CN105263025B - A kind of video Zero watermarking method based on time-space domain - Google Patents
A kind of video Zero watermarking method based on time-space domain Download PDFInfo
- Publication number
- CN105263025B CN105263025B CN201510745110.5A CN201510745110A CN105263025B CN 105263025 B CN105263025 B CN 105263025B CN 201510745110 A CN201510745110 A CN 201510745110A CN 105263025 B CN105263025 B CN 105263025B
- Authority
- CN
- China
- Prior art keywords
- frame
- key frame
- time
- watermark
- time domain
- 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 33
- 238000000605 extraction Methods 0.000 claims abstract description 10
- 230000007246 mechanism Effects 0.000 claims abstract description 7
- 210000000689 upper leg Anatomy 0.000 claims description 9
- 238000007689 inspection Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 238000009792 diffusion process Methods 0.000 claims description 2
- 230000010354 integration Effects 0.000 claims description 2
- 230000035772 mutation Effects 0.000 claims description 2
- 238000012360 testing method Methods 0.000 claims description 2
- 238000013461 design Methods 0.000 abstract description 4
- 230000011218 segmentation Effects 0.000 abstract description 2
- 230000001360 synchronised effect Effects 0.000 abstract description 2
- 230000009977 dual effect Effects 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 19
- 239000012634 fragment Substances 0.000 description 7
- 230000008859 change Effects 0.000 description 6
- 238000003780 insertion Methods 0.000 description 4
- 230000037431 insertion Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000006835 compression Effects 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention discloses a kind of video Zero watermarking method based on time-space domain, including video scene segmentation, synchronized using finite state machine Mechanism Design and key-frame extraction,3D‑HarrisAngle point set up, time-domain gray average data set structure, time-domain extreme value data set structure, dual zero watermarking structure and etc..This method is based on video content features in time-space domain and forms watermark data, the shortcomings of the prior art is overcome, on the basis of resist geometric attacks; color attack and synchronization attack can be fought simultaneously; and zero watermarking ambiguity problem is solved, it is used for the copyright protection of video content, there is high robust feature.
Description
Technical field
The present invention relates to digital watermark technology fields, and in particular to a kind of video Zero watermarking method based on time-space domain.
Background technology
With the development of Internet technology, multimedia technology, including audio, image and video etc. have been enter into our daily lifes
In work.Wherein, video is widely used as important medium role, thus the work of its copyright protection seems and becomes more and more important.Herein
Under background, copyright information is represented with digital watermarking, easily and effectively realize copyright protection task, digital watermark technology obtains
Research extensively.By development for many years, application technology of the video watermark technology as digital watermarking in video copyright protecting,
It reaches its maturity, and there is powerful robustness, can effectively resist all kinds of attacks, including frame attack, compression attack, Filtering Attacks,
Color and contrast attack etc..Existing video watermark technology can be divided into two major classes, Yi Leiwei according to the processing mode to frame image
Keyframe techniques, one kind are space-time field technique.
Keyframe techniques：Key frame selection is carried out with certain standard first, then extracts feature from each key frame
Data realize the insertion of watermark information finally by the change of each key frame characteristic.Such technology again can be into one
Step carefully divides frequency domain and spatial domain treatment technology into.Frequency domain treatment technology Typical Representative is DFT, DCT, DWT, PCA, SVD etc., in frequency
Realize the insertion work of feature information extraction and watermark information in domain.Spatial processing technology Typical Representative is that SIFT and SURF is calculated
Method carries out the generation and insertion of watermark information using video frame space characteristics.Fail to make full use of using keyframe techniques merely
The feature of video media in the time domain, single key frame characteristic do not carry out covering the characteristic set of entire video,
To reduce the anti-attack ability of watermark, especially to the entirety of video file attack, such as overall color Transformation Attack and
Synchronization attack.
