CN103458256A - Methods and devices for watermarking stereoscopic video and extracting watermarks, stereoscopic video generation device and stereoscopic video play device - Google Patents

Methods and devices for watermarking stereoscopic video and extracting watermarks, stereoscopic video generation device and stereoscopic video play device Download PDF

Info

Publication number
CN103458256A
CN103458256A CN 201210169598 CN201210169598A CN103458256A CN 103458256 A CN103458256 A CN 103458256A CN 201210169598 CN201210169598 CN 201210169598 CN 201210169598 A CN201210169598 A CN 201210169598A CN 103458256 A CN103458256 A CN 103458256A
Authority
CN
China
Prior art keywords
described
video
image coefficient
right wing
restriction relation
Prior art date
Application number
CN 201210169598
Other languages
Chinese (zh)
Inventor
伍淳华
袁开国
杨义先
张智辉
Original Assignee
索尼公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 索尼公司 filed Critical 索尼公司
Priority to CN 201210169598 priority Critical patent/CN103458256A/en
Publication of CN103458256A publication Critical patent/CN103458256A/en

Links

Abstract

The invention discloses methods and devices for watermarking stereoscopic video and extracting watermarks, a stereoscopic video generation device and a stereoscopic video playback device. The device for watermarking the stereoscopic video comprises a stereoscopic video decomposition component, a left-sided watermark embedding part, a right-sided watermark embedding part and a stereoscopic video synthesis component. The stereoscopic video decomposition component acquires left-sided video and right-sided video from the stereoscopic video. The left-sided watermark embedding part adjusts image coefficients of an embedded position of the left-sided video according to corresponding left-sided image coefficient limiting relation. The right-sided watermark embedding part adjusts image coefficients of an embedded position of the right-sided video according to corresponding right-sided image coefficient limiting relation. The stereoscopic video synthesis component synthesizes the left-sided video and the right-sided video into the stereoscopic video once again. The left-sided image coefficient limiting relation and the right-sided image coefficient limiting relation corresponding to the same bit value of a watermarking sequence are different, and image coefficient limiting relation relevant to image coefficients of the left-sided video and the right-sided video can be determined through the two relations.

Description

Three-dimensional video-frequency adds watermark, watermark extraction apparatus and method, generation and playing device

Technical field

Present invention relates in general to digital watermark technology, more specifically, the present invention relates to a kind of for stereoscopic video add the apparatus and method of watermark, for extract apparatus and method, three-dimensional video-frequency generating apparatus and the stereoscopic video playing device of watermark from three-dimensional video-frequency.

Background technology

Three-dimensional video-frequency is an important development direction of following video technique.Three-dimensional video-frequency has utilized the binocular parallax principle of human eye, by use two video cameras to gather the vision signal of Same Scene simultaneously, obtain left and right two-path video signal, by certain approach, the two-path video signal is sent to respectively again to people's two of left and right, what make that left eye sees is the image of taking back, and what right eye was seen is the image taken over, and forms binocular parallax, finally, by the synthetic next scene that is virtually reality like reality of people's brain, reach the effect of three-dimensional video-frequency.Three-dimensional film, since coming out, has obtained numerous spectators' favor, and its market share rapidly increases.But a kind of as digital multimedia, three-dimensional film still has and is easy to the characteristic of revising, can perfect copying and can worldwide need hardly the distribution of cost ground, and its copyright protection becomes problem demanding prompt solution.

Digital watermark technology is one of effective and key means of protection digital media content.With other technologies, compare, digital watermarking has can not perception, with the host data seamless combination and the advantage such as be not easy to be removed and distort.The above-mentioned advantage of digital watermarking makes its copyright protection at image, audio frequency and single channel video and authenticated connection obtain application extremely widely, but need perfect in the application aspect the three-dimensional video-frequency copyright protection.

For the single channel video, three-dimensional video-frequency has left and right two-path video.Watch two-path video by special-purpose equipment simultaneously, can obtain the effect of 3 D stereo video.Simultaneously, any road video also can be used as the single channel video-see.Therefore, not only need to extract watermark in complete stereo video data, also need to extract watermark from any road single channel video.

Existing three-dimensional video-frequency digital watermark can be divided into two classes.List of references 1(John N.Ellinas:Reversible Watermarking on Stereo Image Sequences.International Journal of Information and Communication Engineering.5:3 2009,210-215) proposed a kind of reversible three-dimensional video-frequency watermarking method, the data of right wing video have been hidden in the video of left road.Like this, only need to transmit the data of left road video, just can obtain three-dimensional video-frequency at receiving terminal.List of references 2(Zhang Z, Zhu Z, Xi L:Novel scheme for watermarking stereo video.International Journal of Nonlinear Science.3:(2007), 74-80) watermark is hidden in the distinctive parallax data of three-dimensional video-frequency.Yet this two classes three-dimensional video-frequency digital watermark all can't meet the requirement of extracting watermark from any single channel video.

Summary of the invention

Provided hereinafter about brief overview of the present invention, in order to the basic comprehension about some aspect of the present invention is provided.Should be appreciated that this general introduction is not about exhaustive general introduction of the present invention.It is not that intention is determined key of the present invention or pith, and nor is it intended to limit the scope of the present invention.Its purpose is only that the form of simplifying provides some concept, usings this as the preorder in greater detail of discussing after a while.

Above-mentioned condition in view of prior art, an object of the present invention is to propose a kind of for stereoscopic video add the apparatus and method of watermark, for extract apparatus and method, three-dimensional video-frequency generating apparatus and the stereoscopic video playing device of watermark from three-dimensional video-frequency, it not only can extract watermark from complete three-dimensional video-frequency, also can the video independent from any road, extract watermark.

According to an aspect of the present invention, a kind of device that adds watermark for stereoscopic video is provided, comprise: left road watermark built in items, the corresponding left road of value image coefficient restriction relation for the bit in the watermark sequence that will embed on each the predetermined embedded location according to described left road video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in the video of described left road; Right wing watermark built in items, the corresponding right wing image coefficient of value restriction relation for the bit in the described watermark sequence that will embed on each the predetermined embedded location according to described right wing video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in described right wing video; And the three-dimensional video-frequency compound component, for the described left road video that will embed respectively described watermark sequence and right wing video compound stereoscopic video again.Wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation can determine with described left road video in the embedded location of described same bit value on image coefficient and the relevant image coefficient restriction relation all of the image coefficient on the corresponding embedded location of the described same bit value in described right wing video.

According to a further aspect in the invention, provide a kind of for extract the device of watermark from three-dimensional video-frequency, having comprised: the three-dimensional video-frequency Knock-Down Component, for obtain left road video and right wing video from three-dimensional video-frequency; Watermark bit is extracted parts, the stereo-picture coefficient restriction relation satisfied with the image coefficient on corresponding embedded location in described right wing video for the image coefficient on each the predetermined embedded location by judging described left road video, extract the corresponding value of described stereo-picture coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the video of described left road and the corresponding embedded location in described right wing video; And watermark sequence forms parts, for the value of the bit of the watermark sequence according to extracted, form described watermark sequence.Wherein, described stereo-picture coefficient restriction relation is determined according to corresponding Zuo road image coefficient restriction relation and right wing image coefficient restriction relation, described left road image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described left road video to embed described watermark in the video of described left road, described right wing image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described right wing video to embed described watermark in described right wing video, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation.

According to a further aspect in the invention, provide a kind of three-dimensional video-frequency generating apparatus, having comprised: left road video acquisition parts, for gathering for three-dimensional video-frequency Zuo road video; Right wing video acquisition parts, for gathering the right wing video for described three-dimensional video-frequency; Left road watermark built in items, the corresponding left road of value image coefficient restriction relation for the bit in the watermark sequence that will embed on each the predetermined embedded location according to described left road video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in the video of described left road; Right wing watermark built in items, the corresponding right wing image coefficient of value restriction relation for the bit in the described watermark sequence that will embed on each the predetermined embedded location according to described right wing video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in described right wing video; And the three-dimensional video-frequency compound component, for described left road video and the right wing video compound stereoscopic video that will embed respectively described watermark sequence.Wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation can determine with described left road video in the embedded location of described same bit value on image coefficient and the relevant image coefficient restriction relation all of the image coefficient on the corresponding embedded location of the described same bit value in described right wing video.

According to a further aspect in the invention, provide a kind of stereoscopic video playing device, having comprised: the watermark extracting parts, extract watermark for the three-dimensional video-frequency from input.Described watermark extracting parts comprise: the three-dimensional video-frequency Knock-Down Component, for obtain left road video and right wing video from described three-dimensional video-frequency; Watermark bit is extracted parts, the stereo-picture coefficient restriction relation satisfied with the image coefficient on corresponding embedded location in described right wing video for the image coefficient on each the predetermined embedded location by judging described left road video, extract the corresponding value of described stereo-picture coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the video of described left road and the corresponding embedded location in described right wing video; Watermark sequence forms parts, for the value of the bit of the watermark sequence according to extracted, forms described watermark sequence; And the watermark sequence verification component, determine for the correlation of the watermark sequence according to extracted and target watermark sequence whether extracted watermark sequence is described target watermark sequence.Described stereoscopic video playing device also comprises the video playback parts, for working as extracted watermark sequence while being described target watermark sequence, plays described three-dimensional video-frequency.Wherein, described stereo-picture coefficient restriction relation is determined according to corresponding Zuo road image coefficient restriction relation and right wing image coefficient restriction relation, described left road image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described left road video to embed described watermark in the video of described left road, described right wing image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described right wing video to embed described watermark in described right wing video, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation.

According to a further aspect in the invention, provide a kind of method that adds watermark for stereoscopic video, having comprised: obtained left road video and right wing video from three-dimensional video-frequency; Be adjusted at the image coefficient on described embedded location according to the corresponding left road of the value image coefficient restriction relation of the bit in the watermark sequence that will embed on each the predetermined embedded location in the video of described left road, to embed described watermark sequence in the video of described left road; Be adjusted at the image coefficient on described embedded location according to the corresponding right wing image coefficient of the value restriction relation of the bit in the described watermark sequence that will embed on each the predetermined embedded location in described right wing video, to embed described watermark sequence in described right wing video; And will embed respectively the described left road video of described watermark sequence and right wing video compound stereoscopic video again.Wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation can determine with described left road video in the embedded location of described same bit value on image coefficient and the relevant image coefficient restriction relation all of the image coefficient on the corresponding embedded location of the described same bit value in described right wing video.

According to a further aspect in the invention, provide a kind of for extract the method for watermark from three-dimensional video-frequency, having comprised: obtain left road video and right wing video from three-dimensional video-frequency; By the stereo-picture coefficient restriction relation of judging that image coefficient on each the predetermined embedded location in the video of described left road is satisfied with the image coefficient on corresponding embedded location in described right wing video, extract the corresponding value of described stereo-picture coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the video of described left road and the corresponding embedded location in described right wing video; And form described watermark sequence according to the value of the bit of extracted watermark sequence.Wherein, described stereo-picture coefficient restriction relation is determined according to corresponding Zuo road image coefficient restriction relation and right wing image coefficient restriction relation, described left road image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described left road video to embed described watermark in the video of described left road, described right wing image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described right wing video to embed described watermark in described right wing video, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation.

Method and apparatus according to above-mentioned aspect of the present invention, use for limiting in the video of image coefficient restriction relation Zuo road, frame of video image coefficient Zuo road, left road watermarked, and use the right wing image coefficient restriction relation for limiting right wing frame of video image coefficient watermarked in the right wing video, therefore, can extract watermark sequence from the video of independent Zuo road according to left road image coefficient restriction relation, and extract watermark sequence according to right wing image coefficient restriction relation from independent right wing video.

In addition, when watermarked in three-dimensional video-frequency, the difference Zuo road video watermark sequence identical with embedding in the right wing video, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation is different from right wing image coefficient restriction relation, and, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation can be determined all relevant image coefficient restriction relations of image coefficient on the corresponding embedded location of image coefficient on the embedded location of the described same bit value in the video of Yu Zuo road and the same bit value in the right wing video.Thus, can be according to according to left road image coefficient restriction relation and the definite image coefficient restriction relation of right wing image coefficient restriction relation, as stereo-picture coefficient restriction relation, extracting watermark sequence from three-dimensional video-frequency.

In addition, when two-path video suffers same attack, certain variation can occur in image coefficient, but the change direction of the image coefficient in two-path video is consistent.Like this, the relation between the coefficient of two-path video can be relatively stable.Therefore, according to according to left road image coefficient restriction relation and the definite stereo-picture coefficient restriction relation of right wing image coefficient restriction relation, extracting watermark, can improve the robustness of watermark extracting.

