CN101393635A - Watermark information embedding, extracting method and processing system - Google Patents
Watermark information embedding, extracting method and processing system Download PDFInfo
- Publication number
- CN101393635A CN101393635A CNA2008101712455A CN200810171245A CN101393635A CN 101393635 A CN101393635 A CN 101393635A CN A2008101712455 A CNA2008101712455 A CN A2008101712455A CN 200810171245 A CN200810171245 A CN 200810171245A CN 101393635 A CN101393635 A CN 101393635A
- Authority
- CN
- China
- Prior art keywords
- depth image
- zone
- block
- pixels
- embedding
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The embodiment of the invention provides a method for embedding and extracting watermark information, a device and a processing system thereof. The embedding method comprises the steps of conducting the occlusion detection for a range image corresponding to an original image; selecting a reference embedding region of the range image according to the result of the occlusion detection and the range image; matching up a target embedding region on the original image corresponding to the reference embedding region; and embedding acquired watermark into the target embedding region. After the occlusion detection for the range image, the embodiment of the invention can select the target embedding region of the original image according to the result of the occlusion detection, and embed the watermark information in the target embedding region. Since the occlusion detection is effectively conducted for an embedding position when the watermark information is embedded, the watermark information can not be embedded in to an occluded region of a new viewpoint image, so that the watermark information can be completely reserved in the new viewpoint image, thereby effectively extracting the watermark information.
Description
Technical field
The present invention relates to image processing field, particularly a kind of embedding of watermark information, extracting method and device and disposal system.
Background technology
Digital watermark technology is a up-to-date in the world Information Hiding Techniques.So-called digital watermark technology is exactly that watermark informations such as numeral, sequence number, literal, logos are embedded into the information processing technology in the multi-medium data.Image is the abundantest Digital Media of interconnection internet resource, because the attribute of human eye vision redundancy and image file itself makes image become the main carrier of watermark information.
At present watermark information is embedded in the new viewpoint image that utilization generates based on drafting (Depth Image BasedRendering the is called for short DIBR) technology of depth image normally on the basis that light field is drawn embed watermark information on each light field image.The light field method that generates the new viewpoint image mainly contains two kinds, a kind of method that is based on nearest light field image, and another kind is based on the method that two kinds of images that close on generate the new viewpoint image.
In realizing process of the present invention, the inventor finds that there are the following problems at least in the prior art: watermark information might can't be retained in the new viewpoint image, thereby can't effectively be extracted.
Summary of the invention
The embodiment of the invention provides a kind of embedding, extracting method and device and disposal system of watermark information, in order to realize that watermark information can intactly be retained in the new viewpoint image, can be extracted out effectively.
The embodiment of the invention provides a kind of embedding grammar of watermark information, comprising:
Depth image to the original image correspondence carries out occlusion detection;
The reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the zone;
On described original image, match with described and embed the zone with reference to embedding the corresponding target in zone;
The watermark information that obtains is embedded into described target to embed in the zone.
The embodiment of the invention provides a kind of extracting method of watermark information, comprising:
Depth image to the original image correspondence carries out occlusion detection;
The reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the district;
On described original image, match with described and embed the zone with reference to embedding the corresponding target in zone;
Embed the zone according to described target and in the new viewpoint image, match embedding to be extracted zone;
From described embedding to be extracted zone, extract watermark information.
The embodiment of the invention also provides a kind of flush mounting of watermark information, comprising:
The first occlusion detection module is used for the depth image of original image correspondence is carried out occlusion detection;
Primary importance is chosen module, be used for the result of the occlusion detection of carrying out and reference that described depth image is chosen on the described depth image and embed the zone, and on described original image, match and describedly embed regional with reference to embedding the corresponding target in zone according to the described first occlusion detection module;
The watermark acquisition module is used to obtain watermark information;
The watermark merge module is used for that the watermark information that described watermark acquisition module obtains is embedded into described primary importance and chooses the target embedding zone that module is chosen.
The embodiment of the invention also provides a kind of extraction element of watermark information, comprising:
The second occlusion detection module is used for the depth image of original image correspondence is carried out occlusion detection;
The second place is chosen module, be used for the result of the occlusion detection of carrying out and reference that described depth image is chosen on the described depth image and embed the zone, and on described original image, match and describedly embed regional with reference to embedding the corresponding target in zone according to the described second occlusion detection module;
The location matches module is used for choosing target that module chooses according to the described second place and embeds the zone to match embeddings to be extracted at the new viewpoint image regional;
The watermark extracting module is used for extracting watermark information from the embedding to be extracted zone that described second place matching module matches.
