CN109919821A - Insertion, extracting method and the storage medium of the dual blind watermatking of three-dimensional digital model - Google Patents
Insertion, extracting method and the storage medium of the dual blind watermatking of three-dimensional digital model Download PDFInfo
- Publication number
- CN109919821A CN109919821A CN201711322951.0A CN201711322951A CN109919821A CN 109919821 A CN109919821 A CN 109919821A CN 201711322951 A CN201711322951 A CN 201711322951A CN 109919821 A CN109919821 A CN 109919821A
- Authority
- CN
- China
- Prior art keywords
- vertex
- neighbours
- watermark
- heavy water
- value
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Architecture (AREA)
- Computer Graphics (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Image Processing (AREA)
- Editing Of Facsimile Originals (AREA)
- Processing Or Creating Images (AREA)
Abstract
The present invention provides insertion, extracting method and the storage medium of a kind of dual blind watermatking of three-dimensional digital model, and embedding grammar includes: to print insertion conditional filtering first kind vertex according to the first heavy water from three-dimensional digital model；Obtain the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print；The position on corresponding first kind vertex is adjusted by pixel characteristic attribute in three-dimensional digital model；Screening meets the second class vertex of preset condition from three-dimensional digital model adjusted；From the two ring neighbours vertex on the second class vertex, the two ring neighbours vertex for meeting the second heavy water print insertion condition are selected；Obtain the watermark sequence of values for being used as the second heavy water print；Watermark sequence of values includes the watermark value at each numerical digit；When watermark value at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit mismatches, the position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value.
Description
Technical field
The present invention relates to three-dimensional digital model digital watermark fields, more particularly to a kind of dual blind water of three-dimensional digital model
Insertion, extracting method and the storage medium of print.
Background technique
With the development of 3 dimension imaging technology, shown especially with computer graphics, virtual reality, 3D game, 3D
Deng development, three-dimensional digital model is increasingly becoming a kind of important Digital Media form.For illegally accounting for for three-dimensional digital model
The behaviors such as have, usurp, modifying, propagating more and more thus extremely urgent for the intellectual property protection of three-dimensional digital model.Three
Dimension word model digital watermark is the important technical for protecting the intellectual property of three-dimensional digital model.Three-dimensional digital model
Whether need archetype that can be divided into blind watermatking and non-blind watermatking when watermark is according to detection watermark, the latter needs when detecting watermark
Archetype, and the former does not need then, under the conditions of current three-dimensional digital model retrieval technique, blind watermatking technology has more
Actual application value.Three-dimensional digital model watermark is divided into again by content can examine watermark, Readable Watermarking, visual watermark.Wherein, may be used
The testing result for examining watermark is the binary sequence of 0 or 1 composition, and the testing result of Readable Watermarking is character string, the inspection of visual watermark
Survey result is image.
Compared with image watermark, audio frequency watermark and video watermark, the carrier of three-dimensional digital model watermark is different.3-dimensional digital
The vertex of model is irregular sampling, and what is do not determined puts in order, because handling without direct mathematical tool it,
It is big so as to cause watermark insertion difficulty.The non-blind watermatking of three-dimensional digital model can be aligned mentioning for progress watermark by means of resampling and again
It takes, and three-dimensional digital model blind watermatking then depends on itself watermark carrier to carry out watermark extracting, thus the blind water of three-dimensional digital model
The insertion difficulty and extraction difficulty of print are bigger, this is also one of the reason of most three-dimensional digital model watermarks is using non-blind watermatking.
Although the research of three-dimensional digital model blind watermatking is very difficult under the conditions of current three-dimensional digital model retrieval technique, have
Important practical application value.It is also more and more for the attack pattern of watermark with the continuous development of digital watermark, how
Guarantee that the robustness for guaranteeing that it resists attack under the invisibility of watermark is still a Research Challenges.Current three-dimensional digital model
Watermark is difficult to still remain or be properly detected after various attacks mode is attacked, thus the robustness for resisting attack is low.
Summary of the invention
Based on this, it is necessary to for the low technical problem of the robustness of current three-dimensional digital model blind watermatking, provide one kind
Insertion, extracting method and the storage medium of the dual blind watermatking of three-dimensional digital model.
A kind of embedding grammar of the dual blind watermatking of three-dimensional digital model, which comprises
Obtain three-dimensional digital model；
Insertion conditional filtering first kind vertex is printed according to the first heavy water from the three-dimensional digital model；
Obtain the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print；
The position on corresponding first kind vertex is adjusted by the pixel characteristic attribute in the three-dimensional digital model；
Screening meets the second class vertex of preset condition from three-dimensional digital model adjusted；The preset condition includes
Insertion condition is printed in first heavy water；
From the two ring neighbours vertex on second class vertex, the two ring neighbours for meeting the second heavy water print insertion condition are selected
Vertex；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to print according to first heavy water and be embedded in
Vertex referenced when first kind vertex described in conditional filtering；
Obtain the watermark sequence of values for being used as the second heavy water print；The watermark sequence of values includes the water at each numerical digit
Print value；
Watermark value at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit is not
The position on the corresponding two rings neighbours vertex of the numerical digit is adjusted to match with the watermark value by timing.
A kind of flush mounting of the dual blind watermatking of three-dimensional digital model, described device include:
Module is obtained, for obtaining three-dimensional digital model；
Screening module, for printing insertion conditional filtering first kind top according to the first heavy water from the three-dimensional digital model
Point；
Module is obtained, is also used to obtain the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print；
First heavy watermark embedding module, for adjusting phase by the pixel characteristic attribute in the three-dimensional digital model
Answer the position on first kind vertex；
Screening module is also used to screen the second class vertex for meeting preset condition from three-dimensional digital model adjusted；
The preset condition is included in the first heavy water print insertion condition；
Screening module is also used to from the two ring neighbours vertex on second class vertex, select meet the second heavy water print it is embedding
Enter two ring neighbours vertex of condition；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to according to institute
State vertex referenced when first kind vertex described in the first heavy water print insertion conditional filtering；
Module is obtained, the watermark sequence of values for being used as the second heavy water print is also obtained；The watermark sequence of values includes being in
Watermark value at each numerical digit；
Second heavy watermark embedding module, for when the position of picking out, two ring neighbours corresponding with numerical digit vertex
When being mismatched with the watermark value at corresponding numerical digit, the position on the corresponding two rings neighbours vertex of the numerical digit is adjusted to and the water
The matching of print value.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the calculating
When machine program is executed by the processor, so that the processor executes the embedding grammar of the dual blind watermatking of three-dimensional digital model
Step.
A kind of computer readable storage medium is stored with computer program, when the computer program is executed by processor,
So that the processor executes the step of embedding grammar of the dual blind watermatking of three-dimensional digital model.
