CN101980280A - Watermark embedding method, detection method, devices and watermark processing system - Google Patents
Watermark embedding method, detection method, devices and watermark processing system Download PDFInfo
- Publication number
- CN101980280A CN101980280A CN 201010511725 CN201010511725A CN101980280A CN 101980280 A CN101980280 A CN 101980280A CN 201010511725 CN201010511725 CN 201010511725 CN 201010511725 A CN201010511725 A CN 201010511725A CN 101980280 A CN101980280 A CN 101980280A
- Authority
- CN
- China
- Prior art keywords
- watermark information
- embedded
- watermark
- scale
- zone
- 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.)
- Pending
Links
Images
Abstract
The invention discloses a watermark embedded method, a watermark detection method, a watermark embedding device, a watermark detection device and a watermark processing system. The watermark embedding method comprises the following steps of: determining adjacent areas of a preset embedded point; calculating the curvature changes of each adjacent area; selecting the adjacent areas of which the curvature changes do not exceed a preset change threshold as areas where the embedding is performed; converting watermark information to be embedded into binary watermark information; and embedding the binary watermark information into the areas where the embedding is performed, wherein the number of the areas embedded with the binary watermark information is the same as the watermark information length of the binary watermark information, and each areas where the embedding is performed is embedded with only one bit of data of the binary watermark information. Through the technical scheme, the watermark information is embedded into the areas with the small curvature changes, and when a three-dimensional grid model is operated, the areas embedded with the watermark information are difficult to change and the embedded watermark information is accurate to ensure the accuracy of extracted watermark information to further reduce the possibility of misjudgment.
Description
Technical field
The present invention relates to the Information Hiding Techniques field, more particularly, relate to a kind of watermark embedding method, detection method, device and system for processing watermark.
Background technology
3D (three-dimensional, three-dimensional) grid model when expression 3D object with its more convenient simpler advantage, be widely used in every field such as industrial manufacturing, three-dimensional animation, virtual reality, become the requisite message form that each field 3D shape is described.Be accompanied by the develop rapidly of kownledge economy, various three-dimensional modeling datas, be presented on the network as casper, product model etc., viewed, download or propagate, the application of three-dimensional modeling data, shared and exploitation have greatly been promoted, yet also make the bootlegger can duplicate and propagate unwarranted three-dimensional modeling data with extremely cheap cost, grievous injury the rights and interests of knowledge author.How protecting the intellecture property of three-dimensional modeling data effectively, concerning that can kownledge economy develop in a healthy way on the internet, is the IT circles problem demanding prompt solutions.
3D grid digital watermarking embedding and detection method are at present the simplest, quick, objective and compellent to the kinds of protect means of three-dimensional model intellecture property.It embeds the form of copyright information with watermark information and be hidden in the three-dimensional grid model data under the prerequisite of the normal use that does not influence data, is sightless to the user, protects the intellecture property of three-dimensional grid model data effectively.Simultaneously, by watermark information in the extraction model data and original watermark information contrast, can judge whether the watermark information information in the extraction model data is distorted.
But, 3D grid digital watermarking embedding at present and detection method are at random the form of version information with watermark information to be embedded and be hidden in the three-dimensional grid model data, when the three-dimensional network model is for example carrying out: when rotation, convergent-divergent, cutting, noise, fairing operation, tend to its watermark information that is embedded in is caused change, therefore when extracting watermark information, the probability of makeing mistakes of watermark information increases, and then causes judging by accident disconnected probability increase.
Summary of the invention
In view of this, the object of the present invention is to provide a kind of watermark embedding method, detection method, device and system for processing watermark, to solve in the prior art when the three-dimensional grid model of embed watermark is operated at random, the watermark information that causes embedding changes, therefore when extracting watermark information, the probability of makeing mistakes of watermark information increases, and then causes judging by accident the problem that disconnected probability increases.
In order to address the above problem, the technical scheme of the embodiment of the invention is as follows:
The embodiment of the invention provides a kind of watermark embedding method, comprising:
Determine the default some v that embeds
iNeighboring region Nv (v
i, r), the distance radius r of described default embedding point and described neighboring region sets according to the watermark information intensity of watermark information to be embedded and watermark information capacity;
Calculate each neighboring region Nv (v
i, curved transition D r)
K(v
i, r);
Choose curved transition D
K(v
i, r) do not surpass the neighboring region Nv (v that presets change threshold
i, r) as zone to be embedded;
Changing watermark information to be embedded is scale-of-two watermark information W;
Described scale-of-two watermark information W is embedded in the zone to be embedded, the number in the zone to be embedded of described embedding scale-of-two watermark information W is identical with the watermark information length l en (W) of described scale-of-two watermark information W, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information W.
Corresponding with it, the embodiment of the invention also provides a kind of watermark flush mounting, comprising:
The neighboring region determination module is used for determining the default some v that embeds
iNeighboring region Nv (v
i, r), the distance radius r of described default embedding point and described neighboring region sets according to the watermark information intensity of watermark information to be embedded and watermark information capacity;
The curved transition computing module that links to each other with the neighboring region determination module is used to calculate each neighboring region Nv (v
i, curved transition D r)
K(v
i, r);
Module is chosen in the zone to be embedded that links to each other with the curved transition computing module, is used to choose curved transition D
K(v
i, r) do not surpass the neighboring region Nv (v that presets change threshold
i, r) as zone to be embedded;
The watermark modular converter, being used to change watermark information to be embedded is scale-of-two watermark information W;
Choose the watermark merge module that module links to each other with watermark modular converter and zone to be embedded, be used for described scale-of-two watermark information W is embedded into zone to be embedded, the number in the zone to be embedded of described embedding scale-of-two watermark information W is identical with the watermark information length l en (W) of described scale-of-two watermark information W, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information W.