The embodiment of the invention provides a kind of disposal system of watermark information again, comprising:
The flush mounting of watermark information, be used for the depth image of original image correspondence is carried out occlusion detection, the reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the zone, on described original image, match with described and embed the zone, the watermark information that obtains is embedded into described target embeds in the zone with reference to embedding the corresponding target in zone;
The extraction element of watermark information, be used for the depth image of original image correspondence is carried out occlusion detection, the reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the zone, on described original image, match with described and embed the zone with reference to embedding the corresponding target in zone, embed the zone according to described target and in the new viewpoint image, match embedding to be extracted zone, from described embedding to be extracted zone, extract watermark information.
As shown from the above technical solution, the embodiment of the invention is by carrying out depth image after the occlusion detection, the target that can choose on the original image according to the result of occlusion detection embeds the zone, the watermark information that obtains is embedded into described target to embed in the zone, owing to during embed watermark information embedded location has been carried out effective occlusion detection, the occlusion area that watermark information can be embedded into the new viewpoint image, make watermark information intactly to be retained in the new viewpoint image, thereby can be extracted out effectively.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art, to do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below, apparently, accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to these accompanying drawings.
The schematic flow sheet of the embedding grammar of the watermark information that Fig. 1 provides for the embodiment of the invention one;
The schematic flow sheet of the embedding grammar of the watermark information that Fig. 2 (a) provides for the embodiment of the invention two;
Fig. 2 (b) is for carrying out the principle schematic of occlusion detection to block of pixels among Fig. 2 (a);
The schematic flow sheet of the extracting method of the watermark information that Fig. 3 provides for the embodiment of the invention three;
The schematic flow sheet of the extracting method of the watermark information that Fig. 4 provides for the embodiment of the invention four;
The structural representation of the flush mounting of the watermark information that Fig. 5 (a) provides for the embodiment of the invention five;
Fig. 5 (b) is the structural representation of the first occlusion detection module 51 among Fig. 5 (a);
The structural representation of the extraction element of the watermark information that Fig. 6 (a) provides for the embodiment of the invention six;
Fig. 6 (b) is the structural representation of the second occlusion detection module 61 among Fig. 6 (a);
The structural representation of the disposal system of the watermark information that Fig. 7 provides for the embodiment of the invention seven.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that is obtained under the creative work prerequisite.
The schematic flow sheet of the embedding grammar of the watermark information that Fig. 1 provides for the embodiment of the invention one, as shown in Figure 1, present embodiment can may further comprise the steps:
The occlusion detection of present embodiment by depth image is carried out, the target that can choose effectively on the original image embeds the zone, embeds in the zone thereby make watermark information can be embedded into target.Carried out effective occlusion detection owing to during embed watermark information the embedded location watermark is embedded the zone, the occlusion area of new viewpoint image can not appear watermark information is embedded into, make watermark information intactly to be retained in the new viewpoint image, thereby can be extracted out effectively.
The schematic flow sheet of the embedding grammar of the watermark information that Fig. 2 (a) provides for the embodiment of the invention two, shown in Fig. 2 (a), present embodiment can may further comprise the steps:
To choose a n * n block of pixels B is example, and the size of n value can depend on experimental result, through experimental verification, finds that n=16 has preferable performance.Be elaborated in the present embodiment step 204 selected n * n block of pixels being carried out occlusion detection.Fig. 2 (b) is among Fig. 2 (a) block of pixels being carried out the principle schematic of occlusion detection, and shown in Fig. 2 (b), Ω is that corner detects boundary; F represents whether be pixel on selected n * n block of pixels B, F=1 represents is, F=0 represents that F=-1 does not represent the pixel on non-this object; The projecting block of plane BP remarked pixel piece B; Plane FB represents the nearly face of the cubical view directions of three dimensions (rectangle ABCD place face); BPJ represents the projection (rectangle EFGH) of BP at the nearly face of three dimensions; BPJB represent plane BP and BPJ (rectangle EFGH) and between the stage body that constitutes of line; Block number a little among the SN remarked pixel piece B, SN=0 is an initial value.Set in advance the Ω corner and detect boundary, the numerical value of Ω equals the numerical value of the rotation angle (angle between new viewpoint and the original viewpoint) of new viewpoint image and original image, for example: the angle of the rotation angle of new viewpoint image and original image is 5 °, and Ω=5 ° are set so.
The pixel of depth value maximum is made as P1 among the selected pixels piece B, the plane F1 that does the vertical depth direction makes pixel P1 on the F1 of plane, F1 is as the mapping plane, with the projecting block BP of all pixels among the block of pixels B in plane of plane F1 projection formation, shown in the shade among the figure.With Ω is swing angle, doing two planes that become the 90+ Ω number of degrees with projecting block BP is plane F2, plane F3, intersection with plane FB (rectangle ABCD place face) is straight line EG, straight line FH respectively, and block of pixels BP goes up vertical projection straight line EF, the straight line GH of lower limb on the FB of plane, constitute BPJ (rectangle EFGH), BPJ (rectangle EFGH) and plane BP and between line constitute the BPJB truncated cone, the point that detects F=0 in BPJB promptly blocks number a little in this block of pixels, is designated as SN.