The accompanying drawing explanation

The present invention can, by reference to hereinafter given description and being better understood by reference to the accompanying drawings, wherein use same or analogous Reference numeral to mean identical or similar parts in institute's drawings attached.Described accompanying drawing comprises in this manual and forms the part of this specification together with following detailed description, and is used for further illustrating the preferred embodiments of the present invention and explains principle and advantage of the present invention.In the accompanying drawings:

Fig. 1 illustrates the schematic block diagram that adds the device of watermark for stereoscopic video according to an embodiment of the invention;

Fig. 2 illustrates the indicative flowchart that adds the method for watermark for stereoscopic video according to an embodiment of the invention;

Fig. 3 illustrate according to another embodiment of the present invention add the schematic block diagram of the device of watermark for stereoscopic video;

Fig. 4 illustrate according to another embodiment of the invention add the indicative flowchart of the method for watermark for stereoscopic video;

Fig. 5 illustrates the schematic block diagram of left road watermark built in items according to an embodiment of the invention;

Fig. 6 illustrates the schematic block diagram of right wing watermark built in items according to an embodiment of the invention;

Fig. 7 illustrates the schematic block diagram of Zuo road watermark built in items according to another embodiment of the invention;

Fig. 8 illustrates the indicative flowchart of left road according to an embodiment of the invention watermark embed process;

Fig. 9 illustrates the schematic block diagram of right wing watermark built in items according to another embodiment of the invention;

Figure 10 illustrates the indicative flowchart of right wing watermark embed process according to an embodiment of the invention;

Figure 11 a-11d illustrates the schematic diagram of image coefficient matrix according to an embodiment of the invention;

Figure 12 a-12c illustrates the schematic diagram of image coefficient matrix according to another embodiment of the invention;

Figure 13 illustrates according to an embodiment of the invention for extract the schematic block diagram of the device of watermark from three-dimensional video-frequency;

Figure 14 illustrates according to an embodiment of the invention for extract the indicative flowchart of the method for watermark from three-dimensional video-frequency;

Figure 15 illustrates the schematic block diagram for the device from three-dimensional video-frequency extraction watermark according to another embodiment of the invention;

Figure 16 illustrates the indicative flowchart for the method from three-dimensional video-frequency extraction watermark according to another embodiment of the invention

Figure 17 illustrates the schematic block diagram for the device from three-dimensional video-frequency extraction watermark according to another embodiment of the invention;

Figure 18 illustrates the schematic block diagram of three-dimensional video-frequency generating apparatus according to an embodiment of the invention;

Figure 19 illustrates the indicative flowchart of three-dimensional video-frequency generation method according to an embodiment of the invention;

Figure 20 illustrates the schematic block diagram of stereoscopic video playing device according to an embodiment of the invention;

Figure 21 illustrates the indicative flowchart of three-dimensional video-frequency player method according to an embodiment of the invention; And

Figure 22 illustrates the block diagram of the structure of the computer that can realize embodiments of the invention/example.

Embodiment

Embodiments of the invention are described below with reference to accompanying drawings.The element of describing in an accompanying drawing of the present invention or a kind of execution mode and feature can combine with element and feature shown in one or more other accompanying drawing or execution mode.It should be noted that for purposes of clarity, omitted expression and the description of unrelated to the invention, parts known to persons of ordinary skill in the art and processing in accompanying drawing and explanation.

Below, exemplary embodiment will be described in the following sequence.

1. add the apparatus and method of watermark for stereoscopic video

2. for extract the apparatus and method of watermark from three-dimensional video-frequency

3. three-dimensional video-frequency generating apparatus and method

4. stereoscopic video playing device and method

5. can implement the computer configuation of embodiments of the invention/example

<1. for stereoscopic video, add the apparatus and method of watermark >

Fig. 1 illustrates the schematic block diagram that adds the device of watermark for stereoscopic video according to an embodiment of the invention.For brevity, only show the closely-related part with the present invention in figure.As shown in Figure 1, the device (adding watermarking device) 100 that adds watermark for stereoscopic video comprises three-dimensional video-frequency Knock-Down Component 110,Zuo road watermark built in items 120, right wing watermark built in items 130 and three-dimensional watermark compound component 140.The exemplary operation flow process that adds watermarking device 100 is described below with reference to Fig. 2.

Fig. 2 illustrates the indicative flowchart that adds the method for watermark for stereoscopic video according to an embodiment of the invention.As shown in Figure 2, at step S210(, can be carried out by three-dimensional video-frequency Knock-Down Component 110) in, left road video and right wing video obtained from three-dimensional video-frequency.The watermarked stage is identical with the stage of extracting watermark, and extracting watermark is watermarked contrary the processing.

At step S220(Ke Youzuo road watermark built in items 120, carry out) in, be adjusted at the image coefficient on described embedded location, watermarked sequence in the video of Yi Zuo road according to the corresponding left road of the value image coefficient restriction relation of the bit in the watermark sequence that will embed on each the predetermined embedded location in the video of left road.

Accordingly, at step S230(, can be carried out by right wing watermark built in items 130) in, be adjusted at the image coefficient on described embedded location according to the corresponding right wing image coefficient of the value restriction relation of the bit in the watermark sequence that will embed on each the predetermined embedded location in the right wing video, to embed described watermark sequence in the right wing video.The watermark sequence embedded in the watermark sequence Yu Zuo road video embedded in the right wing video is identical.

Digital watermarking is such as comprising text, binary picture, binary sequence etc.And watermarked process is that digital watermarking is converted to corresponding watermark sequence and watermark sequence is embedded to the process such as the destination carrier of video etc.Watermark sequence generally adopts binary sequence.According to the inner capacities of watermark sequence, can pre-determine a complete watermark and will be embedded in the single frame or a plurality of frame in video, and the embedded location of each bit in described single frame or a plurality of frame in watermark sequence.

In this article, image coefficient means the value of the data element in frame of video.If frame of video is in spatial domain (also claiming base band domain), image coefficient means the value (also claiming the base band coefficient) of the pixel in frame of video.If frame of video is in transform domain, image coefficient means the conversion coefficient in frame of video, i.e. the value of the data element in the frame of video after the conversion.

In the watermark embedded technology based on coefficient, the corresponding a kind of image coefficient restriction relation of each bit value in watermark sequence.The image coefficient restriction relation is limited the value of the image coefficient in frame of video.Be adjusted at the image coefficient on the embedded location in video, the corresponding image coefficient restriction relation of the value that makes it meet the watermark bit that will embed at this embedded location, show to have embedded this watermark bit at this embedded location.As example rather than restriction ,Zuo road and right wing image coefficient restriction relation, can be linear inequality or nonlinear inequalities.

At step S240(, can be carried out by three-dimensional video-frequency compound component 140) in, will distinguish watermarked sequence Zuo road video and right wing video compound stereoscopic video again.Should be appreciated that and can carry out by the form identical with original three-dimensional video-frequency compound stereoscopic video again.

In the present embodiment, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation is different from right wing image coefficient restriction relation.For example, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation can be contrary to the adjustment direction of image coefficient.Perhaps, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation can be identical to the adjustment direction of image coefficient but adjusting range is different.This will the following specifically describes.

In addition, in the present embodiment, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation can be determined all relevant image coefficient restriction relations of image coefficient on the corresponding embedded location of image coefficient on the embedded location of the same bit value in the video of Yu Zuo road and the same bit value in the right wing video.Corresponding embedded location in the right wing video refers to the embedded location of embedded location in same position in the video of Yu Zuo road.Determined image coefficient restriction relation can be for extracting watermark sequence from three-dimensional video-frequency.This will the following specifically describes.

In an embodiment of the present invention, use for limiting in the video of image coefficient restriction relation Zuo road, frame of video image coefficient Zuo road, left road watermarked, and use the right wing image coefficient restriction relation for limiting right wing frame of video image coefficient watermarked in the right wing video, therefore, can extract watermark sequence from the video of independent Zuo road according to left road image coefficient restriction relation, and extract watermark sequence according to right wing image coefficient restriction relation from independent right wing video.

In addition, the image coefficient that can determine Yu Zuo road video and right wing video by the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation is relevant image coefficient restriction relation all, as the corresponding stereo-picture coefficient of this bit value restriction relation.Therefore, can also extract watermark sequence from three-dimensional video-frequency according to stereo-picture coefficient restriction relation.

In addition, when two-path video suffers same attack, certain variation can occur in image coefficient, but the change direction of the image coefficient in two-path video is generally consistent.Like this, the relation between the coefficient of two-path video can be relatively stable.Therefore, according to according to left road image coefficient restriction relation and the definite stereo-picture coefficient restriction relation of right wing image coefficient restriction relation, extracting watermark, can improve the robustness of watermark extracting.

As mentioned above, in watermark sequence, the embedded location of each bit in frame of video is predetermined.This can carry out and will determine that result is input to according to an embodiment of the invention by external component and add in watermarking device, or, also can carry out by the embedded location determining means is provided in adding watermarking device according to an embodiment of the invention.Fig. 3 illustrate according to another embodiment of the present invention add the schematic block diagram of the device of watermark for stereoscopic video.Compare with the embodiment shown in Fig. 1, in the embodiment shown in fig. 3, add in watermarking device 100 and increased embedded location determining means 150, identical with in Fig. 1 of miscellaneous part.Correspondingly, as shown in Figure 4, according to another embodiment of the invention add the method for watermark for stereoscopic video, with Fig. 2, comparing has increased step S250(and has been carried out by embedded location determining means 150), determine the embedded location in each bit Zuo road frame of video and right wing frame of video in watermark sequence.Particularly, can be according to the length of watermark sequence, determine embedded location in the frame of the watermark sequence to be embedded of each bit Zuo road video in watermark sequence and the embedded location in the frame of the watermark sequence to be embedded of right wing video.Here the frame of the watermark sequence to be embedded of ,Zuo road video can be single frame, can be also a plurality of frames.Equally, the frame of the watermark sequence to be embedded of right wing video can be single frame, can be also a plurality of frames.

If watermark sequence is longer, be embedded in single frame and may affect the perception of people to single frame, therefore longer watermark sequence can be embedded in a plurality of frames.Embedded location determining means 150 can also determine that single frame or a plurality of frame in the video of Yong Zuo road carry out watermarked sequence according to the length of watermark sequence, and embeds described watermark sequence with the single frame in the right wing video or a plurality of frame.

As the specific embodiment of left road watermark built in items 120 and right wing watermark built in items 130, Fig. 5 and Fig. 6 illustrate respectively the schematic block diagram of left according to an embodiment of the invention road and right wing watermark built in items.,Zuo road watermark built in items 120 can comprise coefficient decision parts 121He Zuo road, left road coefficient adjustment parts 122 as shown in Figure 5.Whether left road coefficient decision parts 121 meet the corresponding left road of the value image coefficient restriction relation that will be embedded in the bit in the watermark sequence on described embedded location for the image coefficient on each embedded location of judging left road video.Left road coefficient adjustment parts 122, in the situation that the image coefficient on embedded location does not meet corresponding left road image coefficient restriction relation, are adjusted at the image coefficient on described embedded location so that the image coefficient through adjusting meets corresponding Zuo road image coefficient restriction relation.As shown in Figure 6, right wing watermark built in items 130 can comprise right wing coefficient decision parts 131 and right wing coefficient adjustment parts 132.Whether right wing coefficient decision parts 131 meet the corresponding right wing image coefficient of the value restriction relation that will be embedded in the bit in the watermark sequence on described embedded location for the image coefficient on each embedded location of judging the right wing video.Right wing coefficient adjustment parts 132, in the situation that the image coefficient on embedded location does not meet corresponding right wing image coefficient restriction relation, are adjusted at the image coefficient on described embedded location so that the image coefficient through adjusting meets corresponding right wing image coefficient restriction relation.

When image coefficient is the pixel value in spatial domain, do not need frame of video is converted.When image coefficient is the conversion coefficient in transform domain, need first frame of video to be transformed in transform domain.In certain embodiments, in order to save computational resource, the frame of video of watermark to be embedded can also be divided into to the piece of pre-sizing, in watermark sequence, the embedded location of each bit is distributed in each piece.In the situation that frame of video has been carried out to piecemeal, if frame of video is converted, can be converted each piece in frame of video.As example rather than restriction, following Description Image coefficient is conversion coefficient situation Xia Zuo road in transform domain and the example of right wing watermark built in items.

Fig. 7 illustrates the schematic block diagram of Zuo road watermark built in items according to another embodiment of the invention.In Fig. 7, except comprising coefficient decision parts 121He Zuo road, Zuo same as shown in Figure 5 road coefficient adjustment parts 122, left road watermark built in items 120 ' also comprises left road transform component 124, and (by empty wire frame representation) comprises that left road divides block part 123 alternatively.The schematic workflow of left road watermark built in items 120 ' is described below with reference to Fig. 8.