Embedding grammar, computer equipment and the storage medium of the above-mentioned dual blind watermatking of three-dimensional digital model, by according to
The corresponding position for adjusting the first kind vertex filtered out of pixel point feature for making the two dimensional image of the first heavy water print, by the first heavy water
In print insertion three-dimensional digital model.For the three-dimensional digital model after insertion the first heavy water print, filter out as the second weight of insertion
The two ring neighbours vertex on the vertex of watermark, according to the two ring neighbours vertex filtered out and the watermark numerical value sequence printed as the second heavy water
Second heavy water is printed insertion 3-dimensional digital mould by the corresponding relationship of column, the position on the two ring neighbours vertex that corresponding adjustment filters out
In type, to be embedded in dual blind watermatking in three-dimensional digital model, the robustness of watermark in three-dimensional digital model is improved.It is logical
The restrict for crossing the second heavy water print insertion condition ensure that the first heavy water print being embedded in three-dimensional digital model and the second heavy water
Print is mutually indepedent, does not interfere with each other, further improves the robustness of watermark.
A kind of extracting method of the dual blind watermatking of three-dimensional digital model, which comprises
Obtain three-dimensional digital model；
Insertion conditional filtering first kind vertex is printed according to the first heavy water from the three-dimensional digital model；
Corresponding pixel characteristic attribute is determined according to the position on the first kind vertex；
The two dimensional image as the first heavy water print is restored according to the pixel characteristic attribute；
Screening meets the second class vertex of preset condition from the three-dimensional digital model；The preset condition is included in institute
State the first heavy water print insertion condition；
From the two ring neighbours vertex on second class vertex, the two ring neighbours for meeting the second heavy water print insertion condition are selected
Vertex；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to print according to first heavy water and be embedded in
Vertex referenced when first kind vertex described in conditional filtering；
According to the position on the two ring neighbours vertex picked out matched watermark value and the two ring neighbours vertex picked out
Corresponding numerical digit restores the watermark sequence of values as the second heavy water print.
A kind of extraction element of the dual blind watermatking of three-dimensional digital model, described device include:
Module is obtained, for obtaining three-dimensional digital model；
Screening module, for printing insertion conditional filtering first kind top according to the first heavy water from the three-dimensional digital model
Point；
First heavy water prints extraction module, for determining corresponding pixel point feature category according to the position on the first kind vertex
Property；
First heavy water prints extraction module, for restoring two as the first heavy water print according to the pixel characteristic attribute
Tie up image；
Screening module is also also used to screen the second class vertex for meeting preset condition from the three-dimensional digital model；Institute
It states preset condition and is included in the first heavy water print insertion condition；
Screening module is also used to from the two ring neighbours vertex on second class vertex, select meet the second heavy water print it is embedding
Enter two ring neighbours vertex of condition；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to according to institute
State vertex referenced when first kind vertex described in the first heavy water print insertion conditional filtering；
Second heavy water prints extraction module, for according to the matched watermark value of position institute on the two ring neighbours vertex picked out with
And the two corresponding numerical digits in ring neighbours vertex picked out, restore the watermark sequence of values as the second heavy water print.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the calculating
When machine program is executed by the processor, so that the processor executes the extracting method of the dual blind watermatking of three-dimensional digital model
Step.
A kind of computer readable storage medium is stored with computer program, when the computer program is executed by processor,
So that the processor executes the step of extracting method of the dual blind watermatking of three-dimensional digital model.
Extracting method, computer equipment and the storage medium of the above-mentioned dual blind watermatking of three-dimensional digital model, according to the first weight
Watermark insertion condition and preset condition filter out corresponding first kind vertex and the second class vertex respectively from three-dimensional digital model,
Two dimensional image is restored according to the first kind vertex filtered out, i.e., extracts the first heavy water print from three-dimensional digital model.From screening
The second class vertex in screening meet two ring neighbours vertex of the second heavy water print insertion condition, and according to the two ring neighbours filtered out
Vertex restores corresponding watermark sequence of values, i.e., the second heavy water print is extracted from three-dimensional digital model, to realize dual
The extraction of watermark improves the robustness of watermark in three-dimensional digital model.In doubly time series model extraction process, independently of each other, and not
It is related to being not embedded into the three-dimensional digital model of watermark, further improves the Shandong that attack is resisted in embedded watermark in three-dimensional digital model
Stick.
Detailed description of the invention
Fig. 1 is the schematic diagram of internal structure of computer equipment in one embodiment；
Fig. 2 is the flow diagram of the embedding grammar of the dual blind watermatking of three-dimensional digital model in one embodiment；
Fig. 3 is the flow diagram of the embedding grammar of the dual blind watermatking of three-dimensional digital model in another embodiment；
Fig. 4 is the flow diagram of the extracting method of the dual blind watermatking of three-dimensional digital model in one embodiment；
Fig. 5 is the flow diagram of the extracting method of the dual blind watermatking of three-dimensional digital model in another embodiment；
Fig. 6 is the insertion of the dual blind watermatking of three-dimensional digital model and the flow diagram of extracting method in one embodiment；
Fig. 7 is the structural block diagram of the flush mounting of the dual blind watermatking of three-dimensional digital model in one embodiment；
Fig. 8 is the structural block diagram of the extraction element of the dual blind watermatking of three-dimensional digital model in one embodiment；
Fig. 9 is the structural block diagram of the extraction element of the dual blind watermatking of three-dimensional digital model in another embodiment.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
As shown in Figure 1, in one embodiment, providing a kind of schematic diagram of internal structure of computer equipment.The calculating
Machine equipment includes the processor and memory connected by system bus.Wherein, memory include non-volatile memory medium and
Built-in storage.It is stored with computer program in the memory of the computer equipment, it, can when which is executed by processor
So that processor realizes the insertion and extracting method of a kind of dual blind watermatking of three-dimensional digital model.The interior storage of the computer equipment
Device provides environment for the operation of the computer program in non-volatile memory medium.The processor calculates and controls energy for providing
Power supports the operation of entire computer equipment.The computer equipment can also include network interface, input unit and display screen.
The network interface of the computer equipment such as receives three-dimensional digital model, two dimensional image for carrying out network communication with outside accordingly
Or watermark sequence of values.The display screen of computer equipment can be liquid crystal display or electric ink display screen etc., input dress
It sets and can be the touch layer covered on display screen, be also possible to the button being arranged on computer equipment shell, trace ball or touch-control
Plate is also possible to external keyboard, Trackpad or mouse etc..The display screen of computer equipment can be used for showing before insertion watermark
Three-dimensional digital model later and the two dimensional image as the first heavy water print.The input unit of computer equipment can be used for
Input is used as the watermark sequence of values of the second heavy water print, and insertion or the trigger action instruction for extracting watermark.