The embodiment of the invention provides a kind of method of detecting watermarks, comprising:
Extract the embedded scale-of-two watermark information W in zone to be embedded, described zone to be embedded does not surpass default change threshold according to this zone curved transition and chooses, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information W, the number in zone to be embedded is identical with the watermark information length of described scale-of-two watermark information W, and described scale-of-two watermark information W is that watermark information conversion to be embedded draws;
According to default original watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
Corresponding with it, the embodiment of the invention also provides a kind of watermark detection device, comprising:
The watermark extracting module, be used to extract the embedded scale-of-two watermark information in zone to be embedded, described zone to be embedded does not surpass default change threshold according to this zone curved transition and chooses, and each zone to be embedded only embeds data of scale-of-two watermark information, the number in zone to be embedded is identical with the watermark information length of described scale-of-two watermark information, and described scale-of-two watermark information is that watermark information conversion to be embedded draws;
With the watermark detection module that the watermark extracting module links to each other, be used for according to default original watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
The embodiment of the invention also provides a kind of system for processing watermark, comprising:
The watermark flush mounting, be used to choose curved transition and do not surpass the neighboring region of default change threshold as zone to be embedded, watermark information to be embedded is converted to the scale-of-two watermark information, and described scale-of-two watermark information is embedded in the zone to be embedded, the number in the zone to be embedded of described embedding scale-of-two watermark information is identical with the watermark information length of described scale-of-two watermark information, and each zone to be embedded only embeds data of scale-of-two watermark information;
The watermark detection device is used to extract the embedded scale-of-two watermark information in zone to be embedded, and according to default original watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
The applicant is through repeatedly experiment discovery, the little zone of curved transition three-dimensional grid model carry out such as: when rotation, convergent-divergent, cutting, noise, fairing operation, the probability of its variation is obviously low with respect to the probability of curved transition regional change greatly, therefore, for the accuracy that guarantees to judge that watermark information is distorted, watermark information need be embedded into the little zone of curved transition.Technical scheme provided by the invention was at first determined the default some v that embeds before embed watermark information
iNeighboring region Nv (v
i, r), calculate each neighboring region Nv (v
i, curved transition D r)
K(v
i, r), then according to curved transition D
K(v
i, r) choose curved transition D
K(v
i, r) do not surpass the neighboring region Nv (v that presets change threshold
i, r) as zone to be embedded.Scale-of-two watermark information W after the conversion is embedded in the zone to be embedded identical with the watermark information length of described scale-of-two watermark information W, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information W.
Use technique scheme, watermark information is embedded into the little zone of curved transition, therefore when three-dimensional grid model is operated, the zone of embed watermark information is difficult for changing, the watermark information of its embedding is correct, the watermark information that assurance is extracted is accurate, and then reduces the disconnected probability of erroneous judgement, thus the problem that the disconnected probability of the erroneous judgement that the judgement watermark information that embed watermark information caused at random in the solution prior art is distorted increases.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention, to do simple introduction to the accompanying drawing of required use among the embodiment 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.
Fig. 1 is for embedding the situation of change of four kinds of models behind the watermark information with same watermark information strength;
The process flow diagram of the watermark embedding method that Fig. 2 provides for the embodiment of the invention;
Fig. 3 is the neighboring region synoptic diagram;
Fig. 4 is for concerning synoptic diagram between neighboring region point;
Fig. 5 has the model synoptic diagram behind the watermark information of different watermark information intensity for master pattern applies;
Fig. 6 is the variation synoptic diagram behind each embed watermark information in the model;
The structural representation of the watermark flush mounting that Fig. 7 provides for the embodiment of the invention;
The process flow diagram of the method for detecting watermarks that Fig. 8 provides for the embodiment of the invention;
The structural representation of the watermark detection device that Fig. 9 provides for the embodiment of the invention;
The structural representation of the system for processing watermark that Figure 10 provides for the embodiment of the invention.
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 are not making under the creative work prerequisite, and the every other embodiment that is obtained belongs to protection domain of the present invention.
Curvature: the degree of crook of representation model;
Curved transition: the intensity of variation of expression curvature.
At random watermark information to be embedded is embedded and is hidden in the three-dimensional grid model data in order to solve present watermark embedding and detection method, when the three-dimensional network model is for example carrying out: when rotation, convergent-divergent, cutting, noise, fairing operation, tend to its watermark information that is embedded in is caused change, therefore when extracting watermark information, the probability of makeing mistakes of watermark information increases, and then causes judging by accident disconnected probability increase.The embodiment of the invention provides a kind of watermark embedding method, watermark information to be embedded is embedded in the little zone of curved transition, because the applicant finds after deliberation, when three-dimensional grid model was operated, the probability that the zone that its probability that changes is big with respect to curved transition changes was obviously low.