Space-time field technique：The feature of video in the time domain is made full use of, is combined with spatial domain and frequency domain, in time-space domain three
Video frequency feature data is extracted in dimension space, forms watermark information.Single-frame images characteristic extraction is carried out first, can be space
Domain SIFT algorithms can also be frequency domain DCT scheduling algorithms.Then from whole video angle, generated in time-domain characteristic to
Amount forms vector data collection.Finally, according to respective rule, based on data set, watermark data is generated.Space-time field technique is most
Pipe takes full advantage of the characteristic that video continues in the time domain, but under the premise of being not bound with keyframe techniques, to video frame into
Row whole SIFT, SURF etc. be easy to cause computationally intensive problem, are unsatisfactory for the requirement of video watermark real-time.
Zero watermarking technology：The robustness and invisibility of watermark are a pair of contradictory body always.The enhancing of invisibility is inevitable
Its robustness can be weakened, vice versa.Another problem of watermark is amount of capacity, embedded large capacity watermark information, can be to original
Media file causes the destruction of bigger.Zero watermarking technology has well solved these problems, which utilizes media file itself
Feature to form watermark information, and the watermark information is registered in the third-party institution, the copyright as media file is believed
Breath, to realize the copyright protection that do not destroy on the basis of raw media file itself.There are ambiguity to ask for certain zero watermarking technology
Topic, i.e., different media files may will produce same watermark information.
Robustness：Robustness embodies ability of the digital watermark to attack resistance.Extension of the video as image in the time domain,
Attacker can not only operate its single-frame images, can also carry out integrated operation, water to series of successive frames image
Print technology needs to resist various attacks.Typical video watermarking attack has：Geometric attack includes the rotation of video frame images
Turn, burst forth, translating and affine transformation；Frame is attacked or synchronization attack, including frame is cut, frame is inserted into, Frame switch and frame averagely etc.；
Color is attacked, including the variation of color change, luminance information and contrast variation etc.；Further include compression attack and Filtering Attacks in addition
Deng.Either keyframe techniques or space-time field technique, robustness have the characteristics that respective, and all outstanding behaviours is in one aspect
Or two aspects, do not have while the ability of resisting various attacks, the especially attack type to video integrated operation, such as color
Change attack and synchronization attack with contrast.
Invention content
Problem to be solved by this invention is：How a kind of video Zero watermarking method based on time-space domain, this method are provided
Overcome defect of the existing technology, effectively improve video watermark robustness, can make full use of video overall permanence, support simultaneously
Resist geometric attacks, frame attack, synchronization attack, color and contrast attack.
Technical problem proposed by the invention solves in this way：A kind of video zero watermarking side based on time-space domain is provided
Method, which is characterized in that include the following steps：
Step 1 scene cut
It using the method based on gray value, calculates per frame gradation of image mean value, cross threshold values using Change in Mean carries out as standard
Scene cut, and packet marking is done to scene；
Step 2 synchronization mechanism designs and key frame extraction
According to the research to video time domain characteristic, audio video synchronization mechanism is designed using finite state machine, to each
Video scene, the finite state machine current state and subsequent time state are for determining to be formed residing for watermark data key frame
Region and specific location, and the state of finite state machine is determined by gray average relationship between frame in region；
Step 3 3D-Harris angle point collection is built
Harris angle point calculating is carried out to each key frame, chooses stable point, and Harris points are extended in time domain,
It sets up, forms the Harris angle points combination in three dimensions, and record time domain coordinate；
Step 4 gray average data set is built
Angle point data gray mean value computation is carried out as unit of time domain to 3D-Harris angle point collection, forms gray average number
According to collection；
Step 5 extreme point data set is built
Gray average data set is segmented with predetermined distance in the time domain, extreme value calculating is carried out in section, respectively shape
At the maximum and minimum data set in time-domain；
Step 6 zero watermarking data generate
(1) it is based on gray average data set generation watermark data, sets numerical value threshold values, ash of the judgement larger and smaller than threshold values
Mean value number is spent, and two numbers are compared, watermark data is generated according to comparison result；
(2) it is based on extreme point data integration and gives birth to watermark data, extreme value is compared apart from interior in given time domain, is compared
As a result it is used as watermark data to generate standard；
According to the watermark data that above two rule generates, then the randomly ordered operation of step-by-step is carried out, generates final watermark
Data.