It will be understood by those skilled in the art that structure shown in Fig. 1, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
As shown in Fig. 2, in one embodiment, providing a kind of embedding grammar of dual blind watermatking of three-dimensional digital model.
The present embodiment is applied to the computer equipment in above-mentioned Fig. 1 in this way to illustrate.This method specifically comprises the following steps:
S202 obtains three-dimensional digital model.
Wherein, three-dimensional digital model is the three-dimensional model connected and composed by multiple vertex.Three-dimensional digital model specifically may be used
Being obtained by surface sweeping three-dimension object, it is also possible to be obtained according to data parameters by three-dimensional modeling.Specifically, it calculates
Machine equipment can also obtain three-dimensional from middle acquisition three-dimensional digital model is locally stored from other computer equipments or storage equipment
Mathematical model.
S204 prints insertion conditional filtering first kind vertex according to the first heavy water from three-dimensional digital model.
Wherein, watermark is the information added in the carrier, for realizing the copyright protection of carrier.First heavy water print is used for area
Divide other watermark informations of insertion carrier.First heavy water print insertion condition is used to screen the vertex that can be embedded in the first heavy water print.
First heavy water print insertion condition specifically can be the first kind top that the screening from three-dimensional digital model is used as insertion the first heavy water print
The preset condition of point.First kind vertex representation meets the vertex of the first heavy water print insertion condition, is unsatisfactory for for distinguishing over other
The vertex of the condition.First kind vertex specifically can be in three-dimensional digital model for being embedded in the vertex of the first heavy water print.
Specifically, computer equipment gets three-dimensional digital model, prints insertion condition from institute according to preset first heavy water
First kind vertex is screened in the three-dimensional digital model of acquisition, the first kind vertex filtered out meets the first heavy water print insertion condition,
And the first kind vertex of screening is used as to the vertex of insertion the first heavy water print.
In one embodiment, computer equipment obtains three-dimensional digital model, and obtains for from the three-dimensional digital model
First heavy water on middle screening first kind vertex prints insertion condition, prints insertion condition from three dimensions further according to the first acquired heavy water
Corresponding first kind vertex is screened in word model.
S206 obtains the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print.
Wherein, two dimensional image is the flat image not comprising depth information.Two dimensional image specifically can be bianry image, than
Such as say that content is the bianry image of letter.Pixel characteristic attribute is used to indicate the attributive character of pixel.Pixel point feature category
Property specifically can be the corresponding model attribute of each pixel in two dimensional image.Pixel characteristic attribute may include the row rope of pixel
Draw value, column index value and corresponding pixel value.Ranks index value is used to indicate the row and column at pixel place in two dimensional image.
Specifically, computer equipment obtains two dimensional image, pre-processes acquired two dimensional image, obtains pretreated two
Tie up the pixel characteristic attribute of each pixel of image.Acquired two dimensional image is used as first in insertion three-dimensional digital model
Heavy water print.Corresponding two dimensional image is embedded in three-dimensional digital model according to pixel characteristic attribute obtained by computer equipment
In.
In one embodiment, computer equipment can obtain the two dimensional image of default size.Computer equipment obtains three-dimensional
The sequence and source of mathematical model and two dimensional image are unlimited.
S208 is adjusted the position on corresponding first kind vertex in three-dimensional digital model by pixel characteristic attribute.
Wherein, each vertex in three-dimensional digital model has opposite positional relationship.It is corresponding in three-dimensional digital model
In global coordinate system, each vertex in three-dimensional digital model has determining position.
Specifically, computer equipment obtains the first kind vertex for meeting the first heavy water print insertion condition, and is used as first
The pixel characteristic attribute of the two dimensional image of heavy water print, according to the pixel point feature on acquired first kind vertex and two dimensional image
Attribute determines the corresponding first kind vertex of each pixel of two dimensional image respectively.Computer equipment in three-dimensional digital model,
The position for corresponding to the first kind vertex of the pixel according to the adjustment of the pixel characteristic attribute of pixel, two dimensional image is made
It is printed in insertion three-dimensional digital model for the first heavy water.
In one embodiment, computer equipment is according to each pixel in multiple first kind vertex of acquisition and two dimensional image
Pixel characteristic attribute, determine the corresponding first kind vertex of each pixel in two dimensional image.When the first kind vertex of acquisition
When number is greater than the pixel number of two dimensional image, first kind vertex corresponding with the pixel in two dimensional image is according to corresponding
Pixel characteristic attribute adjusts position, does not determine that the position on the first kind vertex of corresponding relationship is protected with the pixel in two dimensional image
It holds constant.By the way that the pixel characteristic attribute of each pixel of two dimensional image is respectively embedded into corresponding in three-dimensional digital model
In a kind of vertex.
S210, screening meets the second class vertex of preset condition from three-dimensional digital model adjusted；Preset condition packet
It includes and prints insertion condition in the first heavy water.
Wherein, the second class vertex representation meets the vertex of preset condition, for distinguishing other tops for not meeting preset condition
Point.Preset condition is preset for screening the condition on vertex.Preset condition is included in the first heavy water print insertion condition, table
The screening conditions for showing that preset condition includes are relatively fewer, thus include first kind vertex in the second class vertex.
Specifically, computer equipment by adjusting first kind vertex position, using acquired two dimensional image as first
In heavy water print insertion three-dimensional digital model, to obtain the three-dimensional digital model of insertion the first heavy water print.Computer equipment obtains
The preset condition for screening the second class vertex, screening meets the of preset condition from the three-dimensional digital model that the first heavy water of insertion prints
Two class vertex, the second class vertex filtered out include the first kind vertex that the first heavy water print is embedded in three-dimensional digital model.
S212 selects the two rings neighbour for meeting the second heavy water print insertion condition from the two ring neighbours vertex on the second class vertex
Occupy vertex；The two ring neighbours vertex picked out are not belonging to first kind vertex, and are not belonging to print insertion condition sieve according to the first heavy water
Referenced vertex when selecting first kind vertex.
Wherein, two ring neighbours vertex are the vertex being connected by a ring neighbours vertex on vertex with corresponding vertex.One
Ring neighbours vertex is the vertex being connected directly with vertex.One ring neighbours vertex and two ring neighbours vertex are all in three-dimensional digital model
Vertex, a ring neighbours vertex and two ring neighbours vertex are relative concepts, and a ring neighbours vertex and two ring neighbours vertex also have respectively
From a ring neighbours vertex and two ring neighbours vertex.Each vertex can have multiple ring neighbours vertex and multiple two rings neighbours top
Point.Vertex referenced when being embedded in conditional filtering first kind vertex is printed according to the first heavy water, is the first kind vertex filtered out
One ring neighbours vertex.