In the present embodiment step 205 according to block a little number and the depth information of pixel carry out the processing of grade of fit (Fitness), according to the numerical values recited of grade of fit in selected n * n block of pixels, choose can embed watermark information reference embed block of pixels, the big more watermark embedded block that is suitable as more of the numerical value of grade of fit is promptly with reference to embedding block of pixels.The processing of above-mentioned grade of fit can be calculated with following formula, promptly
Grade of fit=α * DepthF-β * SN;
In the formula, the adaptation associated depth value of DepthF remarked pixel point, the value of DepthF can determine according to following formula, promptly
If?depthAVG>DMax×BG,DepthF=depthAVG/2
Else?DepthF=depthAVG;
In the formula, the actual grade value of depth remarked pixel point; The mean depth value of all pixels among the depthAVG remarked pixel piece B; The upper limit of the scope of the depth value of DMax remarked pixel point; BG represents the background weights, and value is 1〉BG〉1/2.Wherein the selection of BG mainly is a depth range of judging background object, and BG is provided with and excessively may causes some backgrounds not to be judged to become background, BG be provided with too small may be as a setting with non-back of the body pixel value, generally choose between the 0.8-0.9 better;
Blocking number a little among the SN remarked pixel piece B, is the result that step 204 drew in the present embodiment;
α, β are two weighting parameters, and above-mentioned two weighting parameters can be chosen, and selection principle is to make that the order of magnitude of DepthF and SN is close, and selection principle also has relation with the value unit of the depth value of pixel.For example, SN is 50, and the degree of depth value of depth is 0-1, and the value of DepthF is 0.5, and the value of weighting parameter α, β is adjusted: if α=100, β=1 are set, then the degree of depth and the situation of blocking are just all balancedly considered; If α=100, β=5 are set, then stress more for the consideration of blocking; If be provided with α=1, β=, when promptly order of magnitude deviation was big, then the degree of depth just almost can be ignored for the influence of grade of fit.So the value of weighting parameter α, β based on the equilibrium number magnitude, is auxilliary to stress the aspect.
Watermark information is embedded into respectively in the target embedding frequency chunks and further can be divided into the embedding of non-blind watermark information and the embedding of blind watermark information in the present embodiment step 208.
For non-blind watermark information, concrete embedding grammar is target to be embedded frequency chunks carry out the intermediate frequency information change, and each intermediate frequency coefficient is through embedding the watermark information of 1 bit (bit) after revising.Concrete embedding grammar can be expressed as:
In the formula, wm represents watermark information, is one group of 0,1 sequence of being made up of the n bit, and wm (i) can represent the i bit of watermark information; X represents that target embeds the intermediate frequency coefficient in the frequency chunks; The target that Xw has represented to embed watermark information embeds the intermediate frequency coefficient in the frequency chunks.Wm (i) is embedded among the X, and CV represents a selected value, and value is 1<CV<32.When the CV value hour, less to the picture quality influence, but the robustness of watermark is relatively poor; When the CV value was big, bigger to the picture quality influence, the robustness of watermark was better.
For blind watermark information, watermark information is that watermark signal adopts band spectrum modulation, and modulated process can be shown below:
In the formula, the watermark signal after wm_signal represents to modulate, wm represents original watermark signal, is one group of 0,1 sequence of being made up of the n bit, wm (i) can represent the i bit of watermark information.In order to realize blind watermark detection when the detection of watermark information, in the time of can supposing i=1, wm (1) equals 1 all the time, and wm_pattern is the noise sequence that n random noise formed, wm_pattern
1May reside in any sample that contains the related watermark information of the embodiment of the invention.
Watermark signal after ovennodulation is embedded in the original signal, and telescopiny can be shown below:
X
w=X+α·wm_signal;
In the formula, α represents embedment strength, and X represents original signal, X
wSignal behind the expression embed watermark signal.Here it should be noted that following formula is that a kind of linear adder embeds, can also adopt multiplication to embed, perhaps non-linear embedding, and can adopt any existing robust watermarking embedding grammar, about these methods is not the emphasis of the embodiment of the invention, is not described in detail here.
In some watermaking system, watermark information can accurately be extracted, and this process is known as the extraction of watermark information.For example: in integrity validation is used, must accurately extract the watermark information of embedding, and confirm the integrality of multi-medium data by the integrality of watermark information.If detect the variation that part has taken place the watermark information that is extracted, can also determine the position that raw data is distorted by the position of the watermark information that changes.