Before the watermark embedding method that the elaboration embodiment of the invention provides, at first why explanation chooses the little zone of curved transition.See also Fig. 1, Fig. 1 is the situation of change that embeds four kinds of models behind the watermark information with same watermark information strength.
As shown in Figure 1, solid line is depicted as master pattern, and four kinds of models among the figure, and curvature is ascending successively from left to right.Four kinds of models are all applied identical disturbance, promptly embed the watermark information of same watermark information strength.The dotted line that obtains is the model after the disturbance.As can be seen, the model that curvature is big more, the translation of its generation is more little, and is difficult more from visually perceived.
But pure curvature can not reflect the visual characteristic of three-dimensional model fully.For example: a flat site, in the promptly low area of curvature, the outshot that has higher curvature has important perceived characteristic; In the dense higher curvature outshot, a flat site, promptly low area of curvature also has important perceived characteristic.What curved transition was represented is the intensity of variation of curvature, can well reflect the variation of the curvature integral body in certain zone, thereby reflects the bulk property in zone.Have the zone that changes than deep camber and more be subjected to the concern of vision, have higher vision attention compared with the zone of less curved transition.
Yet, the zone that has a high vision attention is with respect to the zone of low vision attention, i.e. the zone littler than curved transition, the zone that curved transition is big is when operating three-dimensional grid model, easy more changing, for example: the curved transition zone is worn distortion or data change.When extracting watermark information, the number that causes meeting in this zone the point of customizing messages surpasses 70% with the ratio of always counting, then can't extract this regional watermark information, and reality should the zone embedding have watermark information, and then cause the watermark information mistake extracted, when comparing, judge that watermark information is distorted with original watermark information.Yet in fact the watermark information that embeds in the three-dimensional grid model is not distorted, and it is disconnected erroneous judgement to occur.Therefore, when selecting the embed watermark information zone, need avoid the zone of selecting curved transition big, and choose the little zone of curved transition, make it can be relatively easy to hide watermark information, can prevent to judge by accident disconnected generation again.
Therefore,, when promptly verifying copyright, guarantee to judge that the accuracy whether watermark information is distorted increases, need choose the little zone of curved transition in order to reach the purpose that watermark information embeds.
Through above-mentioned research, watermark embedding method provided by the invention is embedded into watermark information to be embedded in the little zone of curved transition.Fig. 2 is the process flow diagram of the watermark embedding method that provides of the embodiment of the invention, comprising:
S101: determine the default some v that embeds
iNeighboring region Nv (v
i, r), described default embedding point v
iSet according to the watermark information intensity of watermark information to be embedded and watermark information capacity with the distance radius r of described neighboring region;
The default some v that embeds
iBe to choose according to the curved transition of model, neighboring region Nv (v
i, r) with the default some v that embeds
iBe shown with the relation table of distance radius r: Nv (v
i, r)={ v
j| ‖ v
jOne v
i‖≤r}, as shown in Figure 3.The default some v that embeds of Fig. 3 hollow core point expression
i, solid dot v
jForm neighboring region Nv (v
i, r), as can be seen from Figure 3 the zone of three-dimensional grid model is made up of different triangular facets, and promptly each zone comprises at least one triangular facet.
Distance radius r is determined by the watermark information intensity and the watermark information capacity of watermark information to be embedded.For example: when the watermark information intensity of watermark information to be embedded is 1% watermark information capacity of three-dimensional grid model bounding box length when being 128, watermark information intensity r is 4% of a three-dimensional grid model bounding box length.
S102: calculate each neighboring region Nv (v
i, curved transition D r)
K(v
i, r);
G
k(v
i, r) be neighboring region Nv (v
i, the r) mean curvature of using Gauss's method of weighting to calculate, D
K(v
i, r) be neighboring region Nv (v
i, curved transition r), computing formula is as follows:
Wherein: Num (v
i, r) expression neighboring region Nv (v
i, the number of the point that is comprised in r), K (v
i) be neighboring region Nv (v
i, r) in each the some v
iCurvature value.
For the curvature value of three-dimensional grid model, discrete curvature evaluation method on the good triangle mesh curved surface has been arranged at present.In watermark embedding method provided by the invention, adopt the Voronoi method of propositions such as Meyer to estimate curvature.The basic thought of Meyer method is the limit or a linear proximity of smooth curved surface being regarded as gang's grid, the metric property on each summit of triangle gridding is seen carried out four than the average tolerance Δ=2Hn=∫ ∫ of space lattice in a little field of this point
AΔ dA/A.The summit of triangle gridding is designated as p
i,
Represent the set on the summit in its 1-neighborhood, N (i) expression and some p
iThe bar number on the limit that is connected, A (p
i) expression comprises a p
iLeg-of-mutton area sum.
To area A=A (p
i) discrete follow the example of into: if A (p
i) be oxygon, then get circumcentre of a triangle and connect except that a some p
iThe mid point on other two limits obtains new area A beyond the opposite side
AcuteIf A (p
i) be obtuse triangle, be the mid point of obtuse angle corresponding sides is linked to each other with other mid point on two limits, obtain area A
AbtuseTo triangle gridding arbitrarily, then be to try to achieve whole mixed area A with above two kinds of methods.