Specifically, computer equipment obtains the second heavy water print insertion condition, according to acquired the second heavy water print insertion item
Part is selected from the two ring neighbours vertex on the second class vertex for meeting preset condition and meets the two of the second heavy water print insertion condition
Ring neighbours vertex, the two ring neighbours vertex picked out are used as the vertex of insertion the second heavy water print.Insertion item is printed by the second heavy water
The restrict of part, from the two ring neighbours vertex on the second class vertex, the two ring neighbours vertex selected are not belonging to computer equipment
It is embedded in the first kind vertex of the first heavy water print, is also not belonging to a ring neighbours vertex on the first kind vertex of insertion the first heavy water print.
S214 obtains the watermark sequence of values for being used as the second heavy water print；Watermark sequence of values includes at each numerical digit
Watermark value.
Wherein, watermark sequence of values is the sequence being composed of multiple numerical value ordered arrangements.Watermark sequence of values is specific
The sequence that can be binary numeral composition is also possible to the sequence of other binary values composition.Watermark sequence of values specifically may be used
To be binary sequence, such as " 10001101 ", it is also possible to the sequence being made of other numerical value, such as " 36636633 ".
It constitutes the position of the numerical value of watermark sequence of values in the sequence and is known as numerical digit, numerical digit is corresponding with the numerical value at numerical digit.The application
The middle numerical value for constituting watermark sequence of values can be described as watermark value.
Specifically, computer equipment obtains watermark sequence of values, and acquired watermark sequence of values is used as insertion and has been inserted into
The second heavy water print of the three-dimensional digital model of first heavy water print.Computer equipment can be from being locally stored middle acquisition watermark numerical value sequence
Column can also obtain watermark sequence of values from other storage equipment, can also generate instruction according to watermark sequence of values and generate phase
The watermark sequence of values answered.
In one embodiment, computer equipment is according to the two ring neighbours for meeting the second heavy water print insertion condition picked out
Vertex and acquired watermark sequence of values determine each number in the multiple two rings neighbours vertex and watermark sequence of values picked out
The corresponding relationship of position, to print insertion 3-dimensional digital for watermark sequence of values as the second heavy water according to identified corresponding relationship
In model.
S216, the watermark value at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit is not
When matching, the position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value.
Specifically, computer equipment obtains the position on the two ring neighbours vertex picked out, according to the two ring neighbours picked out
Corresponding relationship between vertex and each numerical digit of acquired watermark sequence of values determines the two ring neighbours' vertex correspondences picked out
The watermark value of numerical digit and the numerical digit.Two ring will be corresponded in the position on two acquired ring neighbours vertex and watermark sequence of values
The watermark value of the numerical digit on neighbours vertex matches, the position when two ring neighbours vertex and the watermark corresponding to the two rings neighbours vertex
When value mismatches, corresponding two rings neighbours vertex is adjusted according to the numerical digit of two ring neighbours' vertex correspondences is corresponding, so that after adjustment
The position on two ring neighbours vertex match with the watermark value at corresponding numerical digit.
The embedding grammar of the above-mentioned dual blind watermatking of three-dimensional digital model, by according to the two dimensional image for being used as the first heavy water print
The position on first kind vertex that filters out of the corresponding adjustment of pixel point feature, the first heavy water is printed in insertion three-dimensional digital model.
For the three-dimensional digital model after insertion the first heavy water print, the two ring neighbours top on the vertex as insertion the second heavy water print is filtered out
Point, according to the corresponding relationship on the two ring neighbours vertex filtered out and the watermark sequence of values printed as the second heavy water, corresponding adjustment
The position on the two ring neighbours vertex filtered out prints the second heavy water in insertion three-dimensional digital model, thus in 3-dimensional digital mould
It is embedded in dual blind watermatking in type, improves the robustness of watermark in three-dimensional digital model.Insertion condition is printed by the second heavy water
Restrict, ensure that the first heavy water print being embedded in three-dimensional digital model and the second heavy water print are mutually indepedent, do not interfere with each other,
Further improve the robustness of watermark.
In one embodiment, in the embedding grammar of the dual blind watermatking of above-mentioned three-dimensional digital model, step S204 includes: root
According to a ring neighbours vertex on vertex each in three-dimensional digital model, the ring neighbours' center of mass point for calculating each vertex is respectively corresponded；
According to ring neighbours' center of mass point on a ring neighbours vertex on each vertex and each vertex, from three-dimensional digital model, screening symbol
It closes preset condition and is not mutually the first kind vertex on neighbours vertex.
Wherein, ring neighbours' center of mass point is the center of mass point on a ring neighbours vertex on vertex.One ring neighbours' center of mass point specifically may be used
To be the center of mass point calculated according to multiple ring neighbours vertex on vertex.Neighbours vertex is used to indicate two adjacent tops
Point.Preset condition is preset for screening the condition on vertex.
Specifically, for each vertex in acquired three-dimensional digital model, computer equipment is respectively according to each top
One ring neighbours vertex of point calculates ring neighbours' center of mass point on the vertex, and according to a ring neighbours vertex on each vertex and calculating
Resulting ring neighbours' center of mass point screens first kind vertex from three-dimensional digital model.Computer equipment is from three-dimensional digital model
The first kind vertex of screening meets preset condition, and the first kind vertex filtered out is not mutually neighbours vertex.
In one embodiment, computer equipment establishes global coordinate system to acquired three-dimensional digital model, and obtains
Coordinate value of each vertex in global spherical coordinate system in three-dimensional digital model, acquired coordinate value include abscissa (X), indulge
Coordinate (Y) and ordinate (Z).For each vertex in three-dimensional digital model, computer equipment obtains each vertex respectively
The total quantity on one ring neighbours vertex and the respective coordinate value in multiple ring neighbours vertex on the vertex, belong to one for acquired
The coordinate value on the one ring neighbours vertex on a vertex respectively corresponds summation, then the resulting coordinate value that will sum is according to corresponding ring neighbours
The total quantity on vertex is averaging, to obtain the coordinate value of ring neighbours' center of mass point on the vertex.
In above-described embodiment, according to a ring neighbours vertex on each vertex and corresponding ring neighbours' center of mass point, from three-dimensional
The first kind vertex for being embedded in the first heavy water print is filtered out in mathematical model, the first kind vertex filtered out meets default item
Part and be not mutually neighbours vertex, so that it is mutually indepedent between each first kind vertex of insertion the first heavy water print, to improve the first weight
The robustness of watermark.