Fig. 8 illustrates the indicative flowchart of left road according to an embodiment of the invention watermark embed process.As shown in Figure 8, on optional (by empty wire frame representation) step S223(Ke Youzuo road, divide block part 123 to carry out) in the frame of watermark sequence to be embedded of ,Jiang Zuo road video be divided into the piece of pre-sizing.At step S224(Ke Youzuo road transform component 124, carry out) in carry out predetermined map in the frame of watermark sequence to be embedded of ,Zuo road video.Here, image coefficient is the conversion coefficient of described predetermined map.In the situation that carried out the piecemeal of step S223, in the piece of the embedded location of each piece in the frame of the watermark sequence to be embedded of step S224 Zhong Duizuo road video in carrying out predetermined map ,Zuo road video in being distributed in left road video.Concern judging part 121 execution on step 221(Ke Youzuo road) in, judge whether the image coefficient on each embedded location in the video of left road meets the corresponding left road of the value image coefficient restriction relation that will be embedded in the bit in the watermark sequence on described embedded location.In the situation that the image coefficient on embedded location does not meet corresponding left road image coefficient restriction relation, at step S222(Ke Youzuo road coefficient adjustment parts 122, carry out) in be adjusted at the image coefficient on described embedded location so that the image coefficient through adjusting meets corresponding Zuo road image coefficient restriction relation.Judge whether to have carried out for whole embedded locations the processing of step S221 and step S222 in step S225.If all embedded location has completed the watermark embed process end of ,Ze Zuo road.If still have embedded location untreated, process proceeds to step S221, proceeds the processing of step S221 and S222.

Fig. 9 illustrates the schematic block diagram of right wing watermark built in items according to another embodiment of the invention.In Fig. 9, except comprising right wing coefficient decision parts 131 and right wing coefficient adjustment parts 132 same as shown in Figure 6, right wing watermark built in items 130 ' also comprises right wing transform component 134, and (by empty wire frame representation) comprises that right wing divides block part 133 alternatively.The schematic workflow of right wing watermark built in items 130 ' is described below with reference to Figure 10.

Figure 10 illustrates the indicative flowchart of right wing watermark embed process according to an embodiment of the invention.As shown in figure 10, at optional (by empty wire frame representation) step S233(, can be divided by right wing block part 133 to carry out) in, the frame of the watermark sequence to be embedded of right wing video is divided into to the piece of pre-sizing.At step S234(, can be carried out by right wing transform component 134) in, carry out predetermined map in the frame of the watermark sequence to be embedded of right wing video.The predetermined map of carrying out in the frame of video of described predetermined map Yu Zuo road is identical.Here, image coefficient is the conversion coefficient of described predetermined map.In the situation that carried out the piecemeal of step S233, in step S234, each piece in the frame of the watermark sequence to be embedded of right wing video is carried out to predetermined map, the embedded location in the right wing video is distributed in the piece in the right wing video.Can be concerned by right wing judging part 131 execution at step 231() in, judge whether the image coefficient on each embedded location in the right wing video meets the corresponding right wing image coefficient of the value restriction relation that will be embedded in the bit in the watermark sequence on described embedded location.In the situation that the image coefficient on embedded location does not meet corresponding right wing image coefficient restriction relation, at step S232(, can be carried out by right wing coefficient adjustment parts 132) in be adjusted at the image coefficient on described embedded location so that the image coefficient through adjusting meets corresponding right wing image coefficient restriction relation.Judge whether to have carried out for whole embedded locations the processing of step S231 and step S232 in step S235.If all embedded location completes, the right wing watermark embed process finishes.If still have embedded location untreated, process proceeds to step S231, proceeds the processing of step S231 and S232.

As example rather than restriction, below describe on the corresponding embedded location of the different Zuo of several utilizations road image coefficient restriction relation and right wing image coefficient restriction relation Zuo road video and right wing video and embed corresponding watermark bit Zuo road and right wing watermark embed process example.Value as the bit of the watermark sequence of binary sequence for example can comprise { 0,1} or { 1,1}, that is, watermark sequence is by 0,1 binary sequence formed, or by-1,1 binary sequence formed.In following example, use bit value to comprise the { watermark sequence of 0,1}.For bit value comprise the watermark sequence of 1,1}, processing mode is identical, for example can easily the bit value 0 in following example be replaced with to-1, bit value 1 still remains 1, carries out thus bit value and comprises that { embedding of the watermark sequence of 1,1} is processed.In addition, in following example, in Yu Jiangzuo road and right wing video, the frame of watermark sequence to be embedded is divided into the piece of a plurality of 8 * 8 pixels, and uses discrete cosine transform (Discrete Cosine Transform, DCT) to be converted each piece.The embedded location of watermark bit can be distributed in one or more in frame, and in a piece, one or more embedded locations can be set.

In the first example, left road and right wing image coefficient restriction relation are all linear inequalities, limit respectively DCT coefficient on the embedded location in the video of left road and single order deviation and the DCT coefficient on the embedded location in the right wing video and the single order deviation of adjacent DCT coefficient of adjacent DCT coefficient.

In this example, the corresponding left road of the bit value 0 in watermark sequence image coefficient restriction relation is by following formula (1) expression, and the corresponding left road of bit value 1 image coefficient restriction relation is meaned by following formula (2):

(c i,8-i) left-mean((c i,8-i+1) left,(c i+1,8-i) left,(c i+1,8-i+1) left)≥a,?(1)

(c i,8-i) left-mean((c i,8-i+1) left,(c i+1,8-i) left,(c i+1,8-i+1) left)<b,?(2)

Wherein, (c i, 8-i) leftdCT coefficient on embedded location (i, 8-i) in a piece in the video of Shi Zuo road, (c i, 8-i+1) left, (c i+1,8-i) left, (c i+1,8-i+1) leftembedded location (i, 8-i) position (i, 8-i+1), (i+1,8-i), (i+1 on every side in described in the video of difference Shi Zuo road, DCT coefficient 8-i+1), the function that mean() expression is averaged, i=1 ..., 7, a and b are integers, and a>=b.

The bit value that for example will embed on an embedded location in the video of ,Dang Zuo road is 0, and the DCT coefficient on this embedded location is adjusted the value of the DCT coefficient on this embedded location while not meeting formula (1), makes it meet formula (1).The bit value that will embed on an embedded location in the video of Dang Zuo road is 0, and the DCT coefficient on this embedded location is while meeting formula (1), without the value of adjusting the DCT coefficient on this embedded location.The value of the DCT coefficient on this embedded location meets formula (1), is equivalent to embed a bit that value is 0 on this embedded location, rather than this bit is write direct on this embedded location.The bit value that in like manner will embed on an embedded location in the video of ,Dang Zuo road is 1, and the DCT coefficient on this embedded location is adjusted the value of the DCT coefficient on this embedded location while not meeting formula (2), makes it meet formula (2).The bit value that will embed on an embedded location in the video of Dang Zuo road is 1, and the DCT coefficient on this embedded location is while meeting formula (2), without the value of adjusting the DCT coefficient on this embedded location.

In this example, the corresponding right wing image coefficient of bit value 0 restriction relation in watermark sequence is by following formula (3) expression, and the corresponding right wing image coefficient of bit value 1 restriction relation is meaned by following formula (4):

(c i,8-i) right-mean((c i,8-i+1) right,(c i+1,8-i) right,(c i+1,8-i+1) right)≤m,?(3)

(c i,8-i) right-mean((c i,8-i+1) right,(c i+1,8-i) right,(c i+1,8-i+1) right)>n,?(4)

Wherein, (c i, 8-i) rightthe DCT coefficient on the embedded location (i, 8-i) in a piece in the right wing video, (c i, 8-i+1) right, (c i+1,8-i) right, (c i+1,8-i+1) rightbe respectively the DCT coefficient on position (i, 8-i+1) on every side of embedded location (i, 8-i) in described in the right wing video, (i+1,8-i), (i+1,8-i+1), m and n are integers, and n>=m.

For example, the bit value that will embed on an embedded location in the right wing video is 0, and the DCT coefficient on this embedded location is adjusted the value of the DCT coefficient on this embedded location while not meeting formula (3), makes it meet formula (3).The bit value that will embed on an embedded location in the right wing video is 0, and the DCT coefficient on this embedded location is while meeting formula (3), without the value of adjusting the DCT coefficient on this embedded location.The value of the DCT coefficient on this embedded location meets formula (3), is equivalent to embed a bit that value is 0 on this embedded location.In like manner can learn when the bit value that will embed on an embedded location in the right wing video is 1 the adjustment process to the DCT coefficient on this embedded location according to formula (4), repeat no more here.

For convenient, understand, Figure 11 a-11d illustrates the schematic diagram of image coefficient matrix according to an embodiment of the invention.As shown in Figure 11 a, the frame in left side is divided into a plurality of fritters, and each piece is through dct transform; The Matrix C on right side 8the DCT coefficient matrix of a piece after dct transform in frame on the left of being.For for purpose of brevity, not shown Matrix C 8in whole elements, but the Partial Elements relevant with this example only is shown.For example, whether 4 DCT coefficients that each square frame frame goes out can meet for the DCT coefficient of judging this right upper angle of square frame corresponding left road or right wing image coefficient restriction relation, in order to adjust the DCT coefficient at this right upper angle of square frame.

Owing between two two field pictures corresponding in left road video and right wing video, only having parallax, therefore, for the embedded location of the embedded location in the video of left road and the correspondence in the right wing video, can think approx mean ((c i, 8-i+1) left, (c i+1,8-i) left, (c i+1,8-i+1) left)=mean ((c i, 8-i+1) right, (c i+1,8-i) right, (c i+1,8-i+1) right), thus, can derive from above formula (1) and (3) following formula (5), from above formula (2) and the formula (6) below (4) derivation:

(c i,8-i) left-(c i,8-i) right≥a-m,?????(5)

(c i,8-i) left-(c i,8-i) right<b-n。(6)

Wherein formula (5) can be for the watermark bit that is 0 from three-dimensional video-frequency extraction value when the watermark extracting, and formula (6) can be for the watermark bit that is 1 from three-dimensional video-frequency extraction value when the watermark extracting.

Should be appreciated that the above formula provided (1) to (6) only for the purpose of illustration, rather than restriction.For example, in above-mentioned formula (1) to (4), be the embedded location (i in the predetermined block in the video of left/right road, DCT coefficient value 8-i) deducts the position (i around it, 8-i+1), (i+1,8-i), (i+1, the mean value of the DCT coefficient 8-i+1), and think that this mean value of left and right road video equates, and derive formula (5), (6); Those skilled in the art should understand, above-mentioned formula (1) to (4) also can adopt the embedded location (i in the predetermined block in the video of left/right road, DCT coefficient value 8-i) deducts for example (i of its any position on every side, DCT coefficient value 8-i+1), because difference between two two field pictures corresponding in three-dimensional video-frequency Zuo road video and right wing video is little, (c i, 8-i+1) left(c i, 8-i+1) rightalso can think approximately equal.Again for example, in the 8 * 8DCT coefficient matrix shown in Figure 11 a, the coefficient that adjust and adjacent coefficient thereof are positioned on the left clinodiagonal of matrix, and embedded location is positioned on the left clinodiagonal of matrix.Yet embedded location (will adjust this locational coefficient) can be arranged in any position of DCT matrix, and not restriction.For example, the coefficient that adjust and adjacent coefficient thereof also can be positioned on the right clinodiagonal of matrix (as shown in Figure 11 b), on horizontal central line (as shown in Figure 11 c), and vertically on center line (as shown in Figure 11 d), or any one or more precalculated positions in matrix.The coefficient of adjusting also is not limited only to the upper right corner of the square frame shown in each figure, and can be the arbitrary coefficient in square frame.

In addition, for the coefficient that will adjust, the scope of its adjacent coefficient is not restriction also.For example, in Figure 11 c, the coefficient in square frame in the middle of left side is the coefficient that will adjust, has chosen 5 coefficients around the coefficient that will adjust as adjacent coefficient.

Coefficient in above formula (1) to (6) can be made corresponding modify according to the position at coefficient place.For example, for the situation shown in Figure 11 c, following formula (1 ') and (2 ') can be revised as respectively in formula (1) and (2):

(c 5,i) left-mean((c 4,i) left,(c 4,i+1) left,(c 5,i+1) left,(c 6,i) left,(c 6,i+1) left)≥a,(1’)

(c 5,i) left-mean((c 4,i) left,(c 4,i+1) left,(c 5,i+1) left,(c 6,i) left,(c 6,i+1) left)<b。(2’)

Following formula (3 ') and (4 ') can be revised as respectively in formula (3) and (4):

(c 5,i) right-mean((c 4,i) right,(c 4,i+1) right,(c 5,i+1) right,(c 6,i) right,(c 6,i+1) right)≤m,??(3’)

(c 5,i) right-mean((c 4,i) right,(c 4,i+1) right,(c 5,i+1) right,(c 6,i) right,(c 6,i+1) right)>n。(4’)

Following formula (5 ') and (6 ') can be revised as respectively in formula (5) and (6):

(c 5,i) left-(c 5,i) right≥a-m,????(5’)

(c 5,i) left-(c 5,i) right<b-n。(6’)

In like manner can obtain the corresponding image coefficient restriction relation of the situation shown in other figure, not describe one by one here.