For strong watermark, usually can not accurately extract the original watermark information of embedding, because strong if desired watermark of application, illustrate that this application is subjected to the attack of various malice probably, after the watermark information data were gone through these operations, the watermark information that extracts was changed beyond recognition usually.At this moment only need carry out the detection of watermark information.Usually the first step of the detection of watermark information is the extraction of watermark information, is the judgement of watermark information then.The common practice of the judgement of watermark information can adopt correlation detection.Select a correlativity judgement standard, the correlation of the watermark information that calculating extracts and the watermark information of appointment if correlation is enough high, can conclude substantially that then detected new viewpoint image contains the watermark information of appointment.In addition, the testing result of watermark information depends on a threshold value, when the result of correlation detection surpasses this threshold value, then can draw and contain the conclusion of specifying watermark information.This is actually a test of hypothesis problem in the theory of probability.When threshold value improved, empty inspection probability reduced, and false dismissal probability raises; When threshold value reduced, empty inspection probability raise, and false dismissal probability reduces.In the watermark applications of reality, generally more pay attention to control to void inspection probability.
The schematic flow sheet of the extracting method of the watermark information that Fig. 3 provides for the embodiment of the invention three, as shown in Figure 3, present embodiment can may further comprise the steps:
Present embodiment is actually a kind of extracting method of non-blind watermark information, the occlusion detection of present embodiment by depth image is carried out, the target that can locate effectively on the original image embeds the zone, thereby can will extract the embedding to be extracted zone of watermark information from the new viewpoint image.Carried out effective occlusion detection owing to during embed watermark information the embedded location watermark is embedded the zone, the occlusion area of new viewpoint image can not appear watermark information is embedded into, make watermark information intactly to be retained in the new viewpoint image, thereby can be extracted out effectively.
The schematic flow sheet of the extracting method of the watermark information that Fig. 4 provides for the embodiment of the invention four, as shown in Figure 4, present embodiment can may further comprise the steps:
The principle of the occlusion detection in the occlusion detection in the present embodiment step 404 and the embodiment of the invention two steps 204 is identical, repeats no more herein.
Further, further can also comprise the step that the watermark information that is extracted is detected after the step of from described block of pixels to be extracted, extracting watermark information in the present embodiment step 408.Extract with the step that detects and be specifically as follows: to the embedding in the new viewpoint image that matches the block of pixels to be extracted of watermark information carry out DCT, generate frequency chunks to be extracted, in this frequency chunks to be extracted, carry out the extraction and the detection of watermark information.
The extracting method of above-mentioned two watermark informations that embodiment provided only is applicable to the extraction of non-blind watermark information, when detecting for the extraction of blind watermark information, can adopt the method for traversal simply, calculate the data message and the wm_pattern of the possible block of pixels of institute
1Correlation, correlation value calculation is shown below:
When correlation during, think then and contain watermark signal in this block of pixels, and adopt the data message of following formula calculating pixel piece and the correlation of residual noise that when correlation during greater than thresholding T, then the pairing bit of this noise is 1, otherwise is 0 greater than default thresholding T.
The watermark information method that the embodiment of the invention provides promptly embeds and the occlusion detection of extracting method by depth image is carried out, the target that can choose effectively and locate on the original image embeds the zone, embed in the zone thereby make watermark information can be embedded into target, and can will extract the embedding to be extracted zone of watermark information from the new viewpoint image.Carried out effective occlusion detection owing to during embed watermark information the embedded location target is embedded the zone, the occlusion area of new viewpoint image can not appear watermark information is embedded into, make watermark information intactly to be retained in the new viewpoint image, thereby can be extracted out effectively.
The structural representation of the flush mounting of the watermark information that Fig. 5 (a) provides for the embodiment of the invention five, shown in Fig. 5 (a), the flush mounting of the watermark information of present embodiment can comprise that the first occlusion detection module 51, primary importance choose module 52, watermark acquisition module 53 and watermark merge module 54.Wherein, the depth image of 51 pairs of original image correspondences of the first occlusion detection module carries out after the occlusion detection, primary importance is chosen the result of the occlusion detection that module 52 carries out according to the first occlusion detection module 51 and the reference that depth image is chosen on the depth image embeds the zone, and on original image, match and above-mentionedly embed the zone with reference to embedding the corresponding target in zone, the watermark information that watermark merge module 54 is obtained watermark acquisition module 53 is embedded into primary importance and chooses in the selected targets embedding zone of module 52.Reference wherein embeds the zone can be made up of with reference to embedding block of pixels one or more, and target embeds the zone and can be made up of one or more targets embedding block of pixels.
Further, shown in Fig. 5 (b), the first occlusion detection module 51 further can also comprise the first three-dimensionalreconstruction unit 511, first object segmentation unit 512, first information acquiring unit 513 and the first occlusion detection unit 514 in the present embodiment.Wherein, the first three-dimensionalreconstruction unit 511 obtains the three-dimensional information of pixel on this depth image according to depth image, 512 pairs of these depth images of first object segmentation unit carry out Object Segmentation, choose the block of pixels of default size, the three-dimensional information of pixel obtains the three-dimensional information of pixel on the selected block of pixels of above-mentioned first object segmentation unit 512 on the depth image that first information acquiring unit 513 is obtained according to the above-mentioned first three-dimensionalreconstruction unit 511, the first occlusion detection unit 514 carries out occlusion detection according to the three-dimensional information of the pixel that above-mentioned first information acquiring unit 513 is obtained to above-mentioned first object segmentation unit, 512 selected block of pixels, determines to block in the described block of pixels number a little.