According to Laplace-Beltrami operator character, ∫ ∫ is arranged
AΔ dA=∫ ∫
AΔ
U, vDudv utilizes Gauss theorem again,
Therefore,
α
Ij, β
IjBe respectively ∠ p
ip
J-1p
j, ∠ p
ip
J+1p
j, some p
i, the some p
j, the some p
J+1, the some pj
-1Concern that synoptic diagram as shown in Figure 4.
Use θ
iExpression three-dimensional grid model mid point p
iThe angle at place, the Gauss-Bonnet theorem in the application of differential geometry has
So, utilizing above-mentioned discrete method equally, can get:
Wherein: N is three-dimensional grid model mid point p
iBe adjacent the angle number that each point is formed in the zone.With Fig. 4 is example, and the value of N is 3, promptly removes summit p in the value of N and the three-dimensional grid model
iOutside the number of other points identical.
Therefore, for each the some v in the 3D grid
i, the curvature value that estimates is designated as K (v
i), and K (v
i) computing formula be:
S103: choose curved transition D
K(v
i, r) do not surpass the neighboring region Nv (v that presets change threshold
i, r) as regional Mv (v to be embedded
i, r);
S104: changing watermark information to be embedded is scale-of-two watermark information W;
Watermark information to be embedded is converted to one, and len (W) is a watermark information length by the 0 and 1 scale-of-two watermark information W that forms, wherein the unit watermark information be expressed as w ∈ 0,1}.When embed watermark information, each watermark information w of unit is embedded into independent to be embedded regional Mv (v
i, r) in.
S105: described scale-of-two watermark information W is embedded into regional Mv (v to be embedded
i, r) in, the to be embedded regional Mv (v of described embedding scale-of-two watermark information W
i, number r) is identical with the watermark information length l en (W) of described scale-of-two watermark information W, and each regional Mv (v to be embedded
i, r) only embed the one digit number certificate of scale-of-two watermark information W.Wherein:
Step S105 may further comprise the steps:
S1051: with regional Mv (v to be embedded
i, r) according to curved transition D
K(v
i, r) sort, constitute regional ensemble to be embedded;
Suppose that default change threshold is ε
D, regional Mv (v to be embedded
i, r) in all curved transition D
K(v
i, r) less than ε
DThe zone form to embed regional ensemble B={Mv (v
i, r) | D
k(v
i, r)＜ε
D, and regional Mv (v to be embedded in the set B
i, r) according to D
K(v
i, r) value is ascending or descending sorts;
S1052: calculate regional Mv (v to be embedded in the regional ensemble to be embedded
i, number len r) (B);
S1053: choose regional Mv (v to be embedded
i, position r) is positioned at regional ensemble to be embedded
The to be embedded regional Mv (v at place
i, r),, embed described scale-of-two watermark information, each regional Mv (v to be embedded as embedding the zone
i, r) only embed the one digit number certificate of scale-of-two watermark information, i.e. each regional Mv (v to be embedded
i, r) only embed a watermark information w of unit.
Elaborate below and how the watermark information w of unit is embedded into regional Mv (v to be embedded
i, r).
For the face f in the three-dimensional grid model
i, face f
iBe compositing area Nv (v
i, certain face r), its normal vector are normal (f
i), area is area (f
i), the normal vector normal (f of three-dimensional grid model mid point
i) can calculate according to the adjacent surface of point:
Wherein: Mv (v
i) expression and vertex v
iThe bar number on the limit that is connected.
Embedding algorithm can be expressed as:
Wherein:
Be expressed as the zone behind the embed watermark information, F () expression watermark information imbedding function.
Further analyze Mv (v
i, the r) set of forming by some points, and to Mv (v
i, the effect that r) applies will finally embody Mv (v
i, therefore r) the variation of the coordinate of interior each point decomposes F () in each point and goes.Specific practice as shown in Figure 5, Fig. 5 is the model synoptic diagram after master pattern applies the watermark information with different watermark information intensity, wherein: solid line is represented master pattern, dotted line represents to apply the model after the disturbance of different watermark information intensity.As seen from Figure 5, more little for the variation that master pattern applies, the fragility that is watermark information is big more, the deformation of its generation is also just more little, the result who brings thus is exactly that its anti-tamper aggressiveness is strong more, but antijamming capability is poor more, easily by in the Internet communication those change unintentionally disturb, therefore need choose watermark information intensity d according to specific tamper-proof requirement
iCome embed watermark information.Watermark information intensity d
iFor F () to Mv (v
i, the disturbance that r) applies is to determine according to the conditions such as fragility of watermark information to be embedded.
To Mv (v
i, the effect that r) applies will finally embody Mv (v
i, the r) variation of the coordinate of interior each point, therefore, at definite watermark information intensity d
iAfter, need be with watermark information intensity d
iBe applied to Mv (v
i, r) in the point in, for the watermark information intensity d (v of each point
j) have:
V wherein
j∈ Nv (v
i, r).
Be illustrated in when embedding and consider regional Mv (v to be embedded
i, the visual characteristic of integral body r) is for Mv (v
i, r) interior point near the center applies less influence, and applies bigger influence near the part on border, can guarantee the fragility of watermark information itself with this, also can guarantee the fluency of the vision of zone boundary part.As shown in Figure 6, Fig. 6 is the variation synoptic diagram behind each embed watermark information in the model, and the point of light color is represented the preceding Mv (v of embed watermark information among the figure
i, the r) point in is after dark point represents to embed 0 watermark information
In the point.(1-a) K (v of formula
i) show that the size of curvature also exerts an influence to embedment strength; Weights a then represents the preceding shared proportion of two parts, need decide according to concrete model and demand, and the value of weights a is 0.5 among the present invention.