In the second example, left road and right wing image coefficient restriction relation are all nonlinear inequalities, limit respectively DCT coefficient on the embedded location in the video of left road and second order deviation and the DCT coefficient on the embedded location in the right wing video and the second order deviation of adjacent DCT coefficient of adjacent DCT coefficient.The corresponding left road of bit value 0 in watermark sequence image coefficient restriction relation is by following formula (7) expression, and the corresponding left road of bit value 1 image coefficient restriction relation is meaned by following formula (8):

( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c 8 - i , i ) left 2 + ( c 8 - i , i + 1 ) left 2 + ( c 8 - i + 1 , i ) left 2 + ( c 8 - i + 1 , i + 1 ) left 2 ) &GreaterEqual; a , - - - ( 7 )

( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c 8 - i , i ) left 2 + ( c 8 - i , i + 1 ) left 2 + ( c 8 - i + 1 , i ) left 2 + ( c 8 - i + 1 , i + 1 ) left 2 ) < b , - - - ( 8 )

Wherein, (c i, 8-i) left, (c i, 8-i+1) left, (c i+1,8-i) left, (c i+1,8-i+1) leftdCT coefficient on embedded location ((i, 8-i), (i, 8-i+1), (i+1,8-i), (i+1,8-i+1)) (2 * 2 coefficient block) in a piece in the video of difference Shi Zuo road, (c 8-i, i) left, (c 8-i, i+1) left, (c 8-i+1, i) left, (c 8-i+1, i+1) leftrespectively the DCT coefficient on the position ((8-i, i), (8-i, i+1), (8-i+1, i), (8-i+1, i+1)) (2 * 2 coefficient block) in described in the video of described left road, i=1 ..., 8, a and b are integers, and a>=b. ( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) With ( ( c 8 - i , i ) left 2 + ( c 8 - i , i + 1 ) left 2 + ( c 8 - i + 1 , i ) left 2 + ( c 8 - i + 1 , i + 1 ) left 2 ) The energy that means respectively above two 2 * 2 coefficient block.

The bit value that for example will embed on an embedded location in the video of ,Dang Zuo road (2 * 2 coefficient block) is 0, and the DCT coefficient on this embedded location is adjusted DCT the coefficient ((c on this embedded location while not meeting formula (7) i, 8-i) left, (c i, 8-i+1) left, (c i+1,8-i) left, (c i+1,8-i+1) leftin one or more) value, make it meet formula (7).The bit value that will embed on an embedded location in the video of Dang Zuo road is 0, and the DCT coefficient on this embedded location is while meeting formula (7), without the value of adjusting the DCT coefficient on this embedded location.The value of the DCT coefficient on this embedded location meets formula (7), is equivalent to embed a bit that value is 0 on this embedded location.In like manner, can learn the bit value that will embed on an embedded location in the video of Dang Zuo road be 1 o'clock according to formula (8) adjustment process to the DCT coefficient on this embedded location, repeat no more here.

In this example, the corresponding right wing image coefficient of bit value 0 restriction relation in watermark sequence is by following formula (9) expression, and the corresponding right wing image coefficient of bit value 1 restriction relation is meaned by following formula (10):

( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) - ( ( c 8 - i , i ) right 2 + ( c 8 - i , i + 1 ) right 2 + ( c 8 - i + 1 , i ) right 2 + ( c 8 - i + 1 , i + 1 ) right 2 ) &le; m , - - - ( 9 )

( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) - ( ( c 8 - i , i ) right 2 + ( c 8 - i , i + 1 ) right 2 + ( c 8 - i + 1 , i ) right 2 + ( c 8 - i + 1 , i + 1 ) right 2 ) > n , - - - ( 10 )

Wherein, (c i, 8-i) right, (c i, 8-i+1) right, (c i+1,8-i) right, (c i+1,8-i+1) rightrespectively the discrete cosine transform coefficient on the embedded location ((i, 8-i), (i, 8-i+1), (i+1,8-i), (i+1,8-i+1)) (2 * 2 coefficient block) in a piece in the right wing video, (c 8-i, i) right, (c 8-i, i+1) right, (c 8-i+1, i) right, (c 8-i+1, i+1) rightbe respectively the discrete cosine transform coefficient on the position ((8-i, i), (8-i, i+1), (8-i+1, i), (8-i+1, i+1)) (2 * 2 coefficient block) in described in the right wing video, m and n are integers, and n>=m. ( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) With ( ( c 8 - i , i ) right 2 + ( c 8 - i , i + 1 ) right 2 + ( c 8 - i + 1 , i ) right 2 + ( c 8 - i + 1 , i + 1 ) right 2 ) The energy that means respectively above two 2 * 2 coefficient block.

Similar to the coefficient adjustment process in the adjustment process Zuo road video of the DCT coefficient on this embedded location according to formula (9) or (10) when the bit value that will embed on an embedded location in the right wing video is 0 or 1, repeat no more here.

For convenient, understand, Figure 12 a-12c illustrates the schematic diagram of image coefficient matrix according to another embodiment of the invention.Image coefficient matrix in Figure 12 a-12c can be the corresponding image coefficient matrix of a piece in the frame of video in Figure 11 a.For for purpose of brevity, not shown Matrix C 8in whole elements, but the Partial Elements relevant with this example only is shown.For example, in Figure 12 a, left side 2 * 2 coefficient block that gone out by the square frame frame are embedded locations, can judge with the difference of the energy of right side 2 * 2 coefficient block that gone out by the square frame frame whether the coefficient in this left side block meets corresponding left road image coefficient restriction relation according to the energy of this piece, in order to adjust the coefficient in this left side block.

Similarly, for the embedded location of the embedded location in the video of left road and the correspondence in the right wing video, can think approx ( ( c 8 - i , i ) left 2 + ( c 8 - i , i + 1 ) left 2 + ( c 8 - i + 1 , i ) left 2 + ( c 8 - i + 1 , i + 1 ) left 2 ) = ( ( c 8 - i , i ) right 2 + ( c 8 - i , i + 1 ) right 2 + ( c 8 - i + 1 , i ) right 2 + ( c 8 - i + 1 , i + 1 ) right 2 ) , thus, can derive from above formula (7) and (9) following formula (11), from above formula (8) and the formula (12) below (10) derivation:

( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) )

&GreaterEqual; a - m , - - - ( 11 )

( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) )

< b - n , - - - ( 12 )

Wherein formula (11) can be for the watermark bit that is 0 from three-dimensional video-frequency extraction value when the watermark extracting, and formula (12) can be for the watermark bit that is 1 from three-dimensional video-frequency extraction value when the watermark extracting.

Should be appreciated that the above formula provided (7) to (12) only for the purpose of illustration, rather than restriction.In the 8 * 8DCT coefficient matrix shown in Figure 12 a, embedded location is positioned on the left clinodiagonal of matrix.Yet embedded location can be arranged in any position of DCT matrix, and not restriction.For example, the coefficient that adjust and adjacent coefficient thereof also can be positioned on the right clinodiagonal of matrix (as shown in Figure 12 b), on horizontal central line (as shown in Figure 12 c), and vertically (not shown) on center line, or any one or more precalculated positions in matrix.

In addition, embedded location also can be corresponding to the coefficient block of other sizes.The size of the coefficient block that for example, the embedded location shown in Figure 12 c is corresponding is 3 * 2 coefficient block.

In addition, for the coefficient that will adjust, the scope of its adjacent coefficient is not restriction also.For example, in Figure 12 c, the adjacent coefficient of the coefficient that adjust (being arranged in a coefficient block) is arranged in 2 adjacent coefficient block.

Coefficient in above formula (7) to (12) can be made corresponding modify according to the position at coefficient place.For example, for the situation shown in Figure 12 c, formula (7) to (12) can be revised as respectively following formula (7 ') to (12 '):

( ( c 4 , i ) left 2 + ( c 4 , i + 1 ) left 2 + ( c 5 , i ) left 2 + ( c 5 , i + 1 ) left 2 + ( c 6 , i ) left 2 + ( c 6 , i + 1 ) left 2 ) - mean ( ( ( c 4 , i - 2 ) left 2 + ( c 4 , i - 1 ) left 2

+ ( c 5 , i - 2 ) left 2 + ( c 5 , i - 1 ) left 2 + ( c 6 , i - 2 ) left 2 + ( c 6 , i - 1 ) left 2 ) + ( ( c 4 , i + 2 ) left 2 + ( c 4 , i + 3 ) left 2 + ( c 5 , i + 2 ) left 2 + ( c 5 , i + 3 ) left 2 + ( c 6 , i + 2 ) left 2

+ ( c 6 , i + 3 ) left 2 ) ) &GreaterEqual; a , - - - ( 7 , )

( ( c 4 , i ) left 2 + ( c 4 , i + 1 ) left 2 + ( c 5 , i ) left 2 + ( c 5 , i + 1 ) left 2 + ( c 6 , i ) left 2 + ( c 6 , i + 1 ) left 2 ) - mean ( ( ( c 4 , i - 2 ) left 2 + ( c 4 , i - 1 ) left 2 + ( c 5 , i - 2 ) left 2

+ ( c 5 , i - 1 ) left 2 + ( c 6 , i - 2 ) left 2 + ( c 6 , i - 1 ) left 2 ) ( ( c 4 , i + 2 ) left 2 + ( c 4 , i + 3 ) left 2 + ( c 5 , i + 2 ) left 2 + ( c 5 , i + 3 ) left 2 + ( c 6 , i + 2 ) left 2 + ( c 6 , i + 3 ) left 2 ) ) < b , - - - ( 8 , )

( ( c 4 , i ) right 2 + ( c 4 , i + 1 ) right 2 + ( c 5 , i ) right 2 + ( c 5 , i + 1 ) right 2 + ( c 6 , i ) right 2 + ( c 6 , i + 1 ) right 2 ) - mean ( ( ( c 4 , i - 2 ) right 2 + ( c 4 , i - 1 ) right 2 + ( c 5 , i - 2 ) right 2

+ ( c 5 , i - 1 ) right 2 + ( c 6 , i - 2 ) right 2 + ( c 6 , i - 1 ) right 2 ) + ( ( c 4 , i + 2 ) right 2 + ( c 4 , i + 3 ) right 2 + ( c 5 , i + 2 ) right 2 + ( c 5 , i + 3 ) right 2 + ( c 6 , i + 2 ) right 2

+ ( c 6 , i + 3 ) right 2 ) ) &le; m

, - - - ( 9 , )

( ( c 4 , i ) right 2 + ( c 4 , i + 1 ) right 2 + ( c 5 , i ) right 2 + ( c 5 , i + 1 ) right 2 + ( c 6 , i ) right 2 + ( c 6 , i + 1 ) right 2 ) - mean ( ( ( c 4 , i - 2 ) right 2 + ( c 4 , i - 1 ) right 2

+ ( c 5 , i - 2 ) right 2 + ( c 5 , i - 1 ) right 2 + ( c 6 , i - 2 ) right 2 + ( c 6 , i - 1 ) right 2 ) + ( ( c 4 , i + 2 ) right 2 + ( c 4 , i + 3 ) right 2 + ( c 5 , i + 2 ) right 2 + ( c 5 , i + 3 ) right 2 + ( c 6 , i + 2 ) right 2

+ ( c 6 , i + 3 ) right 2 ) ) > n - - - ( 10 , )

( c 4 , i ) left 2 + ( c 4 , i + 1 ) left 2 + ( c 5 , i ) left 2 + ( c 5 , i + 1 ) left 2 + ( c 6 , i ) left 2 + ( c 6 , i + 1 ) left 2 ) , - - - ( 11 , )

- ( ( c 4 , i ) right 2 + ( c 4 , i + 1 ) right 2 + ( c 5 , i ) right 2 + ( c 5 , i + 1 ) right 2 + ( c 6 , i ) right 2 + ( c 6 , i + 1 ) right 2 ) &GreaterEqual; a - m

( c 4 , i ) left 2 + ( c 4 , i + 1 ) left 2 + ( c 5 , i ) left 2 + ( c 5 , i + 1 ) left 2 + ( c 6 , i ) left 2 + ( c 6 , i + 1 ) left 2 ) . - - - ( 12 , )

- ( ( c 4 , i ) right 2 + ( c 4 , i + 1 ) right 2 + ( c 5 , i ) right 2 + ( c 5 , i + 1 ) right 2 + ( c 6 , i ) right 2 + ( c 6 , i + 1 ) right 2 ) < b - n

In like manner can obtain the corresponding image coefficient restriction relation of the situation shown in other figure, not describe one by one here.

In above the first example and the second example, the same bit value of watermark sequence corresponding left road image coefficient restriction relation and the right wing image coefficient restriction relation adjustment opposite direction to image coefficient.Particularly, inequality (formula (1) and (3) of bit value 0 corresponding left road image coefficient restriction relation and right wing image coefficient restriction relation, and formula (7) and (9)) symbol contrary, the symbol of the inequality of bit value 1 corresponding left road image coefficient restriction relation and right wing image coefficient restriction relation (formula (2) and (4), and formula (8) and (10)) is contrary.Like this, adjustment opposite direction due to Zuo road video and the right wing video image coefficient corresponding to same bit value, relation while making three-dimensional video-frequency be attacked between the coefficient of two-path video is highly stable, when according to left road image coefficient restriction relation and the definite stereo-picture coefficient restriction relation of right wing image coefficient restriction relation, extracting watermark, can farthest guarantee the robustness of watermark extracting.

In following the 3rd example and the 4th example, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation are identical but adjusting range is different to the adjustment direction of image coefficient.