In one embodiment, screening meet preset condition and mutually be neighbours vertex the first kind vertex the step of packet
It includes: according to a ring neighbours center of mass point on vertex each in three-dimensional digital model and a ring neighbours vertex, according to two kinds of normal direction meters
Calculation mode calculates separately two normal vectors of ring neighbours' center of mass point；First kind vertex is selected from three-dimensional digital model；Selection
First kind vertex a ring neighbours vertex and the distance between ring neighbours' center of mass point maxima and minima ratio it is small
Normal vector angle between default fractional threshold, two normal vectors of ring neighbours' center of mass point on the first kind vertex of selection is big
In default angle threshold value, and the first kind vertex selected is not mutually neighbours vertex.
Wherein, the normal vector of point can be being averaged for the normal vector in face belonging to the point.Default fractional threshold is to preset
For threshold value compared with distance ratio, such as 4.Default angle threshold value is preset is used for and two normal vectors
Angle compare, so that screening meets the threshold value on the vertex of condition, such as 0.1.
Specifically, for each vertex in three-dimensional digital model, computer equipment is according to the ring neighbours on each vertex
Vertex calculates separately ring neighbours' center of mass point of respective vertices, and according to a ring neighbours vertex on each vertex and ring neighbours' matter
Heart point calculates separately two normal vectors of ring neighbours' center of mass point on each vertex according to two kinds of normal vector calculations.It calculates
Machine equipment is according to two methods on a ring neighbours vertex on each vertex and a ring neighbours center of mass point and ring neighbours' center of mass point
Vector, selection meets the first kind vertex of the first heavy water print insertion condition from three-dimensional digital model.
Further, three conditions in first kind vertex while satisfaction that computer equipment filters out:
(1) maxima and minima of the distance between a ring neighbours vertex on first kind vertex and ring neighbours' center of mass point
Ratio be less than default fractional threshold.
(2) the normal vector angle between two normal vectors of ring neighbours' center of mass point on first kind vertex is greater than default angle
Threshold value.
(3) first kind vertex is not mutually neighbours vertex.
It, can be with when computer equipment screens from three-dimensional digital model while meeting the first kind vertex of above three condition
It is screened by arbitrary built-up sequence, such as can first select the vertex for meeting condition (1), then choose from the vertex of selection
Be full the vertex of sufficient condition (2), finally screened from the vertex selected mutually for neighbours vertex vertex as the filtered out
A kind of vertex.
In one embodiment, if not considering the attribute information on each vertex in three-dimensional digital model, three-dimensional digital model M
It is represented by M={ Vm,Km, wherein Vm={ 0,1,2 ... m-1 } indicate that the vertex set of three-dimensional digital model M, m indicate top
The number of point, KmIt is the set of the topological connection relation between each vertex and each vertex of M, set KmIn element can
Including 3 seed types: vertex Vi={ i }, side Ei,j={ i, j }, face Fi,j,k={ i, j, k }.If side { i, j } ∈ Km, then vertex { i }
{ j } mutually claims neighbours vertex.The one ring neighbours vertex on vertex { i } may be defined as N1(i)=j | { i, j } ∈ Km, vertex { i's }
Two ring neighbours vertex N2It (i) is a ring neighbours vertex N1(i) a ring neighbours vertex.
For three-dimensional digital model M, computer equipment is according to each vertex ViA ring neighbours vertex N1(i) it calculates separately
Corresponding ring neighbours' center of mass point Zi, then by ring neighbours' center of mass point ZiWith correspond to the same vertex ViEach ring neighbours
Vertex N1(i) it is connected, obtains multiple with ring neighbours' center of mass point ZiFor the triangle of common vertex, it is denoted as T (Zi), triangle
Number be denoted as N.The normal vector of triangle obtained is denoted as nij, with the normal vector n of N number of triangleijMean value as one
Ring neighbours' center of mass point ZiNormal vector ne.Specific formula for calculation is as follows:
T (Z is calculated separately againi) in each triangle area Sij, according to the normal vector n of each triangleijAnd triangle
Shape area SijSquare calculate another normal vector nt, specific formula for calculation is as follows:
It calculates separately and corresponds to each vertex ViEach ring neighbours vertex N1(i) He Yihuan neighbours' center of mass point ZiBetween
Distance, constitute corresponding distance set L (i).Computer equipment selects L from three-dimensional digital modelmax(i)/Lmin(i) small
In the vertex of default fractional threshold, wherein Lmax(i) and Lmin(i) maximum value and minimum value in distance set are respectively indicated.
Further, computer equipment calculates separately each vertex V selectediRing neighbours' center of mass point ZiTwo
Normal vector neAnd ntBetween normal vector angle Φ.From the vertex selected, selecting method vector angle Φ is greater than default angle threshold
The vertex of value.Wherein, angle threshold value custom size as needed is preset.Computer equipment is rejected from the vertex picked out
The vertex on neighbours vertex each other, filtering out mutually is not the vertex on neighbours vertex as first kind vertex.
In above-described embodiment, by the first heavy water print insertion condition be specifically divided into three conditions, based on three conditions from
Screening is used as the first kind vertex of insertion the first heavy water print in three-dimensional digital model.Based on three conditional filtering first kind tops
Point ensure that the invisibility of embedded watermark in first kind vertex, improves the accuracy of watermark insertion, effectively prevents water
The ambiguousness for printing insertion, to improve the robustness of watermark.
In one embodiment, the embedding grammar of the dual blind watermatking of above-mentioned three-dimensional digital model, further includes: the second class vertex
Middle exist and the two ring neighbours vertex corresponding second class vertex selected, and the two ring neighbours vertex selected only with corresponding second class
Two ring neighbours vertex on vertex are connected；The triangle that the two ring neighbours vertex and two ring neighbours vertex selected are constituted
In, base angle locating for two ring neighbours vertex is acute angle；The method of one ring neighbours' center of mass point on the two ring neighbours vertex selected
Vector angle is less than default angle threshold value；The respective ring neighbours' center of mass point in one ring neighbours vertex on the two ring neighbours vertex selected
Normal vector angle be less than default angle threshold value.
Specifically, computer equipment is filtered out from the three-dimensional digital model that the first heavy water of insertion prints meets preset condition
Second class vertex, the second class vertex filtered out include the first kind vertex for being embedded in the first heavy water print.Computer equipment from
In the two ring neighbours vertex on the second class vertex filtered out, the two ring neighbours vertex for meeting the second heavy water print insertion condition are selected.
The two ring neighbours vertex selected while meeting following four condition:
(1) two ring neighbours vertex is only connected with two ring neighbours vertex on corresponding second class vertex.Wherein, this two
Ring neighbours vertex is the two ring neighbours vertex on the second class vertex.
In the triangle that (2) two ring neighbours vertex and above-mentioned two ring neighbours vertex are constituted, two ring neighbours vertex
Locating base angle is acute angle.
The normal vector angle of one ring neighbours' center of mass point on (3) two ring neighbours vertex is less than default angle threshold value.Wherein, two ring
One ring neighbours' center of mass point on neighbours vertex is the ring neighbours' mass center calculated according to a ring neighbours vertex on the two rings neighbours vertex
Point.