The occlusion detection that the flush mounting of the watermark information that the embodiment of the invention provides carries out depth image by the occlusion detection module, make the position choose the target that module can choose on the original image effectively and embed the zone, embed in the zone thereby make the watermark merge module watermark information can be embedded into target.Carried out effective occlusion detection owing to during embed watermark information the embedded location target is embedded the zone, the occlusion area of new viewpoint image can not appear watermark information is embedded into, make watermark information intactly to be retained in the new viewpoint image, thereby can be extracted out effectively.
The structural representation of the extraction element of the watermark information that Fig. 6 (a) provides for the embodiment of the invention six, shown in Fig. 6 (a), the extraction element of the watermark information of present embodiment can comprise that the second occlusion detection module 61, the second place choose module 62, location matches module 63 and watermark extracting module 64.Wherein, the depth image of 61 pairs of original image correspondences of the second occlusion detection module carries out after the occlusion detection, the second place is chosen the result of the occlusion detection that module 62 carries out according to the second occlusion detection module 61 and the reference that depth image is chosen on the depth image embeds the zone, and on original image, match with above-mentioned with reference to embedding corresponding target embedding zone, zone, location matches module 63 is chosen the regional embedding to be extracted zone that matches of target embedding that module 62 is chosen according to the second place in the new viewpoint image, watermark extracting module 64 is extracted watermark information from the embedding to be extracted zone that location matches module 63 matches.Reference wherein embeds the zone can be made up of with reference to embedding block of pixels one or more, and target embeds the zone and can be made up of one or more targets embedding block of pixels.
Further, shown in Fig. 6 (b), the second occlusion detection module 61 further can also comprise the second three-dimensionalreconstruction unit 611, second object segmentation unit 612, second information acquisition unit 613 and the second occlusion detection unit 614 in the present embodiment.Wherein, the second three-dimensionalreconstruction unit 611 obtains the three-dimensional information of pixel on this depth image according to depth image, 612 pairs of these depth images of second object segmentation unit carry out Object Segmentation, choose the block of pixels of default size, the three-dimensional information of pixel obtains the three-dimensional information of pixel on the selected block of pixels of above-mentioned second object segmentation unit 612 on the depth image that second information acquisition unit 613 is obtained according to the above-mentioned second three-dimensionalreconstruction unit 611, the second occlusion detection unit 614 carries out occlusion detection according to the three-dimensional information of the pixel that above-mentioned second information acquisition unit 613 is obtained to above-mentioned second object segmentation unit, 612 selected block of pixels, determines to block in the described block of pixels number a little.
The occlusion detection that the extraction element of the watermark information that the embodiment of the invention provides carries out depth image by the occlusion detection module, make the position choose the target that module can locate on the original image effectively and embed the zone, thereby make the watermark extracting module will extract the embedding to be extracted zone of watermark information from the new viewpoint image.Carried out effective occlusion detection owing to during embed watermark information the embedded location target is embedded the zone, the occlusion area of new viewpoint image can not appear watermark information is embedded into, make watermark information intactly to be retained in the new viewpoint image, thereby can be extracted out effectively.
The structural representation of the disposal system of the watermark information that Fig. 7 provides for the embodiment of the invention seven, as shown in Figure 7, present embodiment can comprise the flush mounting 71 of watermark information and the extraction element 72 of watermark information.The depth image of 71 pairs of original image correspondences of flush mounting of watermark information carries out occlusion detection, the reference of choosing on the depth image according to the result and the depth image of above-mentioned occlusion detection embeds the zone, on above-mentioned original image, match with above-mentioned and embed the zone, the watermark information that obtains is embedded into above-mentioned target embeds in the zone with reference to embedding the corresponding target in zone; The depth image of 72 pairs of original image correspondences of extraction element of watermark information carries out occlusion detection, the reference of choosing on the depth image according to the result and the depth image of above-mentioned occlusion detection embeds the zone, on above-mentioned original image, match with above-mentioned and embed the zone with reference to embedding the corresponding target in zone, embed the zone according to above-mentioned target and in the new viewpoint image, match embedding to be extracted zone, from above-mentioned embedding to be extracted zone, extract watermark information.Reference wherein embeds the zone can be made up of with reference to embedding block of pixels one or more, and target embeds the zone and can be made up of one or more targets embedding block of pixels.
The flush mounting of the watermark information that the flush mounting 71 of the watermark information in the present embodiment can provide for the embodiment of the invention five, the extraction element of the watermark information that the extraction element 72 of watermark information can provide for the embodiment of the invention six.