On the 3rd example Zhong,Zuo road and right wing image coefficient, are all linear inequalities.Identical with in the first example of bit value 0 in watermark sequence and 1 corresponding left road image coefficient restriction relation, used respectively formula (1) and (2) to mean.The corresponding right wing image coefficient of bit value 0 restriction relation in watermark sequence is by following formula (13) expression, and the corresponding right wing image coefficient of bit value 1 restriction relation is meaned by following formula (14):

(c i,8-i) right-mean((c i,8-i+1) right,(c i+1,8-i) right,(c i+1,8i+1) right)≥m,??(13)

(c i,8-i) right-mean((c i,8-i+1) right,(c i+1,8-i) right,(c i+1,8-i+1) right)<n,?(14)

Wherein, (c i, 8-i) rightthe DCT coefficient on the embedded location (i, 8-i) in a piece in the right wing video, (c i, 8-i+1) right, (c i+1,8-i) right, (c i+1,8-i+1) rightbe respectively the DCT coefficient on position (i, 8-i+1) on every side of embedded location (i, 8-i) in described in the right wing video, (i+1,8-i), (i+1,8-i+1), m and n are integers, m>=n.Different from the coefficient adjustment amplitude of right wing video for Shi Zuo road video, in this example, m, n in a, b in formula (1), (2) and formula (13), (14) are carried out to following restriction: a ≠ m, b ≠ n.

Identical with the first example with the process that embeds the bit in corresponding watermark sequence on embedded location in the video of (2) Zuo road according to formula (1).Adjust image coefficient on the embedded location in the right wing video according to formula (13) and (14) similar with process and the coefficient adjustment process in the right wing video in the first example that embeds bit 0 in corresponding watermark sequence or 1, repeat no more here.

Similarly, the formula (5) more than also can deriving from formula (1) and (13), take for the watermark bit that is 0 from three-dimensional video-frequency extraction value when the watermark extracting; Formula (6) more than deriving from formula (2) and (14), take for the watermark bit that is 1 from three-dimensional video-frequency extraction value when the watermark extracting.Yet, in the 3rd example, when using formula (5) and (6), need to meet a-m >=b-n(in the first example due to the relation between each formula, this condition is met naturally) so that do not have overlapping between bit value 0 corresponding stereo-picture coefficient restriction relation and the corresponding stereo-picture coefficient of bit value 1 restriction relation.

On the 4th example Zhong,Zuo road and right wing image coefficient, are all nonlinear inequalities.Identical with in the second example of bit value 0 in watermark sequence and 1 corresponding left road image coefficient restriction relation, used respectively formula (7) and (8) to mean.The corresponding right wing image coefficient of bit value 0 restriction relation in watermark sequence is by following formula (15) expression, and the corresponding right wing image coefficient of bit value 1 restriction relation is meaned by following formula (16):

( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) - ( ( c 8 - i , i ) right 2 + ( c 8 - i , i + 1 ) right 2 + ( c 8 - i + 1 , i ) right 2 + ( c 8 - i + 1 , i + 1 ) right 2 ) &GreaterEqual; m , - - - ( 15 )

( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) - ( ( c 8 - i , i ) right 2 + ( c 8 - i , i + 1 ) right 2 + ( c 8 - i + 1 , i ) right 2 + ( c 8 - i + 1 , i + 1 ) right 2 ) < n , - - - ( 16 )

Wherein, (c i, 8-i) right, (c i, 8-i+1) right, (c i+1,8-i) right, (c i+1,8-i+1) rightrespectively the DCT coefficient on the embedded location ((i, 8-i), (i, 8-i+1), (i+1,8-i), (i+1,8-i+1)) in a piece in the right wing video, (c 8-i, i) right, (c 8-i, i+1) right, (c 8-i+1, i) right, (c 8-i+1, i+1) rightbe respectively the DCT coefficient on the position ((8-i, i), (8-i, i+1), (8-i+1, i), (8-i+1, i+1)) in described in the right wing video, m and n are integers, m>=n.Different from the coefficient adjustment amplitude of right wing video for Shi Zuo road video, in this example, m, n in a, b in formula (7), (8) and formula (15), (16) are carried out to following restriction: a ≠ m, b ≠ n.

Identical with the second example with the process that embeds the bit in corresponding watermark sequence on embedded location in the video of (8) Zuo road according to formula (7).Adjust image coefficient on the embedded location in the right wing video according to formula (15) and (16) similar with process and the coefficient adjustment process in the right wing video in the second example that embeds bit 0 in corresponding watermark sequence or 1, repeat no more here.

Similarly, the formula (11) more than also can deriving from formula (7) and (15), take for the watermark bit that is 0 from three-dimensional video-frequency extraction value when the watermark extracting; Formula (12) more than deriving from formula (8) and (16), take for the watermark bit that is 1 from three-dimensional video-frequency extraction value when the watermark extracting.Yet, in the 4th example, when using formula (11) and (12), need to meet a-m >=b-n(in the second example due to the relation between each formula, this condition is met naturally) so that do not have overlapping between bit value 0 corresponding stereo-picture coefficient restriction relation and the corresponding stereo-picture coefficient of bit value 1 restriction relation.

In the third and fourth above example, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation are identical but adjusting range is different to the adjustment direction of image coefficient.Particularly, the symbol of the inequality of bit value 0 corresponding left road image coefficient restriction relation and right wing image coefficient restriction relation (formula (1) and (13), and formula (7) and (15)) is identical, but a ≠ m; The symbol of the inequality of bit value 1 corresponding left road image coefficient restriction relation and right wing image coefficient restriction relation (formula (2) and (14), and formula (8) and (16)) is identical, but b ≠ n.Like this, although the adjustment direction of the image coefficient that Zuo road video is corresponding to same bit value with the right wing video is identical, but adjusting range difference, therefore when according to left road image coefficient restriction relation and the definite stereo-picture coefficient restriction relation of right wing image coefficient restriction relation, extracting watermark, also can guarantee to a certain extent the robustness of watermark extracting.

In above-mentioned example, (formula (1), (2), (7) (8) are only for limiting the image coefficient of left road frame of video for left road image coefficient restriction relation, (formula (3), (4), (9) (10) are only for limiting the image coefficient of right wing frame of video for right wing image coefficient restriction relation, therefore, can extract watermark sequence from the video of independent Zuo road according to left road image coefficient restriction relation, and extract watermark sequence according to right wing image coefficient restriction relation from independent right wing video.

Although in above example, the frame of watermark sequence to be embedded is divided into to 8 * 8 block of pixels,, should be appreciated that piecemeal is not essential.Those skilled in the art can expect the various possible image coefficient restriction relation not to the frame of video piecemeal time according to principle of the present disclosure.For example, can directly apply or suitably revise the image coefficient restriction relation shown in above formula according to the predetermined neighbouring relations of image coefficient.In addition, the size of piece is not restriction also, for example, the frame of watermark sequence to be embedded can be divided into to other big or small pieces, and according to the described image coefficient restriction relation of the above formula of big or small corresponding modify of piece, in this explanation for example that differs.

In addition, in above example, be to take dct transform to describe as example.But, should be appreciated that and can use any other suitable transform method, as wavelet transformation etc.Those skilled in the art are according to above open, the various possible image coefficient restriction relation in the time of can expecting applying other transform methods.For example, when using other transform methods, the dct transform coefficient in above formula directly can be replaced with to the conversion coefficient of other transform methods, to obtain suitable image coefficient restriction relation.

In addition, more than having provided image coefficient is conversion coefficient Shi Zuo road and right wing and the stereo-picture coefficient restriction relation example in transform domain.But, should be appreciated that conversion is not essential.Those skilled in the art, according to above open, can expect various possible image coefficient restriction relation when image coefficient is the pixel value in spatial domain.For example, can replace the dct transform coefficient in above formula with the pixel value on correspondence position, obtain the image coefficient restriction relation in spatial domain.

<2. for extract the apparatus and method of watermark from three-dimensional video-frequency >

The process of extracting watermark from three-dimensional video-frequency is to add the inverse process of the process of watermark with stereoscopic video basically.In the watermarking method according to the embodiment of the present invention and device, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation is different from right wing image coefficient restriction relation, and the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation can be determined all relevant image coefficient restriction relations of image coefficient on the corresponding embedded location of image coefficient on the embedded location of the described same bit value in the video of Yu Zuo road and the described same bit value in the right wing video.Therefore, when from three-dimensional video-frequency, extracting watermark, both can be according in the video of the independent Cong Zuo of left road image coefficient restriction relation road, having extracted watermark sequence or extracted watermark sequence separately from the right wing video according to right wing image coefficient restriction relation, also can be according to according to corresponding left road image coefficient restriction relation and the definite image coefficient restriction relation of right wing image coefficient restriction relation, as stereo-picture coefficient restriction relation, extracting watermark sequence from three-dimensional video-frequency.

Figure 13 illustrates according to an embodiment of the invention for extract the schematic block diagram of the device of watermark from three-dimensional video-frequency.For brevity, only show the closely-related part with the present invention in figure.As shown in figure 13, comprise that for the device (also claiming watermark extraction apparatus) 1300 that extracts watermark from three-dimensional video-frequency three-dimensional video-frequency Knock-Down Component 1310, watermark bit are extracted parts 1320 and watermark sequence forms parts 1330.The exemplary operation flow process of watermark extraction apparatus 1300 is described below with reference to Figure 14.

Figure 14 illustrates according to an embodiment of the invention for extract the indicative flowchart of the method for watermark from three-dimensional video-frequency.As shown in figure 14, at step S1410(, can be carried out by three-dimensional video-frequency Knock-Down Component 1310) in, left road video and right wing video obtained from three-dimensional video-frequency.Extract the watermark process for three-dimensional video-frequency form and watermarked process for the form of three-dimensional video-frequency identical.

Can be extracted by watermark bit parts 1320 at step S1420(carries out) the stereo-picture coefficient restriction relation satisfied with the image coefficient on corresponding embedded location in the right wing video by the image coefficient on each the predetermined embedded location in judging left road video, extract the corresponding value of stereo-picture coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the video of Zuo road and the corresponding embedded location in the right wing video.

Here, stereo-picture coefficient restriction relation is determined according to corresponding Zuo road image coefficient restriction relation and right wing image coefficient restriction relation.Left road image coefficient restriction relation is for the image coefficient on the embedded location of the value corresponding with stereo-picture coefficient restriction relation that is adjusted at left road video, watermarked in the video of Yi Zuo road.Right wing image coefficient restriction relation is for the image coefficient on the embedded location of the value corresponding with stereo-picture coefficient restriction relation that is adjusted at the right wing video, with watermarked in the right wing video.The corresponding left road of the same bit value of watermark sequence image coefficient restriction relation is different from right wing image coefficient restriction relation.About the more details of stereo-picture coefficient restriction relation ,Zuo road image coefficient restriction relation and right wing image coefficient restriction relation, can with reference to<1. for stereoscopic video, add the apparatus and method of watermark the description of part.

Equally, in watermark sequence, the embedded location of each bit in frame of video is predetermined.This can carry out by external component, or, also can be by providing embedded location determining means (not shown) to carry out in watermark extraction apparatus according to an embodiment of the invention.For example, can three-dimensional video-frequency Knock-Down Component 1310 and watermark bit in watermark extraction apparatus 1300 extract the insertion embedded location determining means identical with the embedded location determining means 150 shown in Fig. 3 between parts 1320.

Can be formed by watermark sequence parts 1330 at step S1430(carries out) in, watermark sequence formed according to the value of the bit of extracted watermark sequence.For example, can the bit value extracted on each embedded location be formed to watermark sequence according to predetermined order.

When the watermark sequence according to extracting is confirmed the copyright of three-dimensional video-frequency, the watermark sequence that extracts can be compared with the watermark sequence (also claiming the target watermark sequence) of the known copyright that can indicate three-dimensional video-frequency, according to the correlation size of the two, determine whether the watermark sequence extracted is the target watermark sequence.For example, if the correlation of the watermark sequence extracted and target watermark sequence is greater than predetermined threshold, show that extracted watermark sequence is the target watermark sequence, vice versa.

Figure 17 illustrates the schematic block diagram for the device from three-dimensional video-frequency extraction watermark according to another embodiment of the invention.In this embodiment, watermark extraction apparatus 1700 comprises watermark sequence verification component 1360, for the correlation of the watermark sequence according to extracted and target watermark sequence, determines whether extracted watermark sequence is the target watermark sequence.

When image coefficient is the pixel value in spatial domain, while extracting watermark, do not need known watermarked frame of video is converted.When image coefficient is the conversion coefficient in transform domain, need to first known watermarked frame of video be transformed in transform domain.As example rather than restriction, following Description Image coefficient is the example of the watermark extraction apparatus in the conversion coefficient situation in transform domain.