At last, { 0,1} is embedded in a little the coordinate and goes with the watermark information w ∈ of unit.Embedding information 1 for a single point can be write as:
Embedding information 0 can be written as:
Wherein:
Some v behind the expression embed watermark information
j
Use technique scheme, before embed watermark information, determine the default some v that embeds
iNeighboring region Nv (v
i, r), calculate each neighboring region Nv (v
i, curved transition D r)
K(v
i, r), according to curved transition D
K(v
i, r) choose curved transition D
K(v
i, r) do not surpass the neighboring region Nv (v that presets change threshold
i, r) as zone to be embedded.Scale-of-two watermark information after the conversion is embedded in the zone to be embedded identical with the watermark information length of described scale-of-two watermark information, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information.Realize the checking copyright and anti-tamper in, improve the accuracy of judging that watermark information is distorted, thereby solved the problem that erroneous judgement disconnected probability that the judgement watermark information that embed watermark information caused at random in the prior art distorted increases.
Corresponding with it, the embodiment of the invention also provides a kind of watermark flush mounting, its structural representation comprises as shown in Figure 7: module 73, watermark modular converter 74 and watermark merge module 75 are chosen in neighboring region determination module 71, curved transition computing module 72, zone to be embedded.Wherein:
Neighboring region determination module 71 is used for determining the default some v that embeds
iNeighboring region Nv (v
i, r), the distance radius r of described default embedding point and described neighboring region sets according to the watermark information intensity of watermark information to be embedded and watermark information capacity;
Curved transition computing module 72 is used to calculate each neighboring region Nv (v
i, curved transition D r)
K(v
i, r);
Watermark modular converter 74, being used to change watermark information to be embedded is scale-of-two watermark information W;
Region ordering to be embedded unit is used for zone to be embedded according to curved transition D
K(v
i, r) sort, constitute regional ensemble to be embedded;
Regional number computing unit to be embedded is used for calculating the number len (B) in regional ensemble to be embedded zone to be embedded;
The unit is chosen in zone to be embedded, is used to choose be positioned at regional ensemble to be embedded
The zone to be embedded of position is as embedding the zone;
Embed the unit, be used to embed described scale-of-two watermark information.
Whether see also Fig. 8, Fig. 8 is the process flow diagram of a kind of method of detecting watermarks of providing of the embodiment of the invention, distorted to detect embed watermark information, comprising:
S201: extract the embedded scale-of-two watermark information in zone to be embedded, described zone to be embedded does not surpass default change threshold according to this zone curved transition and chooses, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information, the number in zone to be embedded is identical with the watermark information length of described scale-of-two watermark information, and described scale-of-two watermark information is that watermark information conversion to be embedded draws;
The leaching process of watermark information is exactly will be from three-dimensional grid model to be detected
In the data that embed are recovered out.From watermark information embedding method shown in Figure 2 can learn watermark information be embedded into dispersedly the zone each the point in, so the watermark information leaching process at first is that each point in the zone is detected, so carry out Macro or mass analysis, and then the watermark information of hiding in the zone is extracted at watermark information with each point.
At first, analyze a single point.For a v
i∈ M,
Middle corresponding point is
The unit's of extracting watermark information in the following method then:
Secondly, analyze whole zone.For Nv (v
i, r) be the zone of embed watermark information in original three-dimensional grid model M,
Be Nv (v
i, r) exist
Middle corresponding zone, num
1Be designated as
In
Be the number of 1 point, num
0Be designated as
In
Be the number of 0 point, then can extract the watermark information that the zone embeds in the following method:
Num (v
i, r) be the sum of regional mid point, ε
p∈ (0.5,1] be watermark information withdrawal ratio value, choose ε in the watermark information extracting method provided by the invention
p=0.7, if promptly regional
In the number of the point that meets customizing messages surpass at 70% o'clock with the ratio of always counting, corresponding watermark information just can from
In extract.
S202: according to default original watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
Step S202 is embodied as: original binary watermark information that comparison is default and the scale-of-two watermark information that extracts obtain testing result, the scale-of-two watermark information that arrives with Detection and Extraction.If default original binary watermark information is different with the comparison result of the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is distorted; If default original binary watermark information is identical with the comparison result of the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is not distorted.
Need to prove: the watermark information of embedding is the scale-of-two watermark information, therefore, the time is the scale-of-two watermark information that will extract and the original binary watermark information contrast after the conversion in contrast.Certainly the scale-of-two watermark information that extracts can also be converted to the watermark information with the original watermark information same format, again itself and non-switched original watermark information be contrasted, also can realize the detection whether watermark information is distorted.
Use technique scheme, can realize anti-tamper.What choose owing to embedding is regional simultaneously is that the variation of model mean curvature surpasses the zone of presetting change threshold, therefore when three-dimensional grid model is operated, the zone of embed watermark information is difficult for changing, the watermark information information of its embedding is correct, the watermark information that assurance is extracted is accurate, and then reduce the disconnected probability of erroneous judgement, thereby the problem that the disconnected probability of the erroneous judgement that embed watermark information caused at random in the solution prior art increases.