The watermark information device that the embodiment of the invention provides promptly embeds the occlusion detection of by the occlusion detection module depth image being carried out with extraction element, make the position choose the target that module can choose on the original image effectively and embed the zone, and the position is chosen the target that module can locate on the original image effectively and is embedded the zone, embed in the zone thereby make the watermark merge module watermark information can be embedded into target, and the watermark extracting module can will extract the embedding to be extracted zone of watermark information from the new viewpoint image.Carried out effective occlusion detection owing to during embed watermark information the embedded location target is embedded the zone, the occlusion area of new viewpoint image can not appear watermark information is embedded into, make watermark information intactly to be retained in the new viewpoint image, thereby can be extracted out effectively.
One of ordinary skill in the art will appreciate that: all or part of step that realizes said method embodiment can be finished by the relevant hardware of programmed instruction, aforesaid program can be stored in the computer read/write memory medium, this program is carried out the step that comprises said method embodiment when carrying out; And aforesaid storage medium comprises: various media that can be program code stored such as ROM, RAM, magnetic disc or CD.
It should be noted that at last: above embodiment only in order to technical scheme of the present invention to be described, is not intended to limit; Although with reference to previous embodiment the present invention is had been described in detail, those of ordinary skill in the art is to be understood that: it still can be made amendment to the technical scheme that aforementioned each embodiment put down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution break away from the spirit and scope of various embodiments of the present invention technical scheme.
Claims (15)
1, a kind of embedding grammar of watermark information is characterized in that, comprising:
Depth image to the original image correspondence carries out occlusion detection;
The reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the zone;
On described original image, match with described and embed the zone with reference to embedding the corresponding target in zone;
The watermark information that obtains is embedded into described target to embed in the zone.
2, method according to claim 1 is characterized in that, described depth image to the original image correspondence carries out occlusion detection and comprises:
Obtain the three-dimensional information of pixel on the described depth image according to described depth image;
Described depth image is carried out Object Segmentation, choose the block of pixels of default size;
Obtain the three-dimensional information of pixel on the described block of pixels according to the three-dimensional information of pixel on the described depth image;
Three-dimensional information according to described pixel carries out occlusion detection to described block of pixels, determines to block in the described block of pixels number a little.
3, method according to claim 1 is characterized in that,
The described reference embedding zone of choosing on the described depth image according to the result and the described depth image of described occlusion detection comprises: the reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds block of pixels;
Describedly embed the zone and comprise with reference to embedding the corresponding target in zone matching on the described original image with described: according to the embedding of reference on described depth image block of pixels in the target embedding block of pixels that matches on the described original image on the described original image.
4, method according to claim 3, it is characterized in that described result and described depth image according to described occlusion detection chosen reference on the described depth image and embedded block of pixels and be specially: according to described block a little number and the depth information of described pixel in described block of pixels, choose with reference to the embedding block of pixels.
According to claim 3 or 4 described methods, it is characterized in that 5, the described watermark information that obtains is embedded in the described target embedding zone comprises:
Described target is embedded block of pixels carry out discrete cosine transform, generate target and embed frequency chunks;
Described watermark information is embedded into described target respectively to embed in the frequency chunks;
The described target that has embedded watermark information is embedded frequency chunks carry out inverse discrete cosine transform, generate the target that has embedded watermark information and embed block of pixels.
6, a kind of extracting method of watermark information is characterized in that, comprising:
Depth image to the original image correspondence carries out occlusion detection;
The reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the district;
On described original image, match with described and embed the zone with reference to embedding the corresponding target in zone;
Embed the zone according to described target and in the new viewpoint image, match embedding to be extracted zone;
From described embedding to be extracted zone, extract watermark information.
7, method according to claim 6 is characterized in that, described depth image to the original image correspondence carries out occlusion detection and comprises:
Obtain the three-dimensional information of pixel on the described depth image according to described depth image;
Described depth image is carried out Object Segmentation, choose the block of pixels of default size;
Obtain the three-dimensional information of pixel on the described block of pixels according to the three-dimensional information of pixel on the described depth image;
Three-dimensional information according to described pixel carries out occlusion detection to described block of pixels, determines to block in the described block of pixels number a little.
8, method according to claim 6 is characterized in that,
The described reference embedding zone of choosing on the described depth image according to the result and the described depth image of described occlusion detection comprises: the reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds block of pixels;
Describedly embed the zone and comprise with reference to embedding the corresponding target in zone matching on the described original image with described: according to the embedding of reference on described depth image block of pixels in the target embedding block of pixels that matches on the described original image on the described original image.
9, method according to claim 8, it is characterized in that described result and described depth image according to described occlusion detection chosen reference on the described depth image and embedded block of pixels and be specially: according to described block a little number and the depth information of described pixel in described block of pixels, choose with reference to the embedding block of pixels.