According to the video Zero watermarking method provided by the present invention based on time-space domain, which is characterized in that specific side in step 2
Method is：
To each scene, zero or one frame key frame is generated according to every four frame of state, forward one frame is walked after the completion, repeats the behaviour
Make.Finite state machine original state be designated value, subsequent time state according in four frames in addition to key frame three frame images gray scale
Mean value magnitude relationship (co-existing in 6 kinds of relationships) determines that each magnitude relationship represents a kind of state, and determines according to the state
It fixs which frame in one section of four frame is selected as key frame, and records the position of key frame in the time domain.
Entire finite state machine operation is repeated, all scenes is traversed, finally generates all key frames.
According to the video Zero watermarking method provided by the present invention based on time-space domain, which is characterized in that have in step 3,4,5
Body method is：
1. extracting Harris stable points in key frame：Harris angle point operations are carried out to each key frame images in scene,
And descending arrangement is carried out by its response R, top n angle point is chosen as stable point, and H and W is enabled to be respectively frame image level and hang down
Straight coordinate, then its Harris stablize point set be represented by
Hpoint (x, y)=I (x, y) | x=1 ..., H；Y=1 ... W }
2. time-space domain 3D-Harris angle point gray average data sets are built：Indicate that key frame images Harris stablizes with MGV
Point gray average enables N indicate Harris stable point quantity in t-th of key frame, F tables according to the diffusion of key frame in the time domain
Show all number of key frames in scene, then its gray average data set can be expressed as MGVT in the time domain：
MGVT=MGV (t) | t=1 ..., F }
Extreme point when indicating that distance is K key frame with LEV, including maximum and minimum, then it is in the time domain
Extension can be indicated with LEVT, and it is that the extreme value generated in distance K is counted out to enable E, then
LEV(t2)=extreme MGV (j) | j=1 ..., K } t2=1 ..., E/2
LEVT={ LEV (t2)|t2=1 ..., E/2 }.
According to the video Zero watermarking method provided by the present invention based on time-space domain, which is characterized in that specific side in step 6
Method is：
Using scene as base unit, respectively according to time domain gray average and time domain extreme value data set, watermark data is generated.
1. generating watermark information using gray average data set.In the scene, it is from D values with the frame pitch for meeting T inspections
1bit watermarks generate range, calculate the average value of all key frame Harris angle points gray averages (MGV) within the scope of this, are denoted as
Em, two threshold values are set, respectively high-order threshold values and low level threshold values are denoted as Thigh and Tlow respectively,
Thigh=Em+α×255
Tlow=Em+β×255
The parameter of α, β between 0-1, the size for controlling threshold values.
The MGV values of each key frame are compared with threshold values, it is number of key frames more than Thigh to enable H, L be less than
The number of key frames of Tlow.If H is more than L, watermark data " 1 " is generated, otherwise generates watermark data " 0 ".
2. generating watermark information using extreme point data set.According to T test stones, in the scene, to meet the frame of T inspections
Distance D values are that 1bit watermarks generate range, and Nmax and Nmin is enabled to indicate maximum LEV in each segmentmaxWith minimum LEVmin
Number.If Nmax is more than Nmin, watermark data " 1 " is generated, otherwise generates watermark data " 0 ".Finally, two groups of watermarks are believed
Breath carries out the randomly ordered operation of step-by-step, so that it may to generate final watermark data.
Beneficial effects of the present invention：
1, video time-space domain characteristic is made full use of, realizes the extraction to video global feature data.Harris angle points are expanded
Exhibition forms 3D-Harris angle point collection to time-domain.In the research of Contemporary Digital digital watermark, Harris angle points are stablized due to it
Property and resist geometric attacks ability it is strong and be used widely, and practical application focus mostly in image watermark or video watermark from
It dissipates in single-frame images.The present invention makes full use of video in the continuous characteristic of time-domain, and Harris angle points are extended to whole video
Domain chooses stable point from single-frame images Harris angle points, 3D-Harris angle point data sets is formed in the time domain, for extracting generation
The characteristic of table video entirety provides the foundation for the watermark information generation based on time-space domain.
2, multi-angle extracts characteristic, forms doubly time series model information.Common zero watermarking technology faces ambiguity problem,
The watermark information of different media files may there is a situation where identical.The present invention from two angle extraction video frequency feature datas,
Two independent watermark informations are formed, and two groups of watermark informations are subjected to random fusion treatment, ambiguity can be effectively reduced and deposit
Probability.