Corresponding with it, the present invention also provides a kind of watermark detection device, comprising: watermark extracting module 91 and watermark detection module 92.Its structural representation as shown in Figure 9, wherein:
Watermark extracting module 91, be used to extract the embedded scale-of-two watermark information in zone to be embedded, described zone to be embedded does not surpass default change threshold according to this zone curved transition and chooses, and each zone to be embedded only embeds data of scale-of-two watermark information, the number in zone to be embedded is identical with the watermark information length of described scale-of-two watermark information, and described scale-of-two watermark information is that watermark information conversion to be embedded draws;
Watermark detection module 92 is used for according to default original binary watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.Watermark detection module 92 comprises: watermark comparing unit and watermarking detecting results unit.Wherein:
The watermark comparing unit is used to compare default original binary watermark information and the scale-of-two watermark information that extracts, and obtains comparison result;
The watermarking detecting results unit, be used for according to comparison result, obtain testing result, the scale-of-two watermark information that arrives with Detection and Extraction, if comparison result is different for default original binary watermark information and the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is distorted; If comparison result is identical for default original binary watermark information and the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is not distorted.
Below provide method that concrete example and experimental result thereof prove that the embodiment of the invention provides guarantee the checking copyright and anti-tamper in, guarantee that the watermark information that extracts is accurate, and then reduce the disconnected probability of erroneous judgement.When embedding and extract watermark information, what adopt respectively is Fourier transform and inversefouriertransform.
128 watermark informations are embedded in Bunny, Happy_vrip and three three-dimensional grid model data of Dragon, adopt translation, rotation, convergent-divergent, cutting, noise, fairing that three-dimensional grid model is operated respectively.Wherein: the scale-value that Bunny adopted is 4%length, and the scale-value that Happy_vrip and Dragon adopted is 2%length, and watermark information intensity all is 1%length, and length is the length of model bounding box.Table 1 is that the present invention and prior art judge that performance index relatively when adopting above-mentioned experiment.Judge that performance index are: mistake figure place and False Rate.
The anti-attack performance index of table 1 the present invention and prior art relatively
As can be seen from Table 1, method provided by the invention is when judging whether watermark information is distorted, and its False Rate obviously reduces the embed watermark information at random with respect to prior art.
The embodiment of the invention also provides a kind of system for processing watermark, and its structural representation as shown in figure 10.Comprise: watermark flush mounting 101 and watermark detection device 102, wherein:
Watermark flush mounting 101, be used to choose curved transition and do not surpass the neighboring region of default change threshold as zone to be embedded, watermark information to be embedded is converted to the scale-of-two watermark information, and described scale-of-two watermark information is embedded in the zone to be embedded, the number in the zone to be embedded of described embedding scale-of-two watermark information is identical with the watermark information length of described scale-of-two watermark information, and each zone to be embedded only embeds data of scale-of-two watermark information;
Watermark detection device 102 is used to extract the embedded scale-of-two watermark information in zone to be embedded, and according to default original binary watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
Use native system, can realize verifying copyright and anti-tamper simultaneously, and further solve the problem that the disconnected probability of erroneous judgement that embed watermark information caused at random in the prior art judgement watermark information distorted increases.
Each embodiment adopts the mode of going forward one by one to describe in this instructions, and what each embodiment stressed all is and the difference of other embodiment that identical similar part is mutually referring to getting final product between each embodiment.For the disclosed device of embodiment, because it is corresponding with the embodiment disclosed method, so description is fairly simple, relevant part partly illustrates referring to method and gets final product.
To the above-mentioned explanation of the disclosed embodiments, make this area professional and technical personnel can realize or use the present invention.Multiple modification to these embodiment will be conspicuous concerning those skilled in the art, and defined herein General Principle can realize under the situation that does not break away from the spirit or scope of the present invention in other embodiments.Therefore, the present invention will can not be restricted to these embodiment shown in this article, but will meet and principle disclosed herein and features of novelty the wideest corresponding to scope.
Claims (12)
1. a watermark embedding method is characterized in that, comprising:
Determine the default some v that embeds
iNeighboring region Nv (v
i, r), the distance radius r of described default embedding point and described neighboring region sets according to the watermark information intensity of watermark information to be embedded and watermark information capacity;
Calculate each neighboring region Nv (vi, curved transition D r)
K(v
i, r);
Choose curved transition D
K(v
i, r) do not surpass the neighboring region Nv (v that presets change threshold
i, r) as zone to be embedded;
Changing watermark information to be embedded is scale-of-two watermark information W;
Described scale-of-two watermark information W is embedded in the zone to be embedded, the number in the zone to be embedded of described embedding scale-of-two watermark information W is identical with the watermark information length l en (W) of described scale-of-two watermark information W, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information W.
2. watermark embedding method according to claim 1 is characterized in that, described scale-of-two watermark information W is embedded in the zone to be embedded comprise:
With zone to be embedded according to curved transition D
K(v
i, r) sort, constitute regional ensemble to be embedded;
Calculate the number len (B) in zone to be embedded in the regional ensemble to be embedded;
Choose and be positioned at regional ensemble to be embedded
The zone to be embedded of position embeds described scale-of-two watermark information W as embedding the zone.
3. watermark embedding method according to claim 2 is characterized in that, the curved transition D of described each neighboring region of calculating
K(v
i, be r) according to neighboring region Nv (v
i, mean curvature G r)
K(v
i, r) calculate.