According to claim 6,7,8 or 9 described methods, it is characterized in that 10, the described watermark information that extracts also comprises afterwards: described watermark information is detected from described embedding to be extracted zone.
11, a kind of flush mounting of watermark information is characterized in that, comprising:
The first occlusion detection module is used for the depth image of original image correspondence is carried out occlusion detection;
Primary importance is chosen module, be used for the result of the occlusion detection of carrying out and reference that described depth image is chosen on the described depth image and embed the zone, and on described original image, match and describedly embed regional with reference to embedding the corresponding target in zone according to the described first occlusion detection module;
The watermark acquisition module is used to obtain watermark information;
The watermark merge module is used for that the watermark information that described watermark acquisition module obtains is embedded into described primary importance and chooses the target embedding zone that module is chosen.
12, device according to claim 11 is characterized in that, the described first occlusion detection module comprises:
The first three-dimensionalreconstruction unit is used for obtaining according to depth image the three-dimensional information of pixel on the described depth image;
First object segmentation unit is used for described depth image is carried out Object Segmentation, chooses the block of pixels of default size;
First information acquiring unit, the three-dimensional information of pixel obtains the three-dimensional information of pixel on the block of pixels that described first object segmentation unit chooses on the depth image that is used for obtaining according to the described first three-dimensionalreconstruction unit;
The first occlusion detection unit is used for the block of pixels that the three-dimensional information of the pixel that obtains according to described first information acquiring unit chooses described first object segmentation unit and carries out occlusion detection, determines to block in the described block of pixels number a little.
13, a kind of extraction element of watermark information is characterized in that, comprising:
The second occlusion detection module is used for the depth image of original image correspondence is carried out occlusion detection;
The second place is chosen module, be used for the result of the occlusion detection of carrying out and reference that described depth image is chosen on the described depth image and embed the zone, and on described original image, match and describedly embed regional with reference to embedding the corresponding target in zone according to the described second occlusion detection module;
The location matches module is used for choosing target that module chooses according to the described second place and embeds the zone to match embeddings to be extracted at the new viewpoint image regional;
The watermark extracting module is used for extracting watermark information from the embedding to be extracted zone that described second place matching module matches.
14, device according to claim 13 is characterized in that, the described second occlusion detection module comprises:
The second three-dimensionalreconstruction unit is used for obtaining according to depth image the three-dimensional information of pixel on the described depth image;
Second object segmentation unit is used for described depth image is carried out Object Segmentation, chooses the block of pixels of default size;
Second information acquisition unit, the three-dimensional information of pixel obtains the three-dimensional information of pixel on the block of pixels that described second object segmentation unit chooses on the depth image that is used for obtaining according to the described second three-dimensionalreconstruction unit;
The second occlusion detection unit is used for the block of pixels that the three-dimensional information of the pixel that obtains according to described second information acquisition unit chooses described second object segmentation unit and carries out occlusion detection, determines to block in the described block of pixels number a little.
15, a kind of disposal system of watermark information is characterized in that, comprising:
The flush mounting of watermark information, be used for the depth image of original image correspondence is carried out occlusion detection, the reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the zone, on described original image, match with described and embed the zone, the watermark information that obtains is embedded into described target embeds in the zone with reference to embedding the corresponding target in zone;
The extraction element of watermark information, be used for the depth image of original image correspondence is carried out occlusion detection, the reference of choosing on the described depth image according to the result and the described depth image of described occlusion detection embeds the zone, on described original image, match with described and embed the zone with reference to embedding the corresponding target in zone, embed the zone according to described target and in the new viewpoint image, match embedding to be extracted zone, from described embedding to be extracted zone, extract watermark information.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2008101712455A CN101393635B (en) | 2008-10-27 | 2008-10-27 | Watermark information embedding, extracting method and processing system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2008101712455A CN101393635B (en) | 2008-10-27 | 2008-10-27 | Watermark information embedding, extracting method and processing system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101393635A true CN101393635A (en) | 2009-03-25 |
