (en) * | 2010-05-18 | 2010-09-22 | 宁波大学 | H.264 compression bit stream oriented video watermark method |
Non-Patent Citations (5)
| Title |
|---|
| C.H.KUNG: "《Video Watermarking Using Motion Vector》", 《CVGIP2003》 * |
| JUN ZHANG: "《Viedo Watermark Technique in Motion Vector》", 《IEEE》 * |
| 张江等: "《视频水印综述》", 《计算机应用》 * |
| 李玲慧: "《基于运动矢量的视频水印算法》", 《CNKI硕士论文数据库》 * |
| 林克正等: "《基于运动矢量统计的压缩域视频水印算法》", 《哈尔滨理工大学学报》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9367885B2 (en) | 2012-04-20 | 2016-06-14 | Tencent Technology (Shenzhen) Company Limited | Method and system for adding and detecting watermark |
| CN103379101A (en) * | 2012-04-20 | 2013-10-30 | 腾讯科技（深圳）有限公司 | Watermark generation method, client side and server |
| CN104598782A (en) * | 2014-12-04 | 2015-05-06 | 广东欧珀移动通信有限公司 | Data packaging and analysis method and device |
| CN106331744A (en) * | 2015-07-03 | 2017-01-11 | 上海尚恩华科网络科技股份有限公司 | Video interactive method based on audio watermarking |
| CN108781302B (en) * | 2016-02-25 | 2021-09-03 | 西娜媒体有限公司 | Method and system for providing a framework for embedding data in encoded video |
| CN108781302A (en) * | 2016-02-25 | 2018-11-09 | 思科技术公司 | Frame for the embedding data in encoded video |
| US10972807B2 (en) | 2018-04-06 | 2021-04-06 | Deluxe One Llc | Dynamic watermarking of digital media content at point of transmission |
| CN108521613B (en) * | 2018-05-15 | 2019-03-12 | 中南民族大学 | A kind of compressed domain video watermark insertion and extracting method based on moving target |
| CN108521613A (en) * | 2018-05-15 | 2018-09-11 | 中南民族大学 | A kind of compressed domain video watermark insertion and extracting method based on moving target |
| RU2688202C1 (en) * | 2018-07-03 | 2019-05-21 | Закрытое акционерное общество "Перспективный мониторинг" | Method for hidden marking of digital television signal data stream |
| CN109922066A (en) * | 2019-03-11 | 2019-06-21 | 江苏大学 | Dynamic watermark insertion and detection method in a kind of communication network based on time slot feature |
| CN109922066B (en) * | 2019-03-11 | 2020-11-20 | 江苏大学 | Dynamic watermark embedding and detecting method based on time slot characteristics in communication network |
| CN109982157A (en) * | 2019-04-01 | 2019-07-05 | 北京火眼目测科技有限公司 | The method, system and storage medium of implicit fingerprint are embedded in digital video |
| CN110062291A (en) * | 2019-04-29 | 2019-07-26 | 王子孟 | A kind of digital watermarking addition and extracting method, apparatus and system |
| CN112991742A (en) * | 2021-04-21 | 2021-06-18 | 四川见山科技有限责任公司 | Visual simulation method and system for real-time traffic data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102307320B (en) | 2013-07-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102307320B (en) | Piracy tracing watermarking method applicable to streaming media environment | |
| Noorkami et al. | Compressed-domain video watermarking for H. 264 | |
| US7738676B1 (en) | Client-side watermarking using hybrid I-frames | |
| He et al. | A real-time dual watermarking algorithm of H. 264/AVC video stream for Video-on-Demand service | |
| US10277656B2 (en) | Method and system for delivering media data | |
| US11611808B2 (en) | Systems and methods of preparing multiple video streams for assembly with digital watermarking | |
| Ogawa et al. | Watermarking for HEVC/H. 265 stream | |
| Choudry et al. | A survey paper on video steganography | |
| Park et al. | Combined scheme of encryption and watermarking in H. 264/Scalable Video Coding (SVC) | |
| EP2225882A2 (en) | Method of coding a scalable video stream destined for users with different profiles | |
| CN103237209A (en) | H264 video watermarking method based on regional DCT (discrete cosine transform) coefficients | |
| Zhao et al. | Bandwidth efficient fingerprint multicast for video streaming | |
| Janu et al. | Development of an efficient real-time H. 264/AVC advanced video compression encryption scheme | |
| Ayyappan et al. | A review on reversible data hiding techniques | |
| Lian et al. | Secure and traceable multimedia distribution for convergent Mobile TV services | |
| Ali et al. | Multi-layer watermarking of H. 264/AVC video using Differential Expansion on IPCM blocks | |
| Lian et al. | Joint fingerprint embedding and decryption for video distribution | |
| Chattopadhyay et al. | H. 264 compressed domain watermarking in content delivery network (CDN) environment | |
| An et al. | A Fast Data Hiding Algorithm for H. 264/AVC Video in Bitstream Domain | |
| Wang et al. | Robust data hiding in MPEG-2 video against transcoding | |
| CN110012296B (en) | Rapid information embedding method for MP4 video code stream | |
| Kopilovic et al. | Video-DNA: Large-scale server-side watermarking | |
| Basha et al. | Survey on patended and unpatented video watermarking techniques | |
| Liu et al. | Desynchronization in compression process for collusion resilient video fingerprint | |
| Zheng et al. | An efficient method to hide information in MPEG video sequences |
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 |