Figure 15 illustrates the schematic block diagram for the device from three-dimensional video-frequency extraction watermark according to another embodiment of the invention.With the watermark extraction apparatus 1300 of Figure 13, compare, the watermark extraction apparatus in Figure 15 also comprises left road video pre-filtering parts 1340 and right wing video pre-filtering parts 1350.Wherein ,Zuo road video pre-filtering parts 1340 comprise that left road transform component 1342 and optional left road divide block part 1341(by empty wire frame representation).Left road divides block part 1341 He Zuo road transform components 1342 to divide block part 123He Zuo road transform component 124 identical with Fig. 7 Zhong Zuo road.Right wing video pre-filtering parts 1350 comprise that right wing transform component 1352 and optional right wing divide block part 1351(by empty wire frame representation).Right wing divides block part 1351 and right wing transform component 1352 with the right wing in Fig. 9, to divide block part 133 identical with right wing transform component 134.Three-dimensional video-frequency Knock-Down Component 1310 in Figure 15, watermark bit extract parts 1320 and watermark sequence and form identical with in Figure 13 of parts 1330.The exemplary operation flow process of watermark extraction apparatus 1500 is described below with reference to Figure 16.

Figure 16 illustrates the indicative flowchart for the method from three-dimensional video-frequency extraction watermark according to another embodiment of the invention.With the flow process shown in Figure 14, compare, in Figure 16 after three-dimensional video-frequency being decomposed into to left road video and right wing video (step S1410), respectively at optional step S1441(by empty wire frame representation) (Ke Youzuo divides on road block part 1341 to carry out) predetermined frame (the known frame that embeds watermark sequence) of Zhong Jiangzuo road video is divided into the piece of pre-sizing, at step S1442(Ke Youzuo road transform component 1342, carry out) in carry out predetermined map in described predetermined frame; And at optional step S1451(by empty wire frame representation) predetermined frame of right wing video is divided in (can be divided by right wing block part 1351 to carry out) to the piece of pre-sizing, at step S1452(, can be carried out by right wing transform component 1352) in carry out the identical predetermined map of Yu Zuo road video in described predetermined frame.Piece in left road video and right wing video has identical pre-sizing.Here, image coefficient is the conversion coefficient of predetermined map.In the situation of the piecemeal of having carried out step S1441 and S1451, each piece in the video of step S1442 Zhong Duizuo road carries out predetermined map, embedded location in the video of left road is distributed in the piece in the video of left road, in step S1452, each piece in the right wing video is carried out to predetermined map, the embedded location in the right wing video is distributed in the piece in the right wing video.Then, perform step successively S1420 and step S1430.

As example rather than restriction, below describe several from utilize above<1. for stereoscopic video, add the apparatus and method of watermark first to fourth example of part extracts the example of embedded watermark sequence in describing the three-dimensional video-frequency of Zuo road and the watermarked sequence of right wing watermark embed process.

In the 5th example, from through the three-dimensional video-frequency of above the first example process, extracting watermark sequence.For example, for the embedded location (i in a piece in the known frame that embeds watermark sequence in the video of left road, 8-i) and the same embedded location in the known corresponding frame that has embedded watermark sequence in the right wing video, judge the image coefficient (c on this embedded location in the video of left road i, 8-i) leftwith the image coefficient (c on this embedded location in the right wing video i, 8-i) rightwhether meet possible stereo-picture coefficient restriction relation, i.e. formula (5) or (6).If meet formula (5), determine that the value of the bit of the watermark sequence embedded is 0 on this embedded location of three-dimensional video-frequency; If meet formula (6), determine that the value of the bit of the watermark sequence embedded is 1 on this embedded location of three-dimensional video-frequency.If formula (5) and (6) do not meet, can select a-m and b-n middle distance (c i, 8-i) left(c i, 8-i) righta nearer bit value that conduct embeds on this embedded location of difference.

Preferably, in a modification of the 5th example, can be by formula (5) be revised a little, obtain the stereo-picture coefficient restriction relation of formula (17) as bit value 0 correspondence, and formula (6) is revised a little, obtain the stereo-picture coefficient restriction relation of formula (18) as bit value 1 correspondence, as follows:

( c i , 8 - i ) left - ( c i , 8 - i ) right &GreaterEqual; a - m - a - m - ( b - n ) 2 , - - - ( 17 )

( c i , 8 - i ) left - ( c i , 8 - i ) right < b - n + a - m - ( b - n ) 2 . - - - ( 18 )

In the 6th example, from through the three-dimensional video-frequency of above the second example process, extracting watermark sequence.For example, for the embedded location ((i in a piece in the known frame that embeds watermark sequence in the video of left road, 8-i), (i, 8-i+1), (i+1,8-i), (i+1,8-i+1)) known in (one 2 * 2) and right wing video embedded the same embedded location in the corresponding frame of watermark sequence, judges the image coefficient (c on this embedded location in the video of left road i, 8-i) left, (c i, 8-i+1) left, (c i+1,8-i) left, (c i+1,8-i+1) leftwith the image coefficient (c on this embedded location in the right wing video 8-i, i) right, (c 8-i, i+1) right, (c 8-i+1, i) right, (c 8-i+1, i+1) rightwhether meet possible stereo-picture coefficient restriction relation, i.e. formula (11) or (12).If meet formula (11), determine that the value of the bit of the watermark sequence embedded is 0 on this embedded location of three-dimensional video-frequency; If meet formula (12), determine that the value of the bit of the watermark sequence embedded is 1 on this embedded location of three-dimensional video-frequency.If formula (11) and (12) do not meet, can select a-m and b-n middle distance ( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) ) A nearer bit value that conduct embeds on this embedded location.

Preferably, in a modification of the 6th example, can be by formula (11) be revised a little, obtain the stereo-picture coefficient restriction relation of formula (19) as bit value 0 correspondence, and formula (12) is revised a little, obtain the stereo-picture coefficient restriction relation of formula (20) as bit value 1 correspondence, as follows:

( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) )

&GreaterEqual; a - m - a - m - ( b - n ) 2 - - - ( 19 )

( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) )

< b - n + a - m - ( b - n ) 2 . - - - ( 20 )

In the 7th example, from through the three-dimensional video-frequency of above the 3rd example process, extracting watermark sequence.For example, for the embedded location (i in a piece in the known frame that embeds watermark sequence in the video of left road, 8-i) and the same embedded location in the known corresponding frame that has embedded watermark sequence in the right wing video, judge the image coefficient (c on this embedded location in the video of left road i, 8-i) leftwith the image coefficient (c on this embedded location in the right wing video i, 8-i) rightwhether meet possible stereo-picture coefficient restriction relation, i.e. formula (5) or (6), a ≠ m wherein, b ≠ n, and a-m>=b-n.If meet formula (5), determine that the value of the bit of the watermark sequence embedded is 0 on this embedded location of three-dimensional video-frequency; If meet formula (6), determine that the value of the bit of the watermark sequence embedded is 1 on this embedded location of three-dimensional video-frequency.If formula (5) and (6) do not meet, can select a-m and b-n middle distance (c i, 8-i) left(c i, 8-i) righta nearer bit value that conduct embeds on this embedded location of difference.Preferably, in a modification of the 7th example, can use the stereo-picture coefficient restriction relation of formula (17) as bit value 0 correspondence, and use formula (18) is as the stereo-picture coefficient restriction relation of bit value 1 correspondence, a ≠ m wherein, b ≠ n, and a-m >=b-n.

In the 8th example, from through the three-dimensional video-frequency of above the 4th example process, extracting watermark sequence.For example, for the embedded location ((i in a piece in the known frame that embeds watermark sequence in the video of left road, 8-i), (i, 8-i+1), (i+1,8-i), (i+1,8-i+1)) and the right wing video in the known corresponding frame that has embedded watermark sequence in same embedded location, judge the image coefficient (c on this embedded location in the video of left road i, 8-i) left, (c i, 8-i+1) left, (c i+1,8-i) left, (c i+1,8-i+1) leftwith the image coefficient (c on this embedded location in the right wing video 8-i, i) right, (c 8-i, i+1) right, (c 8-i+1, i) right, (c 8-i+1, i+1) rightwhether meet possible stereo-picture coefficient restriction relation, i.e. formula (11) or (12), a ≠ m wherein, b ≠ n, and a-m>=b-n.If meet formula (11), determine that the value of the bit of the watermark sequence embedded is 0 on this embedded location of three-dimensional video-frequency; If meet formula (12), determine that the value of the bit of the watermark sequence embedded is 1 on this embedded location of three-dimensional video-frequency.If formula (11) and (12) do not meet, can select a-m and b-n middle distance ( ( c i , 8 - i ) left 2 + ( c i , 8 - i + 1 ) left 2 + ( c i + 1,8 - i ) left 2 + ( c i + 1,8 - i + 1 ) left 2 ) - ( ( c i , 8 - i ) right 2 + ( c i , 8 - i + 1 ) right 2 + ( c i + 1,8 - i ) right 2 + ( c i + 1,8 - i + 1 ) right 2 ) ) A nearer bit value that conduct embeds on this embedded location.Preferably, in a modification of the 8th example, can use the stereo-picture coefficient restriction relation of formula (19) as bit value 0 correspondence, and use formula (20) is as the stereo-picture coefficient restriction relation of bit value 1 correspondence, a ≠ m wherein, b ≠ n, and a-m >=b-n.

In above example, use stereo-picture coefficient restriction relation to extract watermark sequence from three-dimensional video-frequency.In an embodiment of the present invention, image coefficient in frame in the video of image coefficient restriction relation Jin Yuzuo road, left road is relevant, right wing image coefficient restriction relation is only relevant with the image coefficient in frame in the right wing video, therefore, also can extract watermark sequence from the video of independent Zuo road according to left road image coefficient restriction relation, and extract watermark sequence according to right wing image coefficient restriction relation from independent right wing video.Correspondingly, extract substituting or supplementing of parts 1320 and watermark sequence formation parts 1330 as above-mentioned watermark bit, extract parts according to comprising (unshowned) left road watermark bit in the watermark extraction apparatus of the embodiment of the present invention, for by the image coefficient on each predetermined embedded location of judging left road video, meeting Zuo road image coefficient restriction relation, extract the corresponding value of described left road image coefficient restriction relation, value as the bit in the watermark sequence embedded on the described embedded location in the video of Zuo road, and left road watermark sequence forms parts, form the watermark sequence of left road video for the value of the bit that extracts the watermark sequence that parts extract according to left road watermark bit, and/or the right wing watermark bit is extracted parts, for the satisfied right wing image coefficient restriction relation of image coefficient on each the predetermined embedded location by judging the right wing video, extract the corresponding value of described right wing image coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the right wing video, the right wing watermark sequence forms parts, forms the watermark sequence of right wing video for the value of the bit that extracts the watermark sequence that parts extract according to the right wing watermark bit.The operation principle that the operation principle of left road and right wing watermark bit extraction parts and left road and right wing watermark sequence formation parts and above-mentioned watermark bit are extracted parts 1320 and watermark sequence formation parts 1330 is similar, is not described in detail details more specifically here.

<3. three-dimensional video-frequency generating apparatus and method >

Can apply the technology that adds watermark for stereoscopic video of the present invention in the three-dimensional video-frequency generating apparatus.Figure 18 illustrates the schematic block diagram of three-dimensional video-frequency generating apparatus according to an embodiment of the invention.For brevity, only show the closely-related part with the present invention in figure.As shown in figure 18, three-dimensional video-frequency generating apparatus 1800 comprises left road video acquisition parts 1810, right wing video acquisition parts 1820 ,Zuo road watermark built in items 1830, right wing watermark built in items 1840 and three-dimensional watermark compound component 1850.Wherein, can use respectively according to the device Zhong Zuo road watermark built in items 120, right wing watermark built in items 130 and the three-dimensional watermark compound component 140 that add watermark for stereoscopic video of the embodiment of the present invention and implement Zuo road watermark built in items 1830, right wing watermark built in items 1840 and three-dimensional watermark compound component 1850 here.The following exemplary operation flow process of describing three-dimensional video-frequency generating apparatus 1800 according to Figure 19.

Figure 19 illustrates the indicative flowchart of three-dimensional video-frequency generation method according to an embodiment of the invention.As shown in figure 19, at step S1910(Ke Youzuo road video acquisition parts 1810, carry out) in, gather for three-dimensional video-frequency Zuo road video., at step S1920(, can be carried out by right wing video acquisition parts 1820 simultaneously) in, the right wing video for three-dimensional video-frequency gathered.Then at step S1930(Ke Youzuo road watermark built in items 1830, carry out) in, be adjusted at the image coefficient on described embedded location, watermarked sequence in the video of Yi Zuo road according to the corresponding left road of the value image coefficient restriction relation of the bit in the watermark sequence that will embed on each the predetermined embedded location in the video of left road.Simultaneously, at step S1940(, can be carried out by right wing watermark built in items 1840) in, be adjusted at the image coefficient on described embedded location according to the corresponding right wing image coefficient of the value restriction relation of the bit in the described watermark sequence that will embed on each the predetermined embedded location in the right wing video, with watermarked sequence in the right wing video.Then, at step S1950(, can be carried out by three-dimensional watermark compound component 1850) in, will distinguish watermarked sequence Zuo road video and right wing video compound stereoscopic video.