3, synchronous attack resistant ability is improved.The present invention use finite state machine, by be segmented it is progressive in a manner of carry out key frame
Detection is determined according to finite state machine current state when key frame position in leading portion, and according to when remaining frame average gray of leading portion
Relationship before value determines subsequent time state.The present invention carries out key frame detection using such mode, is effective against synchronization
Attack, including the insertion of frame deletion, frame, Frame switch etc., while Harris angle point calculation amounts can be effectively reduced, improve watermark information
Formation efficiency.
4, anti-color attacking ability is improved.The research of usual video watermark technology, concentrate on anti-Geometrical change, frame attack,
In terms of Filtering Attacks and compression attack.Characteristic extraction is carried out in this hair based on 3D-Harris angle point collection gray values, it can
Strong robustness ability is attacked to provide confrontation color.Either attacker carries out color in human vision tolerance band to video
Single channel/multichannel adjustment, setting contrast, or blank level adjustment is carried out, it can ensure the integrality of watermark information.
Description of the drawings
Fig. 1 scene cuts and fragment schematic diagram in scene；
Fig. 2 synchronization mechanism finite state machine schematic diagrames；
Fig. 3 3D-Harris data set generation flow charts；
Fig. 4 watermark information product process figures.
Specific implementation mode
The invention will be further described below in conjunction with the accompanying drawings：
Video is first carried out scene cut by characteristic of the present invention according to video in the time domain, then is carried out to each scene thin
Change slice, then utilizes finite state machine, traverse each video scene, determine key frame position.According to the crucial frame number of acquisition
According to progress Harris angle point calculating, the angle point for selecting stability high is extended in time domain, forms 3D-Harris angle point data
Collection.Angle steel joint data set carries out further data processing, based on angle point gray value, respectively with two different difference marks
Standard, gray average and extreme value carry out difference, and two groups of different watermark informations are generated with this.Two groups of watermark datas are carried out further
Random fusion, to generate final watermark data.
1, fragment in scene cut and scene
Scene cut and segmentation flow in scene are totally as shown in Figure 1.Gray average meter is carried out to video complete frames image
It calculates, using gray value continuous mutation as video scene partitioning boundary.After completing scene cut, fragment is carried out to each scene, often
One segment number of image frames is taken as N frames, and marks fragment position.In subsequent processing, we will determine the piece from partial segments
A certain frame is key frame in disconnected.
2, synchronization mechanism design and key frame extraction
Synchronization mechanism realizes that finite state machine design schematic diagram is as shown in Figure 2 based on finite state machine.Here, we
If fragment frame number value is 4 in scene, all fragments in each scene are traversed, key frame position is generated with this.
S is set as to each scene original state first0, first fragment in scene is selected to be traversed.State S0Decision is worked as
X frame is key frame in preceding 1-4 frames, except X frame, carries out gradation of image mean value computation to remaining three frames, and carry out to it
Arrangement operation can show that 6 kinds of arrangements are possible, NextState value (S is determined with this altogether0--S5), each state value, which corresponds to, to be determined to take piece
In disconnected key frame is not chosen in key frame position or segment.After completing this operation, key frame extraction is carried out into next segment.When
It has traversed in this scene all to have no progeny, has then entered in later scene and traversed, until completing whole to a vision operation.
3,3D-Harris angle points collection is built
To the key frame that the 2nd step generates, Harris angle point calculating is carried out to each frame first.It may will produce per frame multiple
Angle point, it is also possible to not have angle point, such as black transition frames when transition.Descending arrangement is carried out by its response R, N before choosing
For a angle point as stable point, it is respectively frame image level and vertical coordinate to enable H and W, then its Hariis stablizes point set and can indicate
For
Hpoint (x, y)=I (x, y) | x=1 ..., H；Y=1 ... W }
After obtaining key frame Harris and stablizing angle point, it is extended in time domain scale, key frame figure is indicated with MGV
As Harris stable point gray averages, N is enabled to indicate that Harris stable point quantity in t-th of key frame, F indicate that institute is related in scene
Key frame number, then its gray average data set can be expressed as MGVT in the time domain：
MGVT=MGV (t) | t=1 ..., F }
Extreme point when indicating that distance is K key frame with LEV, including maximum and minimum, then it is in the time domain
Extension can be indicated with LEVT, and it is that the extreme value generated in distance K is counted out to enable E, then
LEV(t2)=extreme MGV (j) | j=1 ..., K } t2=1 ..., E/2
LEVT={ LEV (t2)|t2=1 ..., E/2 }.