The normal vector angle of the respective ring neighbours' center of mass point in one ring neighbours vertex on (4) two ring neighbours vertex is less than default
Angle threshold value.
In above-described embodiment, insertion the second heavy water print is used as based on preset condition and the second heavy water print insertion conditional filtering
Two ring neighbours vertex ensure that the part vertex in three-dimensional digital model is remained unchanged as watermark insertion invariant, improve
The robustness of embedded watermark.And the restrict that insertion condition is printed by the second heavy water, so that embedding in three-dimensional digital model
The the first heavy water print and the second heavy water print entered is mutually indepedent, does not interfere with each other, further improves the robustness of watermark.
In one embodiment, the embedding grammar of the dual blind watermatking of above-mentioned three-dimensional digital model, further includes: by predetermined angle
Range is divided into multiple angle groups；The number of angle group is equal with the numerical digit sum of watermark sequence of values；Each angle group difference
It is corresponding with each numerical digit of watermark sequence of values；Corresponding to the second class vertex on the two ring neighbours vertex picked out and is connecting
In the triangle that two ring neighbours vertex on two ring neighbours vertex and the second class vertex are constituted, according to locating for the second class vertex
Angle group belonging to apex angle determines the numerical digit in watermark sequence of values with two ring neighbours' vertex correspondences；According to watermark sequence of values
Numerical digit, determine the watermark value of two ring neighbours' vertex correspondences.
Wherein, predetermined angle range is preset angular range.Predetermined angle range specifically can be (0, π).Angle
Degree group is used for the angular range for indicating to be made of multiple angle values.
Specifically, computer equipment obtains the numerical digit sum of predetermined angle range and acquired watermark sequence of values,
And acquired predetermined angle range is divided into multiple angle groups, the angle group of acquisition by the numerical digit sum of watermark sequence of values
Number is equal with the numerical digit sum of watermark sequence of values, establishes each numerical digit of each angle group and watermark sequence of values respectively
One-to-one relationship.In three-dimensional digital model, computer equipment is by being connected to two ring neighbours vertex selected and corresponding
In the triangle that two ring neighbours vertex and the second class vertex between second class vertex are constituted, second class is obtained
The angle of apex angle locating for vertex, and judge angle group belonging to acquired angle.
Further, angle group and angle group belonging to computer equipment apex angle according to locating for the second class vertex
With the corresponding relationship of numerical digit each in watermark sequence of values, determine that the second class vertex is corresponding with the numerical digit in watermark sequence of values
Relationship, so that it is determined that the corresponding relationship in watermark sequence of values with the two ring neighbours vertex on the second class vertex.Wherein, with watermark
The numerical digit of sequence of values determines two ring neighbours vertex of corresponding relationship, is to meet the second heavy water print insertion condition and by above-mentioned
Constitute the two ring neighbours vertex that two ring neighbours vertex of triangle are connect with above-mentioned second class vertex.Computer equipment according to
Numerical digit and the corresponding relationship on the two ring neighbours vertex picked out in watermark sequence of values, determine the two ring neighbours picked out respectively
The watermark value of vertex correspondence, so as to according to the corresponding position for adjusting corresponding two rings neighbours vertex of the corresponding relationship.
In one embodiment, the will be met in each of three-dimensional digital model the second class vertex and the second class vertex
Two ring neighbours vertex of double watermark insertion condition, and connect two rings on the second class vertex He the two rings neighbours vertex
Neighbours vertex constitutes corresponding subelement.In each subelement, by the second class vertex and two ring neighbours vertex respectively with two
One ring neighbours vertex constitutes two triangles.It is according to the apex angle where the second class vertex in each subelement, multiple son is single
Member is divided into the quantity multiple unit groups equal with the numerical digit of watermark sequence of values sum, and the second class vertex in the triangle divided
Angle group belonging to the apex angle at place is determined for compliance with the second heavy water print respectively and does not determine multiple unit group and watermark sequence of values
In each numerical digit corresponding relationship.Each subelement in each unit group corresponds to the identical numerical digit in watermark sequence of values, right
Identical watermark value is embedded in two ring neighbours vertex in each unit group.
In one embodiment, the numerical digit sum of watermark sequence of values is w, in each subelement where the second class vertex
Apex angle range is (a π, b π), which is divided into w group, every group of corresponding angular range is
In above-described embodiment, the total equal angle group of numerical digit by obtaining quantity and watermark sequence of values, by the
Two class vertex, and connect the second class vertex and meet two of corresponding two rings neighbours vertex of the second heavy water print insertion condition
In the triangle that one ring neighbours vertex is constituted, determine that corresponding two ring is adjacent according to angle group belonging to apex angle locating for the second class vertex
Numerical value corresponding to vertex is occupied, to correspond to the position for adjusting corresponding two rings neighbours vertex according to the numerical value of the determination, thus will
In the orderly insertion three-dimensional digital model of watermark sequence of values, the order and accuracy of watermark insertion ensure that.
In one embodiment, in the embedding grammar of the dual blind watermatking of above-mentioned three-dimensional digital model, step S216 includes: root
According to projected position of the two ring neighbours vertex picked out on corresponding ring neighbours side, the number of two ring neighbours' vertex correspondences is determined
Value；One ring neighbours side is the line on two ring neighbours vertex corresponding to two ring neighbours vertex；One ring neighbours vertex is connection
The one ring neighbours vertex on two ring neighbours vertex and corresponding second class vertex；When the numerical value and watermark for corresponding to two ring neighbours vertex
When being worth unequal, projected position of the two rings neighbours vertex adjusted on a ring neighbours side is determined according to watermark value, according to throwing
The position on shadow position two ring neighbours vertex of corresponding adjustment.
Wherein, corresponding position when projected position is for indicating for be projected on a ring neighbours side on two ring neighbours vertex.Tool
Body, computer equipment picks out two rings for meeting the second heavy water print insertion condition from the two ring neighbours vertex on the second class vertex
Neighbours vertex, by the company on two ring neighbours vertex between the two ring neighbours vertex for being connected to the second class vertex and selecting
Line, as the one ring neighbours side of the second class vertex for corresponding to the two rings neighbours vertex.Computer equipment is adjacent by two rings selected
It occupies vertex to be projected on corresponding ring neighbours side, obtains homolographic projection position, and determine and select respectively according to projected position
The two corresponding numerical value in ring neighbours vertex.Computer equipment is according to water in the two ring neighbours vertex selected and watermark sequence of values
The corresponding relationship of print value respectively makes comparisons the numerical value for each of selecting two ring neighbours' vertex correspondences and corresponding watermark value, when
When numerical value and unequal corresponding watermark value, according to two ring neighbours vertex of the corresponding adjustment of watermark value on corresponding ring neighbours side
Projected position, according to the corresponding position for adjusting corresponding two rings neighbours vertex of projected position adjusted.