CN101393635B CN101393635B (en) | 2011-08-10 |
Family
ID=40493919
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2008101712455A Expired - Fee Related CN101393635B (en) | 2008-10-27 | 2008-10-27 | Watermark information embedding, extracting method and processing system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101393635B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101990081A (en) * | 2010-11-11 | 2011-03-23 | 宁波大学 | Method for protecting copyright of virtual viewpoint image |
CN103037268A (en) * | 2012-12-07 | 2013-04-10 | 广东威创视讯科技股份有限公司 | Method and device for watermark adding |
CN103198449A (en) * | 2013-04-18 | 2013-07-10 | 山东大学 | Digital watermarking method for three-dimensional image |
CN103366332A (en) * | 2013-06-18 | 2013-10-23 | 河海大学 | Depth information-based image watermarking method |
CN109496325A (en) * | 2016-07-29 | 2019-03-19 | 索尼公司 | Image processing apparatus and image processing method |
CN109685708A (en) * | 2018-12-26 | 2019-04-26 | 珠海奔图电子有限公司 | Image processing method and device, electronic equipment and computer readable storage medium |
CN109919821A (en) * | 2017-12-12 | 2019-06-21 | 深圳大学 | Insertion, extracting method and the storage medium of the dual blind watermatking of three-dimensional digital model |
CN114426923A (en) * | 2022-03-31 | 2022-05-03 | 季华实验室 | Environmental virus sampling robot and method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1105991C (en) * | 1996-07-25 | 2003-04-16 | 明碁电脑股份有限公司 | Image output device |
-
2008
- 2008-10-27 CN CN2008101712455A patent/CN101393635B/en not_active Expired - Fee Related
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101990081A (en) * | 2010-11-11 | 2011-03-23 | 宁波大学 | Method for protecting copyright of virtual viewpoint image |
CN103037268A (en) * | 2012-12-07 | 2013-04-10 | 广东威创视讯科技股份有限公司 | Method and device for watermark adding |
CN103037268B (en) * | 2012-12-07 | 2015-07-08 | 广东威创视讯科技股份有限公司 | Method and device for watermark adding |
CN103198449B (en) * | 2013-04-18 | 2016-01-20 | 山东大学 | A kind of digital watermark method of 3-D view |
CN103198449A (en) * | 2013-04-18 | 2013-07-10 | 山东大学 | Digital watermarking method for three-dimensional image |
CN103366332A (en) * | 2013-06-18 | 2013-10-23 | 河海大学 | Depth information-based image watermarking method |
CN103366332B (en) * | 2013-06-18 | 2015-11-25 | 河海大学 | A kind of image watermark method based on depth information |
CN109496325A (en) * | 2016-07-29 | 2019-03-19 | 索尼公司 | Image processing apparatus and image processing method |
CN109919821A (en) * | 2017-12-12 | 2019-06-21 | 深圳大学 | Insertion, extracting method and the storage medium of the dual blind watermatking of three-dimensional digital model |
CN109685708A (en) * | 2018-12-26 | 2019-04-26 | 珠海奔图电子有限公司 | Image processing method and device, electronic equipment and computer readable storage medium |
CN109685708B (en) * | 2018-12-26 | 2024-02-06 | 珠海奔图电子有限公司 | Image processing method and device, electronic equipment and computer readable storage medium |
CN114426923A (en) * | 2022-03-31 | 2022-05-03 | 季华实验室 | Environmental virus sampling robot and method |
CN114426923B (en) * | 2022-03-31 | 2022-07-12 | 季华实验室 | Environmental virus sampling robot and method |
Also Published As
Publication number | Publication date |
---|---|
CN101393635B (en) | 2011-08-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101393635B (en) | Watermark information embedding, extracting method and processing system | |
Nikolaidis et al. | Region-based image watermarking | |
US7158653B2 (en) | Encoding information in a watermark | |
EP0901102B1 (en) | Watermark embedding method and system | |
CN104680473A (en) | Machine learning-based color image watermark embedding and detecting method | |
KR20060121972A (en) | Electronic watermarking method, electronic watermark detecting method, apparatus and program | |
CN101990081B (en) | Method for protecting copyright of virtual viewpoint image | |
CN112241929B (en) | Watermark embedding method, watermark detecting method, watermark embedding device, watermark detecting device and electronic equipment | |
CN110796582B (en) | Reversible plain watermark method and device based on vector graphic compression | |
JP2000350001A (en) | Method for embedding digital watermark information and method for extracting the same | |
CN108550099A (en) | The method and apparatus of watermark in a kind of removal image | |
CN104766269A (en) | Spread transform dither modulation watermarking method based on JND brightness model | |
Mun et al. | A robust 3D mesh watermarking scheme against cropping | |
KR102027880B1 (en) | Deep learning based 3D model Similarity evaluation method using depth image | |
JP2005524348A (en) | Image edge watermarking method and system | |
CN113628091B (en) | Safety information extraction method and device for electronic display screen content reproduction scene | |
CN106709855B (en) | 3D model watermarking method based on Beamlet line feature positioning | |
KR20070073332A (en) | Robust image watermarking using scale invariant feature transform | |
JP2004038980A (en) | Methods for analyzing image and generating threshold, program executing the method, computer-readable medium storing the program, and apparatus executing the method | |
CN113393363B (en) | Watermark embedding and extracting method, storage medium and electronic device | |
Singh et al. | A novel approach of 3d object watermarking algorithm using vertex normal | |
CN114630130A (en) | Face-changing video tracing method and system based on deep learning | |
KR100697768B1 (en) | The method for digital watermarking using circular hough transform | |
KR100945726B1 (en) | Apparatus for watermarking based on local auto-correlation function and method therefor | |
CN114565500A (en) | Reversible image authentication method and system based on prediction error expansion |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20110810 Termination date: 20161027 |