4, watermark data generates
The generation overview flow chart of watermark data is as shown in Figure 4.Using the two time-space domain data sets generated in step 3,
Gray average MGVT and extreme value LEVT carries out doubly time series model generation.Take random number D as generation 1bit watermarked frame distances first.
To generate the watermark information of sufficient length, AllFrames is enabled to indicate that video totalframes, D values meet
The generation of first group of watermark data：All MGVT values in the D of region are compared.To ensure MGVT values in anti-color
Stability in the variations such as coloured silk, contrast, white balance, increase by two threshold values, one for ensures selection MGVT values apart from
Marginal value 255 is denoted as Thigh, one for ensureing that its value in following edge value 256*20%, is denoted as within 256*20%
Tlow。
Thigh=Em+α×255
Tlow=Em+β×255
The parameter of α, β between 0-1, the size for controlling threshold values.
The MGV values of each key frame are compared with threshold values, it is number of key frames more than Thigh to enable H, L be less than
The number of key frames of Tow.If H is more than L, watermark data " 1 " is generated, otherwise generates watermark data " 0 ".
The generation of second group of watermark data：All LEVT values in the D of region are compared.Nmax and Nmin is enabled to indicate each
Maximum LEV in segmentmaxWith minimum LEVminNumber.If Nmax is more than Nmin, watermark data " 1 " is generated, is otherwise given birth to
At watermark data " 0 ".
Finally, first group of watermark information and second group of watermark information are subjected to out of order fusion, obtain final watermark information,
The watermark information is registered into third party's information bank, to realize the copyright protection to video file.
Claims (3)
1. a kind of video Zero watermarking method based on time-space domain, which is characterized in that include the following steps：
Step 1, using the method based on gray value, calculate per frame gradation of image mean value, be more than threshold values with mean value continuous mutation
Standard carries out scene cut, and does packet marking to scene；
Step 2 chooses key frame using synchronization mechanism to each scene and records the position of key frame in the time domain；
Step 3 carries out Harris angle point calculating to each key frame, chooses stable point, and stable point is extended in time domain,
It sets up, the Harris formed in three dimensions stablizes point set, and records time domain coordinate；
Step 4 gray average data set is built, and angle point data gray mean value is carried out as unit of time domain to 3D-Harris angle point collection
It calculates, forms gray average data set；
Step 5 extreme point data set build, gray average data set is segmented with predetermined distance in the time domain, in section into
Row extreme value calculates, the maximum and minimum data set being respectively formed in time-domain；
Step 6 zero watermarking data generate
(1) it is based on time domain gray average data set generation watermark data, sets numerical value threshold values, ash of the judgement larger and smaller than threshold values
Mean value number is spent, and two numbers are compared, watermark data is generated according to comparison result；
(2) it is based on extreme point data integration and gives birth to watermark data, extreme value is compared apart from interior in given time domain, comparison result
Standard is generated as watermark data；
According to the watermark data that above two rule generates, then the randomly ordered operation of step-by-step is carried out, generates final watermark data；
Specific method is：
Using scene as base unit, respectively according to time domain gray average data set and time domain extreme value data set, watermark data is generated；
1. generating watermark information using gray average data set, all key frame Harris angle point gray averages in scene are calculated
The average value of MGV, is denoted as Em, two threshold values are set, respectively high-order threshold values and low level threshold values are denoted as Thigh and Tlow respectively,
Thigh=Em+α×255
Tlow=Em+β×255
The parameter of α, β between 0-1, the size for controlling threshold values；
The MGV values of each key frame are compared with threshold values, it is the number of key frames more than Thigh to enable H, and L is less than Tlow
Number of key frames, if H be more than L, generate watermark data " 1 ", otherwise generate watermark data " 0 "；
2. using extreme point data set generate watermark information, according to T test stones, in the scene, with meet T inspection frame pitch from
D values are that 1bit watermarks generate range, and Nmax and Nmin is enabled to indicate maximum LEV in each segmentmaxWith minimum LEVminNumber,
If Nmax is more than Nmin, watermark data " 1 " is generated, otherwise generates watermark data " 0 "；
Finally, two groups of watermark informations are subjected to randomly ordered operation, so that it may to generate final watermark data.