In above-mentioned three-dimensional video-frequency generating apparatus and method, the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation is different from right wing image coefficient restriction relation, and the corresponding left road of the same bit value of watermark sequence image coefficient restriction relation and right wing image coefficient restriction relation can be determined all relevant image coefficient restriction relations of image coefficient on the corresponding embedded location of image coefficient on the embedded location of the described same bit value in the video of Yu Zuo road and the described same bit value in the right wing video.

The more specifically details of step S1920 to S1950 can, with reference to the operation of the step S220 to S240 shown in figure 2, be not described in detail here.

<4. stereoscopic video playing device and method >

In stereoscopic video playing device, can apply of the present invention for extract the technology of watermark from three-dimensional video-frequency.Figure 20 illustrates the schematic block diagram of stereoscopic video playing device according to an embodiment of the invention.For brevity, only show the closely-related part with the present invention in figure.As shown in figure 20, stereoscopic video playing device 2000 comprises watermark extracting parts 2010 and video playback parts 2020.Watermark extracting parts 2010 further comprise that three-dimensional video-frequency Knock-Down Component 2011, watermark bit are extracted parts 2012, watermark sequence forms parts 2013 and watermark sequence verification component 2014.Can implement watermark extracting parts 2010 with the watermark extraction apparatus 1700 shown in Figure 17.Three-dimensional video-frequency Knock-Down Component 2011, watermark bit extract parts 2012, watermark sequence and form parts 2013 and watermark sequence verification component 2014 and can have respectively the function identical with corresponding component in watermark extraction apparatus 1700.The exemplary operation flow process of stereoscopic video playing device 2000 is described below with reference to Figure 21.

Figure 21 illustrates the indicative flowchart of three-dimensional video-frequency player method according to an embodiment of the invention.As shown in figure 21, at step S2111(, can be carried out by three-dimensional video-frequency Knock-Down Component 2011), S2112(can extract by watermark bit parts 2012 and carry out) and S2113(can be formed by watermark sequence parts 2013 and carry out) in, extract watermark sequence from the three-dimensional video-frequency of inputting.Step S2111 to S2113 is identical with the operation of the step S1410 to S1430 shown in Figure 14 respectively, is not described in detail here.At step S2114(, can be carried out by watermark sequence verification component 2014) in, determine according to the correlation of extracted watermark sequence and target watermark sequence whether extracted watermark sequence is the target watermark sequence.If judge that in step S2114 extracted watermark sequence is the target watermark sequence, at step S2120(, can be carried out by video playback parts 2020) in the broadcasting three-dimensional video-frequency.If judge that in step S2114 extracted watermark sequence is not the target watermark sequence, process directly finishes, and does not play three-dimensional video-frequency.

In other embodiments of the invention, watermark extracting parts 2010 also extract watermark sequence in Cong Zuo road video or right wing video separately.Correspondingly, extract substituting or supplementing of parts 2012 and watermark sequence formation parts 2013 as watermark bit, can also comprise watermark bit extraction parts in (unshowned) left road in watermark extracting parts 2010, for by the image coefficient on each predetermined embedded location of judging left road video, meeting Zuo road image coefficient restriction relation, extract the corresponding value of described left road image coefficient restriction relation, value as the bit in the watermark sequence embedded on the described embedded location in the video of Zuo road, and left road watermark sequence forms parts, form the watermark sequence of left road video for the value of the bit that extracts the watermark sequence that parts extract according to left road watermark bit, and/or the right wing watermark bit is extracted parts, for the satisfied right wing image coefficient restriction relation of image coefficient on each the predetermined embedded location by judging the right wing video, extract the corresponding value of described right wing image coefficient restriction relation, value as the bit in the watermark sequence embedded on the described embedded location in the right wing video, and the right wing watermark sequence forms parts, form the watermark sequence of right wing video for the value of the bit that extracts the watermark sequence that parts extract according to the right wing watermark bit.

As example rather than restriction, stereoscopic video playing device 2000 can be Blu-ray player for example, and correspondingly, above-mentioned three-dimensional video-frequency can be the blue light three-dimensional video-frequency, and video playback parts 2020 can be for playing the blue light three-dimensional video-frequency.

<5. can implement the computer configuation of embodiments of the invention/example >

Should be appreciated that according to all modules, unit in each device of the embodiment of the present invention and can be configured by the mode of software, firmware, hardware or its combination.Configure spendable concrete means or mode and be well known to those skilled in the art, do not repeat them here.In the situation that realize by software or firmware, from storage medium or network, the program that forms this software is installed to the computer with specialized hardware structure, this computer, when various program is installed, can be carried out various functions etc.

Figure 22 illustrates the block diagram of the structure of the computer that can realize embodiment/example of the present disclosure.In Figure 22, CPU (CPU) 2201 is according to the program of storage in read-only memory (ROM) 2202 or carry out various processing from the program that storage area 2208 is loaded into random-access memory (ram) 2203.In RAM 2203, also store as required data required when CPU 1601 carries out various processing etc.CPU 2201, ROM 2202 and RAM 2203 are connected to each other via bus 2204.Input/output interface 2205 also is connected to bus 2204.

Following parts are connected to input/output interface 2205: importation 2206(comprises keyboard, mouse etc.), output 2207(comprises display, such as cathode ray tube (CRT), liquid crystal display (LCD) etc., with loud speaker etc.), storage area 2208(comprises hard disk etc.), communications portion 2209(comprises that network interface unit is such as LAN card, modulator-demodulator etc.).Communications portion 2209 is processed such as the internet executive communication via network.As required, driver 2210 also can be connected to input/output interface 2205.Detachable media 2211, such as disk, CD, magneto optical disk, semiconductor memory etc. can be installed on driver 2210 as required, makes the computer program of therefrom reading be installed to as required in storage area 2208.

In the situation that realize above-mentioned series of processes by software, from network such as internet or storage medium such as detachable media 2211 is installed the program that forms softwares.

It will be understood by those of skill in the art that this storage medium is not limited to wherein having program stored therein shown in Figure 22, with equipment, distributes separately to provide the detachable media 2211 of program to the user.The example of detachable media 2211 comprises disk (comprising floppy disk (registered trade mark)), CD (comprising compact disc read-only memory (CD-ROM) and digital universal disc (DVD)), magneto optical disk (comprising mini-disk (MD) (registered trade mark)) and semiconductor memory.Perhaps, storage medium can be hard disk comprised in ROM 2202, storage area 2208 etc., computer program stored wherein, and be distributed to the user together with the equipment that comprises them.

The present invention also proposes a kind of program product that stores the instruction code that machine readable gets.When described instruction code is read and carried out by machine, can carry out and above-mentionedly according to the embodiment of the present invention, for stereoscopic video, add the method for watermark.

The present invention also proposes a kind of program product that stores the instruction code that machine readable gets.When described instruction code is read and is carried out by machine, can carry out above-mentioned according to the embodiment of the present invention for extract the method for watermark from three-dimensional video-frequency.

The present invention also proposes a kind of program product that stores the instruction code that machine readable gets.When described instruction code is read and carried out by machine, can carry out the above-mentioned player method of the three-dimensional video-frequency according to the embodiment of the present invention.

Correspondingly, for carrying the above-mentioned storage medium that stores the program product of the instruction code that machine readable gets, be also included within of the present invention open.Described storage medium includes but not limited to floppy disk, CD, magneto optical disk, storage card, memory stick etc.

In the above in the description to the specific embodiment of the invention, the feature of describing and/or illustrating for a kind of execution mode can be used in same or similar mode in one or more other execution mode, combined with the feature in other execution mode, or substitute the feature in other execution mode.

Should emphasize, term " comprises/comprises " existence that refers to feature, key element, step or assembly when this paper is used, but does not get rid of the existence of one or more further feature, key element, step or assembly or add.

In addition, the time sequencing of describing during method of the present invention is not limited to is to specifications carried out, also can according to other time sequencing ground, carry out concurrently or independently.The execution sequence of the method for therefore, describing in this specification is not construed as limiting technical scope of the present invention.

Describe by reference to the accompanying drawings embodiments of the invention in detail although above, it should be understood that execution mode described above is just for the present invention is described, and be not construed as limiting the invention.For a person skilled in the art, can make various modifications and changes and not deviate from the spirit and scope of the invention above-mentioned execution mode.Therefore, scope of the present invention is only limited by appended claim and equivalents thereof.

Claims (19)