In one embodiment, when the numerical value and unequal corresponding watermark value on two ring neighbours vertex, computer equipment
The numerical value of two ring neighbours' vertex correspondences is replaced with into the watermark value corresponding to the two rings neighbours vertex, further according to replaced numerical value
Projected position of the two ring neighbours vertex on corresponding ring neighbours side is adjusted, according to projected position adjusted two rings of corresponding adjustment
Position of the neighbours vertex in three-dimensional digital model.
In one embodiment, computer equipment is according to the watermark value and two rings of the two ring neighbours' vertex correspondences selected
Whether projected position of the neighbours vertex on a ring neighbours side meets preset matching condition, determines the position on corresponding two rings neighbours vertex
The position for whether needing to adjust and adjust set.Watermark value and the two rings neighbours vertex corresponding to the two ring neighbours vertex are one
When projected position on ring neighbours side meets preset matching condition, keep the position on the two rings neighbours vertex constant, when being unsatisfactory for
When, projected position of the corresponding two rings neighbours vertex on a ring neighbours side is adjusted to default projected position according to watermark value, then
According to the position on two ring neighbours vertex of the corresponding adjustment of default projected position.
In one embodiment, the watermark value corresponding to two ring neighbours vertex D is indicated with α, watermark value α can be 0 or 1.
Two ring neighbours vertex indicate that the ring neighbours side being made of the line between A and C is denoted as AC, by two rings with A and C respectively
The projected position that neighbours vertex D is projected on a ring neighbours side AC is denoted as B.If corresponding to two ring neighbours vertex D watermark value α and
Projected position B meets formula simultaneously:
Or
Then keep the position of two ring neighbours vertex D constant.If watermark value α and projection position corresponding to two ring neighbours vertex D
It sets B while meeting formula:
Or
The position for then respectively corresponding two ring neighbours vertex D of adjustment, so that two rings neighbours vertex D adjusted is in a ring neighbours
Side is that the projected position B projected on AC meets formula respectively:
Or
The position adjustment of specific two rings neighbours vertex D are as follows: determine two rings neighbours adjusted according to watermark value α is corresponding
The formula that vertex D is met on a ring neighbours side by the projected position B on AC, determining projected position B location, by identified
The projection plane that projected position B and a ring neighbours are AC when AC is determined perpendicular to a ring neighbours, by corresponding two rings neighbours vertex D
Projection obtains corresponding projected position, using projected position obtained as adjusted to projection plane in projection plane
The position of two ring neighbours vertex D.
In above-described embodiment, according to projected position and watermark numerical value of the two ring neighbours vertex on corresponding ring neighbours side
Corresponding to the corresponding relationship between the watermark value on two ring neighbours vertex in sequence, the two ring neighbours top for needing to adjust position is determined
The position on point and two rings neighbours vertex adjusted, and the corresponding position for adjusting corresponding two rings neighbours vertex, thus by watermark
In the orderly insertion three-dimensional digital model of sequence of values.
In one embodiment, the step of obtaining the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print is wrapped
It includes: obtaining the two dimensional image for being used as the first heavy water print；Two dimensional image is converted into spectrum domain, obtains corresponding spectral matrix；It is right
Spectral matrix takes absolute value；Spectral matrix after taking absolute value according to default scramble condition scramble obtains and corresponds to spectral matrix
Two dimensional image pixel characteristic attribute.
Wherein, spectrum domain is for indicating transform domain corresponding with the airspace indicated with pixel.Spectrum domain specifically can be
The transform domain of information is indicated with frequency and amplitude.Spectral matrix is used to indicate the two-dimensional matrix in spectrum domain.It takes absolute value and is used for
The negative value of spectral matrix is become to the positive value of same size.
Specifically, computer equipment obtains the two dimensional image for the first heavy water print being used as in insertion three-dimensional digital model, will
Acquired two dimensional image is converted into spectrum domain according to default mapping mode, obtains corresponding spectral matrix, to frequency obtained
Spectrum matrix takes absolute value, the spectral matrix after taking absolute value according to default scramble condition scramble, obtains and corresponds to spectral matrix
The pixel characteristic attribute of two dimensional image.
In one embodiment, computer equipment carries out two-dimension discrete cosine transform to acquired two dimensional image, obtains
Corresponding spectral matrix, then Arnold scramble is carried out to the spectral matrix after taking absolute value, obtain each picture of respective two-dimensional image
The pixel characteristic attribute of vegetarian refreshments.
In one embodiment, computer equipment obtains the two dimensional image S that size w1*w2 is, carries out two to two dimensional image
Discrete cosine transform is tieed up, transformed two dimensional image is denoted as S1, and transformation for mula is as follows:
(u=0,1 ..., w1-1；V=0,1 ..., w2-1)
Wherein, x and y indicates airspace coordinate value, i.e. the ranks index value of two dimensional image S.U and v indicates spectrum domain coordinate value,
The ranks index value of i.e. transformed two dimensional image S1.
In above-described embodiment, the two dimensional image for being used as the first heavy water print is pre-processed, and by pretreated two dimension
Image is embedded in three-dimensional digital model, improves the robustness of watermark, and converts each pixel in the spectral matrix of acquisition
Point includes all information of respective two-dimensional image, further improves the robustness of watermark.
In one embodiment, the embedding grammar of the dual blind watermatking of above-mentioned three-dimensional digital model, further includes: obtain to first
Two normal vectors that the one ring neighbours center of mass point on class vertex is calculated separately according to two kinds of normal vector calculations and two
Normal vector angle between normal vector；According to the normal vector angle of ring neighbours' center of mass point on first kind vertex and two dimensional image
Pixel characteristic attribute determines the pixel for corresponding to first kind vertex in two dimensional image；It is adjacent according to a ring on first kind vertex
Two normal vectors for occupying center of mass point and ring neighbours' center of mass point establish the respective local spherical coordinate system in first kind vertex respectively；?
In the respective local spherical coordinate system in first kind vertex, the spherical coordinates value on first kind vertex is determined；Step S208 includes: by X-Y scheme
The ranks index value of each pixel is separately encoded into corresponding angle value according to pre-arranged code condition as in, after being encoded
Pixel characteristic attribute；In three-dimensional digital model, corresponding first kind top is adjusted according to the pixel characteristic attribute after coding
The spherical coordinates value of point.