2. the video Zero watermarking method according to claim 1 based on time-space domain, which is characterized in that specific side in step 2
Method is：
Step 2.1 walks forward one frame, weight after the completion to each scene according to every four frames, zero frame of generation of state or a frame key frame
The multiple operation, finite state machine original state are designated value, subsequent time state according in four frames in addition to key frame three frame images
Gray average magnitude relationship determine that each magnitude relationship represents a kind of state, next section four is determined according to the state
Which frame is selected as key frame in frame, and records the position of key frame in the time domain, next if generating zero frame key frame
Moment goes to original state and continues key frame extraction；
Step 2.2 repeats step 2.1, traverses all scenes, finally generates all key frames.
3. the video Zero watermarking method according to claim 1 based on time-space domain, which is characterized in that have in step 3,4,5
Body method is：
1. extracting Harris stable points in key frame：Harris angle point operations are carried out to each key frame images in scene, and are pressed
Its response R carries out descending arrangement, chooses top n angle point as stable point, it is respectively frame image level and vertical seat to enable H and W
It marks, then its Harris stablizes point set and is represented by
Hpoint (x, y)=I (x, y) | x=1 ..., H；Y=1 ... W }
2. time-space domain 3D-Harris angle point gray average data sets are built：Indicate that each key frame images Harris stablizes with MGV
Point gray average enables N indicate Harris stable point quantity in t-th of key frame, F tables according to the diffusion of key frame in the time domain
Show all number of key frames in region, then its gray average data set can be expressed as MGVT in the time domain：
MGVT=MGV (t) | t=1 ..., F }
I indicates that the Harris that each frame key frame images select stablizes angle point, and value range 0-N indicates that distance is with LEV
Extreme point when K key frame, including maximum and minimum, then its extension in the time domain can be indicated with LEVT, enable E
Extreme value to be generated in distance K is counted out, then
LEV(t2)=extreme MGV (j) | j=1 ..., K } t2=1 ..., E/2
LEVT={ LEV (t2)|t2=1 ..., E/2 }.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510745110.5A CN105263025B (en) | 2015-11-05 | 2015-11-05 | A kind of video Zero watermarking method based on time-space domain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510745110.5A CN105263025B (en) | 2015-11-05 | 2015-11-05 | A kind of video Zero watermarking method based on time-space domain |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105263025A CN105263025A (en) | 2016-01-20 |
| CN105263025B true CN105263025B (en) | 2018-11-02 |
Family
ID=55102506
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510745110.5A Expired - Fee Related CN105263025B (en) | 2015-11-05 | 2015-11-05 | A kind of video Zero watermarking method based on time-space domain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105263025B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107527010B (en) * | 2017-07-13 | 2020-07-10 | 央视国际网络无锡有限公司 | Method for extracting video gene according to local feature and motion vector |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102176208A (en) * | 2011-02-28 | 2011-09-07 | 西安电子科技大学 | Robust video fingerprint method based on three-dimensional space-time characteristics |
| CN102646259A (en) * | 2012-02-16 | 2012-08-22 | 南京邮电大学 | Anti-attack robustness multiple zero watermark method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8214741B2 (en) * | 2002-03-19 | 2012-07-03 | Sharp Laboratories Of America, Inc. | Synchronization of video and data |
| CN103177413B (en) * | 2011-12-20 | 2016-04-13 | 深圳市腾讯计算机系统有限公司 | The method that localization blind watermatking generates, detect and device |
-
2015
- 2015-11-05 CN CN201510745110.5A patent/CN105263025B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102176208A (en) * | 2011-02-28 | 2011-09-07 | 西安电子科技大学 | Robust video fingerprint method based on three-dimensional space-time characteristics |
| CN102646259A (en) * | 2012-02-16 | 2012-08-22 | 南京邮电大学 | Anti-attack robustness multiple zero watermark method |
Non-Patent Citations (1)