1. a device that adds watermark for stereoscopic video comprises:
The three-dimensional video-frequency Knock-Down Component, for obtaining left road video and right wing video from three-dimensional video-frequency;
Left road watermark built in items, the corresponding left road of value image coefficient restriction relation for the bit in the watermark sequence that will embed on each the predetermined embedded location according to described left road video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in the video of described left road;
Right wing watermark built in items, the corresponding right wing image coefficient of value restriction relation for the bit in the described watermark sequence that will embed on each the predetermined embedded location according to described right wing video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in described right wing video; And
The three-dimensional video-frequency compound component, for the described left road video that will embed respectively described watermark sequence and right wing video compound stereoscopic video again,
Wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation can determine with described left road video in the embedded location of described same bit value on image coefficient and the relevant image coefficient restriction relation all of the image coefficient on the corresponding embedded location of the described same bit value in described right wing video.
2. add the device of watermark according to claim 1 for stereoscopic video, wherein, described image coefficient is pixel value in spatial domain or the conversion coefficient in transform domain, and described left road and right wing image coefficient restriction relation are all linear inequality or nonlinear inequalities.
According to claim 1 or 2 add the device of watermark for stereoscopic video, wherein,
Described left road watermark built in items comprises:
Whether left road coefficient decision parts, meet the corresponding left road of the value image coefficient restriction relation that will be embedded in the bit in the described watermark sequence on described embedded location for the image coefficient on each embedded location of judging described left road video; And
Left road coefficient adjustment parts, for in the situation that the image coefficient on described embedded location does not meet corresponding left road image coefficient restriction relation, be adjusted at the image coefficient on described embedded location so that the described image coefficient through adjusting meets described corresponding Zuo road image coefficient restriction relation; And
Described right wing watermark built in items comprises:
Whether right wing coefficient decision parts, meet the corresponding right wing image coefficient of the value restriction relation that will be embedded in the bit in the described watermark sequence on described embedded location for the image coefficient on each embedded location of judging described right wing video; And
Right wing coefficient adjustment parts, for in the situation that the image coefficient on described embedded location does not meet corresponding right wing image coefficient restriction relation, be adjusted at the image coefficient on described embedded location so that the described image coefficient through adjusting meets the right wing image coefficient restriction relation of described correspondence.
According to claim 1 or 2 add the device of watermark for stereoscopic video, also comprise:
The embedded location determining means, for according to the length of described watermark sequence, determine embedded location in the frame of the described watermark sequence to be embedded of described left road video of each bit in described watermark sequence and the embedded location in the frame of the described watermark sequence to be embedded of described right wing video.
5. add the device of watermark according to claim 1 or 2 for stereoscopic video, wherein, the corresponding described left road image coefficient restriction relation of the same bit value of described watermark sequence and the described right wing image coefficient restriction relation adjustment opposite direction to image coefficient.
According to claim 5 add the device of watermark for stereoscopic video, wherein, the bit value of described watermark sequence comprises the first bit value and the second bit value;
The single order or the second order deviation that image coefficient on the embedded location of described the first bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the first bit value road image coefficient restriction relation are more than or equal to the first left road adjustment threshold value, the single order or the second order deviation that image coefficient on the embedded location of described the second bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the second bit value road image coefficient restriction relation are less than the second left road adjustment threshold value, the described first left road is adjusted threshold value and is more than or equal to the described second left road adjustment threshold value, and
Single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the first bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the embedded location of described the first bit value in described right wing video are less than or equal to the first right wing adjustment threshold value, single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the second bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the embedded location of described the second bit value in described right wing video are greater than the second right wing adjustment threshold value, and described the first right wing is adjusted threshold value and is less than or equal to described the second right wing adjustment threshold value.
7. add the device of watermark according to claim 1 or 2 for stereoscopic video, wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation are identical but adjusting range is different to the adjustment direction of image coefficient.
According to claim 7 add the device of watermark for stereoscopic video, wherein, the bit value of described watermark sequence comprises the first bit value and the second bit value;
The single order or the second order deviation that image coefficient on the embedded location of described the first bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the first bit value road image coefficient restriction relation are more than or equal to the first left road adjustment threshold value, the single order or the second order deviation that image coefficient on the embedded location of described the second bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the second bit value road image coefficient restriction relation are less than the second left road adjustment threshold value, the described first left road is adjusted threshold value and is more than or equal to the described second left road adjustment threshold value, and
Single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the first bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the embedded location of described the first bit value in described right wing video are more than or equal to the first right wing adjustment threshold value, single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the second bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the corresponding embedded location in described right wing video are less than the second right wing adjustment threshold value, described the first right wing is adjusted threshold value and is more than or equal to described the second right wing adjustment threshold value, the described first left road is different with right wing adjustment threshold value, the described second left road is different with right wing adjustment threshold value, and the difference that the described first left road and right wing are adjusted threshold value is more than or equal to the poor of the described second left road and right wing adjustment threshold value.
9. one kind for extracting the device of watermark from three-dimensional video-frequency, comprising:
The three-dimensional video-frequency Knock-Down Component, for obtaining left road video and right wing video from three-dimensional video-frequency;
Watermark bit is extracted parts, the stereo-picture coefficient restriction relation satisfied with the image coefficient on corresponding embedded location in described right wing video for the image coefficient on each the predetermined embedded location by judging described left road video, extract the corresponding value of described stereo-picture coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the video of described left road and the corresponding embedded location in described right wing video; And
Watermark sequence forms parts, for the value of the bit of the watermark sequence according to extracted, forms described watermark sequence,
Wherein, described stereo-picture coefficient restriction relation is determined according to corresponding Zuo road image coefficient restriction relation and right wing image coefficient restriction relation, described left road image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described left road video to embed described watermark in the video of described left road, described right wing image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described right wing video to embed described watermark in described right wing video, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation.
According to claim 9 for extract the device of watermark from three-dimensional video-frequency, wherein, described image coefficient is pixel value in spatial domain or the conversion coefficient in transform domain, and described left road and right wing image coefficient restriction relation are all linear inequality or nonlinear inequalities.
11. according to claim 9 or 10 for extract the device of watermark from three-dimensional video-frequency, wherein, the corresponding described left road image coefficient restriction relation of the same bit value of described watermark sequence and the described right wing image coefficient restriction relation adjustment opposite direction to image coefficient.
12. according to claim 11 for extract the device of watermark from three-dimensional video-frequency, wherein, the bit value of described watermark sequence comprises the first bit value and the second bit value;
The single order or the second order deviation that image coefficient on the embedded location of described the first bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the first bit value road image coefficient restriction relation are more than or equal to the first left road adjustment threshold value, the single order or the second order deviation that image coefficient on the embedded location of described the second bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the second bit value road image coefficient restriction relation are less than the second left road adjustment threshold value, the described first left road is adjusted threshold value and is more than or equal to the described second left road adjustment threshold value,
Single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the first bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the embedded location of described the first bit value in described right wing video are less than or equal to the first right wing adjustment threshold value, single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the second bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the embedded location of described the second bit value in described right wing video are greater than the second right wing adjustment threshold value, described the first right wing is adjusted threshold value and is less than or equal to described the second right wing adjustment threshold value, and
Bit value corresponding to the first stereoscopic image coefficient limit restricting the relationship of the left video image is embedded into the first bit position value on the right side with a first bit value of said video embedded image coefficients corresponding to the position of the first or second deviation is greater than or equal to said first threshold value and the adjustment of the left and right to adjust the first threshold value determining a first perspective adjusted threshold value of the second bit image coefficients corresponding to the position of the stereoscopic image image coefficient limiting factor limiting relationship with the positions of the left and right video embedded in the video on the second bit value in the corresponding bit value of said second embedded a first or second order according to the determined deviation is less than the second threshold value and the adjustment of the left and right to adjust the second threshold value of the second threshold adjustment perspective.
13. according to claim 9 or 10 for extract the device of watermark from three-dimensional video-frequency, wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation are identical but adjusting range is different to the adjustment direction of image coefficient.
14. according to claim 13 for extract the device of watermark from three-dimensional video-frequency, wherein, the bit value of described watermark sequence comprises the first bit value and the second bit value;
The single order or the second order deviation that image coefficient on the embedded location of described the first bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the first bit value road image coefficient restriction relation are more than or equal to the first left road adjustment threshold value, the single order or the second order deviation that image coefficient on the embedded location of described the second bit value in the video of described left road are restricted to described image coefficient and adjacent image coefficient with the corresponding Zuo of described the second bit value road image coefficient restriction relation are less than the second left road adjustment threshold value, the described first left road is adjusted threshold value and is more than or equal to the described second left road adjustment threshold value,
Single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the first bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the embedded location of described the first bit value in described right wing video are more than or equal to the first right wing adjustment threshold value, single order or second order deviation that the right wing image coefficient restriction relation corresponding with described the second bit value is restricted to described image coefficient and adjacent image coefficient by the image coefficient on the corresponding embedded location in described right wing video are less than the second right wing adjustment threshold value, described the first right wing is adjusted threshold value and is more than or equal to described the second right wing adjustment threshold value, the described first left road is different with right wing adjustment threshold value, the described second left road is different with right wing adjustment threshold value, and the difference that the described first left road and right wing are adjusted threshold value is more than or equal to the poor of the described second left road and right wing adjustment threshold value, and
Bit value corresponding to the first stereoscopic image coefficient limit restricting the relationship of the left video image is embedded into the first bit position value on the right side with a first bit value of said video embedded image coefficients corresponding to the position of the first or second deviation is greater than or equal to said first threshold value and the adjustment of the left and right to adjust the first threshold value determining a first perspective adjusted threshold value of the second bit image coefficients corresponding to the position of the stereoscopic image image coefficient limiting factor limiting relationship with the positions of the left and right video embedded in the video on the second bit value in the corresponding bit value of said second embedded a first or second order according to the determined deviation is less than the second threshold value and the adjustment of the left and right to adjust the second threshold value of the second threshold adjustment perspective.
15. according to claim 9 or 10 for extract the device of watermark from three-dimensional video-frequency, also comprise:
The watermark sequence verification component, determine for the correlation of the watermark sequence according to extracted and target watermark sequence whether extracted watermark sequence is described target watermark sequence.
16. a three-dimensional video-frequency generating apparatus comprises:
Left road video acquisition parts, for gathering for three-dimensional video-frequency Zuo road video;
Right wing video acquisition parts, for gathering the right wing video for described three-dimensional video-frequency;
Left road watermark built in items, the corresponding left road of value image coefficient restriction relation for the bit in the watermark sequence that will embed on each the predetermined embedded location according to described left road video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in the video of described left road;
Right wing watermark built in items, the corresponding right wing image coefficient of value restriction relation for the bit in the described watermark sequence that will embed on each the predetermined embedded location according to described right wing video is adjusted at the image coefficient on described embedded location, to embed described watermark sequence in described right wing video; And
The three-dimensional video-frequency compound component, for described left road video and the right wing video compound stereoscopic video that will embed respectively described watermark sequence,
Wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation can determine with described left road video in the embedded location of described same bit value on image coefficient and the relevant image coefficient restriction relation all of the image coefficient on the corresponding embedded location of the described same bit value in described right wing video.
17. a stereoscopic video playing device comprises:
The watermark extracting parts, extract watermark for the three-dimensional video-frequency from input, comprising:
The three-dimensional video-frequency Knock-Down Component, for obtaining left road video and right wing video from described three-dimensional video-frequency;
Watermark bit is extracted parts, the stereo-picture coefficient restriction relation satisfied with the image coefficient on corresponding embedded location in described right wing video for the image coefficient on each the predetermined embedded location by judging described left road video, extract the corresponding value of described stereo-picture coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the video of described left road and the corresponding embedded location in described right wing video;
Watermark sequence forms parts, for the value of the bit of the watermark sequence according to extracted, forms described watermark sequence; And
The watermark sequence verification component, determine for the correlation of the watermark sequence according to extracted and target watermark sequence whether extracted watermark sequence is described target watermark sequence; And
The video playback parts, for working as extracted watermark sequence while being described target watermark sequence, play described three-dimensional video-frequency,
Wherein, described stereo-picture coefficient restriction relation is determined according to corresponding Zuo road image coefficient restriction relation and right wing image coefficient restriction relation, described left road image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described left road video to embed described watermark in the video of described left road, described right wing image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described right wing video to embed described watermark in described right wing video, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation.
18. a method that adds watermark for stereoscopic video comprises:
Obtain left road video and right wing video from three-dimensional video-frequency;
Be adjusted at the image coefficient on described embedded location according to the corresponding left road of the value image coefficient restriction relation of the bit in the watermark sequence that will embed on each the predetermined embedded location in the video of described left road, to embed described watermark sequence in the video of described left road;
Be adjusted at the image coefficient on described embedded location according to the corresponding right wing image coefficient of the value restriction relation of the bit in the described watermark sequence that will embed on each the predetermined embedded location in described right wing video, to embed described watermark sequence in described right wing video; And
To embed respectively the described left road video of described watermark sequence and right wing video compound stereoscopic video again,
Wherein, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation and described right wing image coefficient restriction relation can determine with described left road video in the embedded location of described same bit value on image coefficient and the relevant image coefficient restriction relation all of the image coefficient on the corresponding embedded location of the described same bit value in described right wing video.
19. one kind for extracting the method for watermark from three-dimensional video-frequency, comprising:
Obtain left road video and right wing video from three-dimensional video-frequency;
By the stereo-picture coefficient restriction relation of judging that image coefficient on each the predetermined embedded location in the video of described left road is satisfied with the image coefficient on corresponding embedded location in described right wing video, extract the corresponding value of described stereo-picture coefficient restriction relation, as the value of the bit in the watermark sequence embedded on the described embedded location in the video of described left road and the corresponding embedded location in described right wing video; And
Form described watermark sequence according to the value of the bit of extracted watermark sequence,
Wherein, described stereo-picture coefficient restriction relation is determined according to corresponding Zuo road image coefficient restriction relation and right wing image coefficient restriction relation, described left road image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described left road video to embed described watermark in the video of described left road, described right wing image coefficient restriction relation for the image coefficient on the embedded location of the value corresponding with described stereo-picture coefficient restriction relation that is adjusted at described right wing video to embed described watermark in described right wing video, and, the corresponding described left road of the same bit value of described watermark sequence image coefficient restriction relation is different from described right wing image coefficient restriction relation.
CN 201210169598 2012-05-28 2012-05-28 Methods and devices for watermarking stereoscopic video and extracting watermarks, stereoscopic video generation device and stereoscopic video play device CN103458256A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201210169598 CN103458256A (en) 2012-05-28 2012-05-28 Methods and devices for watermarking stereoscopic video and extracting watermarks, stereoscopic video generation device and stereoscopic video play device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201210169598 CN103458256A (en) 2012-05-28 2012-05-28 Methods and devices for watermarking stereoscopic video and extracting watermarks, stereoscopic video generation device and stereoscopic video play device

Publications (1)

Publication Number Publication Date
CN103458256A true CN103458256A (en) 2013-12-18

Family

ID=49740150

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201210169598 CN103458256A (en) 2012-05-28 2012-05-28 Methods and devices for watermarking stereoscopic video and extracting watermarks, stereoscopic video generation device and stereoscopic video play device

Country Status (1)

Country Link
CN (1) CN103458256A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105187817A (en) * 2015-09-09 2015-12-23 武汉大学 Glass-less 3D video encryption method based on digital watermark encryption and system
US10111246B2 (en) 2013-05-10 2018-10-23 Hfi Innovation Inc. Enhanced mechanism of scheduling request to multiple schedulers in a wireless network with inter base station carrier aggregation

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10111246B2 (en) 2013-05-10 2018-10-23 Hfi Innovation Inc. Enhanced mechanism of scheduling request to multiple schedulers in a wireless network with inter base station carrier aggregation
CN105187817A (en) * 2015-09-09 2015-12-23 武汉大学 Glass-less 3D video encryption method based on digital watermark encryption and system

Similar Documents

Publication Publication Date Title
Lei et al. Reversible watermarking scheme for medical image based on differential evolution
Singh et al. A survey of digital watermarking techniques, applications and attacks
Al-Dmour et al. A steganography embedding method based on edge identification and XOR coding
US8655055B2 (en) Method, system and computer program product for converting a 2D image into a 3D image
Lai An improved SVD-based watermarking scheme using human visual characteristics
Wang et al. Robust object removal with an exemplar-based image inpainting approach
CN102667811B (en) Alignment of objects in augmented reality
Tao et al. A wavelet-based particle swarm optimization algorithm for digital image watermarking
Wang et al. A majority-voting based watermarking scheme for color image tamper detection and recovery
EP2352120B1 (en) Network-based access to auxiliary data based on steganographic information
Hsu et al. Copyright protection scheme for digital images using visual cryptography and sampling methods
Potdar et al. A survey of digital image watermarking techniques
US8351643B2 (en) Media fingerprints that reliably correspond to media content
Cox et al. Electronic watermarking: the first 50 years
Su et al. A blind dual color images watermarking based on singular value decomposition
Mishra et al. Image encryption using Fibonacci-Lucas transformation
Tsai et al. Robust lossless image watermarking based on α-trimmed mean algorithm and support vector machine
Zhao et al. Effective watermarking scheme in the encrypted domain for buyer–seller watermarking protocol
CN101523434B (en) Image enhancement
Kaur et al. Image Watermarking Using LSB (Least Significant Bit)
Ouyang et al. Color image watermarking based on quaternion Fourier transform and improved uniform log-polar mapping
JP2009038810A (en) Network system, information terminal, server, detector, and digital content distributing method
CN1930585B (en) Creating a depth map
Hsiao et al. Block-based reversible data embedding
JP3431593B2 (en) Content generation device, digital watermark detection device, content generation method, digital watermark detection method, and recording medium

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20131218

C02 Deemed withdrawal of patent application after publication (patent law 2001)