4. watermark embedding method according to claim 3 is characterized in that, described mean curvature G
K(v
i, r) be to use Gauss's method of weighting to calculate.
5. a watermark flush mounting is characterized in that, comprising:
The neighboring region determination module is used for determining the default some v that embeds
iNeighboring region Nv (v
i, r), the distance radius r of described default embedding point and described neighboring region sets according to the watermark information intensity of watermark information to be embedded and watermark information capacity;
The curved transition computing module that links to each other with the neighboring region determination module is used to calculate each neighboring region Nv (v
i, curved transition D r)
K(v
i, r);
Module is chosen in the zone to be embedded that links to each other with the curved transition computing module, is used to choose curved transition D
K(v
i, r) do not surpass the neighboring region Nv (v that presets change threshold
i, r) as zone to be embedded;
The watermark modular converter, being used to change watermark information to be embedded is scale-of-two watermark information W;
Choose the watermark merge module that module links to each other with watermark modular converter and zone to be embedded, be used for described scale-of-two watermark information W is embedded into zone to be embedded, the number in the zone to be embedded of described embedding scale-of-two watermark information W is identical with the watermark information length l en (W) of described scale-of-two watermark information W, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information W.
6. watermark flush mounting according to claim 5 is characterized in that, described watermark merge module comprises:
Region ordering to be embedded unit is used for zone to be embedded according to curved transition D
K(v
i, r) sort, constitute regional ensemble to be embedded;
The to be embedded regional number computing unit that links to each other with region ordering to be embedded unit is used for calculating the number len (B) in regional ensemble to be embedded zone to be embedded;
Choose the unit with the zone to be embedded that regional number computing unit to be embedded links to each other, be used to choose and be positioned at regional ensemble to be embedded
The zone to be embedded of position is as embedding the zone;
Choose the embedding unit that the unit links to each other with zone to be embedded, be used to embed described scale-of-two watermark information W.
7. a method of detecting watermarks is characterized in that, comprising:
Extract the embedded scale-of-two watermark information W in zone to be embedded, described zone to be embedded does not surpass default change threshold according to this zone curved transition and chooses, and each zone to be embedded only embeds the one digit number certificate of scale-of-two watermark information W, the number in zone to be embedded is identical with the watermark information length of described scale-of-two watermark information W, and described scale-of-two watermark information W is that watermark information conversion to be embedded draws;
According to default original watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
8. method of detecting watermarks according to claim 7, it is characterized in that, the original watermark information that described foundation is default, Detection and Extraction to the scale-of-two watermark information be specially: original binary watermark information that comparison is default and the scale-of-two watermark information that extracts, obtain testing result, the scale-of-two watermark information that arrives with Detection and Extraction.
9. method of detecting watermarks according to claim 8 is characterized in that, if default original binary watermark information is different with the comparison result of the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is distorted; If default original binary watermark information is identical with the comparison result of the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is not distorted.
10. a watermark detection device is characterized in that, comprising:
The watermark extracting module, be used to extract the embedded scale-of-two watermark information in zone to be embedded, described zone to be embedded does not surpass default change threshold according to this zone curved transition and chooses, and each zone to be embedded only embeds data of scale-of-two watermark information, the number in zone to be embedded is identical with the watermark information length of described scale-of-two watermark information, and described scale-of-two watermark information is that watermark information conversion to be embedded draws;
With the watermark detection module that the watermark extracting module links to each other, be used for according to default original watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
11. watermark detection device according to claim 10 is characterized in that, described watermark detection module comprises:
The watermark comparing unit is used to compare default original binary watermark information and the scale-of-two watermark information that extracts, and obtains comparison result;
The watermarking detecting results unit that links to each other with the watermark comparing unit, be used for according to comparison result, obtain testing result, the scale-of-two watermark information that arrives with Detection and Extraction, if comparison result is different for default original binary watermark information and the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is distorted; If comparison result is identical for default original binary watermark information and the scale-of-two watermark information that extracts, then testing result is that embedded scale-of-two watermark information is not distorted.