| Title |
|---|
| Minmum mean squared error time series classification using an echo state network prediction model;M. D. Skowronski, et al;《Circuits and Systems, 2006. ISCAS 2006. Proceedings. 2006 IEEE International Symposium on》;20060524;第3153－3156页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105263025A (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109903223B (en) | Image super-resolution method based on dense connection network and generation type countermeasure network | |
| CN104091302B (en) | A kind of robust watermarking insertion and extracting method based on multiscale space | |
| KR20060121972A (en) | Electronic watermarking method, electronic watermark detecting method, apparatus and program | |
| CN108829711A (en) | A kind of image search method based on multi-feature fusion | |
| CN110349132A (en) | A kind of fabric defects detection method based on light-field camera extraction of depth information | |
| CN102393900A (en) | Video copying detection method based on robust hash | |
| CN106131711B (en) | The hidden watermark insertion of robust and extracting method towards 3D high-definition digital video | |
| TWI646467B (en) | Method and system for generating two dimensional barcode with hidden data | |
| CN103325082A (en) | Vector map reversible information hiding method based on LSD planes | |
| CN102147912A (en) | Adaptive difference expansion-based reversible image watermarking method | |
| CN106204410B (en) | A kind of novel digital watermark method decomposed based on matrix Schur | |
| CN115424209A (en) | Crowd counting method based on spatial pyramid attention network | |
| CN111091151A (en) | Method for generating countermeasure network for target detection data enhancement | |
| CN103366332B (en) | A kind of image watermark method based on depth information | |
| CN101533509B (en) | A three-dimensional grid splitting method of blind watermark | |
| KR101580987B1 (en) | A watermarking method for 3D stereoscopic image based on depth and texture images | |
| CN105263025B (en) | A kind of video Zero watermarking method based on time-space domain | |
| CN111798359A (en) | Deep learning-based image watermark removing method | |
| CN105741222B (en) | A kind of steganography information locating method based on the estimation of pixel subset insertion rate | |
| CN103377455A (en) | Three-dimensional geographic model digital watermarking method with copyright protection service orientation | |
| CN108470318A (en) | The three-dimensional grid doubly time series model method positioned based on grouping strategy and neighborhood relationships | |
| CN104318505A (en) | Three-dimensional mesh model blind watermarking method based on image discrete cosine transformation | |
| CN105828061B (en) | A kind of virtual view quality evaluating method of view-based access control model masking effect | |
| CN102314667B (en) | Vertex weight value-based OBJ (object)-format three-dimensional model digital-watermarking method | |
| Soliman et al. | Robust watermarking approach for 3D triangular mesh using self organization map |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Peng Dezhong Inventor after: Luo Yifan Inventor after: Xiang Yong Inventor after: Zhang Lijun Inventor after: Yin Dawei Inventor after: Jiang Rui Inventor after: Sang Yongsheng Inventor after: Peng Xi Inventor before: Peng Dezhong Inventor before: Luo Yifan Inventor before: Xiang Yong |
|
| TA01 | Transfer of patent application right |
Effective date of registration: 20171221 Address after: 610000 Chengdu City, Chengdu, Sichuan, China (Sichuan) free trade test zone, Chengdu City, Chengdu City, Tianfu New District, Tianfu Road south section, No. 1609, Jing Rong mansion, 16 Applicant after: Sichuan Zhi Qian science and Technology Co.,Ltd. Address before: 610065 1 building, No. 11, No. 11, Xiaojiahe street, Xiaojiahe high new zone, Sichuan Province Applicant before: CHENGDU RUIBEI YINGTE INFORMATION TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20180913 Address after: 610000 1 building, 2 Xiaojiahe Zheng street, Chengdu High-tech Zone, Sichuan, China, 11 Applicant after: CHENGDU RUIBEI YINGTE INFORMATION TECHNOLOGY Co.,Ltd. Address before: 610000 Sichuan Chengdu China (Sichuan) Free Trade Experimental Zone Applicant before: Sichuan Zhi Qian science and Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20181102 |