12. a system for processing watermark is characterized in that, comprising:
The watermark flush mounting, be used to choose curved transition and do not surpass the neighboring region of default change threshold as zone to be embedded, watermark information to be embedded is converted to the scale-of-two watermark information, and described scale-of-two watermark information is embedded in the zone to be embedded, the number in the zone to be embedded of described embedding scale-of-two watermark information is identical with the watermark information length of described scale-of-two watermark information, and each zone to be embedded only embeds data of scale-of-two watermark information;
The watermark detection device is used to extract the embedded scale-of-two watermark information in zone to be embedded, and according to default original watermark information, the scale-of-two watermark information that Detection and Extraction are arrived.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010511725 CN101980280A (en) | 2010-10-12 | 2010-10-12 | Watermark embedding method, detection method, devices and watermark processing system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010511725 CN101980280A (en) | 2010-10-12 | 2010-10-12 | Watermark embedding method, detection method, devices and watermark processing system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101980280A true CN101980280A (en) | 2011-02-23 |
Family
ID=43600781
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010511725 Pending CN101980280A (en) | 2010-10-12 | 2010-10-12 | Watermark embedding method, detection method, devices and watermark processing system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101980280A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102314667A (en) * | 2011-08-31 | 2012-01-11 | 辽宁大学 | Vertex weight value-based OBJ (object)-format three-dimensional model digital-watermarking method |
| CN106599631A (en) * | 2016-12-14 | 2017-04-26 | 吉林大学 | Function call-based software watermark information processing method |
| WO2020000486A1 (en) * | 2018-06-30 | 2020-01-02 | 华为技术有限公司 | Data processing method and device |
| CN111404925A (en) * | 2020-03-12 | 2020-07-10 | 北京航空航天大学 | Vehicle-mounted CAN bus data encryption method based on dynamic digital watermarking |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030223099A1 (en) * | 2002-05-29 | 2003-12-04 | Xerox Corporation | Method of detecting changes occurring in image editing using watermarks |
| JP2005269244A (en) * | 2004-03-18 | 2005-09-29 | Ricoh Co Ltd | Information embedding device and information embedding method, information extracting device and information extracting method, information embedding/extracting system, program, and recording medium |
| CN101533509A (en) * | 2009-03-23 | 2009-09-16 | 福建师范大学 | A three-dimensional grid splitting method of blind watermark |
-
2010
- 2010-10-12 CN CN 201010511725 patent/CN101980280A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030223099A1 (en) * | 2002-05-29 | 2003-12-04 | Xerox Corporation | Method of detecting changes occurring in image editing using watermarks |
| JP2005269244A (en) * | 2004-03-18 | 2005-09-29 | Ricoh Co Ltd | Information embedding device and information embedding method, information extracting device and information extracting method, information embedding/extracting system, program, and recording medium |
| CN101533509A (en) * | 2009-03-23 | 2009-09-16 | 福建师范大学 | A three-dimensional grid splitting method of blind watermark |
Non-Patent Citations (1)
| Title |
|---|
| 《山东大学硕士学位论文》 20100430 韩大志 基于视觉特征的三维网格数字水印算法 正文第17,20-24页 1-12 , 2 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102314667A (en) * | 2011-08-31 | 2012-01-11 | 辽宁大学 | Vertex weight value-based OBJ (object)-format three-dimensional model digital-watermarking method |
| CN106599631A (en) * | 2016-12-14 | 2017-04-26 | 吉林大学 | Function call-based software watermark information processing method |
| WO2020000486A1 (en) * | 2018-06-30 | 2020-01-02 | 华为技术有限公司 | Data processing method and device |
| CN110770725A (en) * | 2018-06-30 | 2020-02-07 | 华为技术有限公司 | Data processing method and device |
| CN110770725B (en) * | 2018-06-30 | 2022-05-31 | 华为云计算技术有限公司 | Data processing method and device |
| CN111404925A (en) * | 2020-03-12 | 2020-07-10 | 北京航空航天大学 | Vehicle-mounted CAN bus data encryption method based on dynamic digital watermarking |
| CN111404925B (en) * | 2020-03-12 | 2021-05-11 | 北京航空航天大学 | Vehicle-mounted CAN bus data encryption method based on dynamic digital watermarking |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zehner et al. | Visualization of gridded scalar data with uncertainty in geosciences | |
| CN101246586B (en) | Vector map watermarking method based on curve segmentation | |
| Marsh et al. | Multi-objective unstructured triangular mesh generation for use in hydrological and land surface models | |
| CN102982500B (en) | Based on vector map watermark embedding grammar and the detection method of virtual grid | |
| CN101980280A (en) | Watermark embedding method, detection method, devices and watermark processing system | |
| CN101622858A (en) | Image processing method and image checking method | |
| Mara et al. | Vectorization of 3D-characters by integral invariant filtering of high-resolution triangular meshes | |
| CN102622721A (en) | Three-dimensional grid model blind-watermarking method based on depth image mapping | |
| CN101692288A (en) | Digital watermark embedding and detecting method of CAD model indicated on basis of NURBS | |
| CN111680459A (en) | Pipeline leakage analysis method and device and computer readable storage medium | |
| CN103247017A (en) | Mesh segmentation-based non-blind watermark realization method for three-dimensional mesh model | |
| Wang et al. | Feature detection of triangular meshes via neighbor supporting | |
| Liu et al. | Zero-watermarking method for resisting rotation attacks in 3D models | |
| CN103198448B (en) | Threedimensional model digital watermark embedding based on vertex curvature and blind checking method | |
| CN108876694B (en) | Three-dimensional model blind digital watermarking algorithm based on Schur decomposition | |
| CN103377455A (en) | Three-dimensional geographic model digital watermarking method with copyright protection service orientation | |
| Han et al. | A trace map comparison algorithm for the discrete fracture network models of rock masses | |
| CN102314667B (en) | Vertex weight value-based OBJ (object)-format three-dimensional model digital-watermarking method | |
| Kim et al. | The perspective silhouette of a canal surface | |
| CN103279915B (en) | Based on the figure watermark embedding method of virtual pattern bar code | |
| Liu et al. | An automatic graph-based method for characterizing multichannel networks | |
| CN103310134A (en) | Vector data watermark anti-counterfeiting method based on geographical semantics support | |
| CN103886540B (en) | Ellipse figure characteristic digital fingerprint embedding and detection method | |
| CN106709855B (en) | 3D model watermarking method based on Beamlet line feature positioning | |
| Kambesis et al. | Morphometric analysis of cave patterns using fractal indices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20110223 |