Wherein, local spherical coordinate system is the spherical coordinate system established in subrange.Local spherical coordinate system specifically can be
The spherical coordinate system established corresponding to each vertex in three-dimensional digital model.Spherical coordinates value is for indicating vertex in corresponding part
Position in spherical coordinate system.Spherical coordinates value can specifically include two angle values and a distance value.Two angle values specifically may be used
To be angle value that the line between vertex and origin is constituted with vertical pivot and horizontal axis respectively, distance value specifically can be vertex and former
The distance between point, the ratio of the mean value on the distance between a ring neighbours vertex on vertex to the vertex.Pixel is to constitute
The basic unit of two dimensional image.Pixel point value is the value of each pixel in two dimensional image.Pixel characteristic attribute may include picture
Row index values, column index value and the pixel point value of vegetarian refreshments.Angle value is the value for being expressed as angle form.Pre-arranged code condition is pre-
What is first set is used to become ranks index value the condition of corresponding angle value.
Specifically, computer equipment filters out the first kind vertex for meeting the first heavy water print insertion condition, according to each the
The one ring neighbours vertex on a kind of vertex obtains corresponding ring neighbours' center of mass point, calculates separately one according to two kinds of normal vector calculations
Two normal vectors of ring neighbours' center of mass point, and calculate the normal vector angle between two normal vectors.Computer equipment is according to first
The pixel value of each pixel is big in the size and two dimensional image of the normal vector angle of one ring neighbours' center of mass point on class vertex
It is small, determine first kind vertex corresponding with each pixel in two dimensional image.Wherein, computer equipment can be according to a ring neighbours
The normal vector angle of center of mass point from it is as low as big or from greatly to small sequence and pixel pixel value from greatly to small or from as low as
Big sequence, establishes one-to-one relationship respectively, so that it is determined that the corresponding relationship of pixel and first kind vertex.
Further, for each first kind vertex, computer equipment is with ring neighbours' center of mass point on first kind vertex
One normal vector of origin, ring neighbours' center of mass point is vertical pivot, with another normal vector of ring neighbours' center of mass point by origin
It is projected as horizontal axis in the plane determined with ordinate, the local spherical coordinate system on corresponding first kind vertex is established, determines the first kind
Spherical coordinates value of the vertex in respective local spherical coordinate system.Computer equipment obtains the two dimensional image for being used as the first heavy water print,
Convert two dimensional image and obtain corresponding spectral matrix, scramble take absolute value after spectral matrix, respectively by the X-Y scheme after scramble
The row index values and column index value of each pixel of picture are separately encoded into corresponding angle value according to pre-arranged code condition, obtain
Pixel characteristic attribute after each pixel coding.According to the corresponding pass of pixel and first kind vertex in two dimensional image
System, the corresponding relationship of the spherical coordinates value of pixel characteristic attribute and first kind vertex after determining pixel coding.In three dimensions
In word model, the spherical coordinates value on first kind vertex is replaced with into corresponding pixel characteristic attribute.It is replaced according to first kind vertex
Spherical coordinates value afterwards, the position on corresponding adjustment first kind vertex.
In one embodiment, for each first kind vertex V in three-dimensional digital model MiRing neighbours' center of mass point
Zi, calculate separately corresponding normal vector neAnd ntAnd the normal vector angle Φ between two normal vectors.For in two dimensional image
Each pixel encoded after pixel characteristic attribute (W1, W2, I).According to normal vector angle Φ and pixel point feature
Attribute (W1, W2, I) determines the corresponding relationship of first kind vertex and pixel.Computer equipment is with ring neighbours' center of mass point ZiFor
Origin, normal vector neFor the ordinate (Z axis) of spherical coordinate system, by normal vector ntIn ring neighbours' center of mass point ZiWith normal vector neReally
Projection, projects the projection vector of acquisition as horizontal axis (X-axis), establishes first kind vertex V in fixed planeiLocal spherical coordinates
System.Obtain first kind vertex ViSpherical coordinates value in corresponding local spherical coordinate systemBy first kind vertex ViBall sit
Scale valueReplace with corresponding pixel characteristic attribute (W1, W2, I).According to first kind vertex ViReplaced spherical coordinates
It is worth (W1, W2, I), calculates first kind vertex ViCoordinate value in the corresponding global coordinate system of three-dimensional digital model, by the first kind
Vertex ViAdjust position corresponding to the coordinate value.
Wherein, each first kind vertex V in three-dimensional digital modeliSpherical coordinates value in respective local spherical coordinate system
Solved by customized calculation formula, specific formula for calculation is as follows:
R=sqrt ((x-x0)2+(y-y0)2+(z-z0)2)/ρ
θ=arccos (z-z0|/r)
Wherein, x, y and z respectively indicate first kind vertex ViCoordinate value in global spherical coordinate system.x0、y0And z0Respectively
Indicate first kind vertex ViRing neighbours' center of mass point ZiCoordinate value in global spherical coordinate system.ρ is first kind vertex ViTo
A kind of vertex ViA ring neighbours vertex N1The distance between (i) mean value.
In one embodiment, by the row index values w1 of each pixel in transformed two dimensional image S1iWith column rope
Draw value w2iPre-arranged code condition is pressed respectively, is separately encoded into corresponding angle value W1iAnd W2i, obtain each pixel after coding
(w1i,w2i) pixel characteristic attribute (W1i,W2i,Ii).Wherein, IiIt is pixel (w1i,w2i) pixel point value.Consider embedding
Enter the invisibility of watermark, λ1Value can be set as 0.5, λ2Value can be set as any value between 0.25 to 0.4, λ3Value can be set as 1.In advance
If encoding condition is as follows:
In above-described embodiment, according in the normal vector angle and two dimensional image of ring neighbours' center of mass point on first kind vertex
Pixel characteristic attribute after each pixel coding determines the insertion sequence and each first kind of insertion the first heavy water print
Corresponding watermark information is embedded in 3-dimensional digital by the position on corresponding adjustment first kind vertex by the watermark information being embedded in vertex
In model, it ensure that the order of watermark insertion, and the invisibility of insertion watermark, improve the robustness of watermark.
As shown in figure 3, in a specific embodiment, providing a kind of insertion of dual blind watermatking of three-dimensional digital model
Method.Method includes the following steps:
S302 obtains three-dimensional digital model.
S304 prints insertion conditional filtering first kind vertex according to the first heavy water from three-dimensional digital model.
S306 establishes the respective local spherical coordinate system in each first kind vertex respectively；In the respective part in first kind vertex
In spherical coordinate system, the spherical coordinates value on first kind vertex is determined.
S308 obtains the two dimensional image for being used as the first heavy water print；Two dimensional image is converted into spectrum domain, obtains corresponding frequency
Spectrum matrix；It takes absolute value to spectral matrix；Spectral matrix after taking absolute value according to default scramble condition scramble, is corresponded to
The pixel characteristic attribute of the two dimensional image of spectral matrix.
The ranks index value of pixel each in two dimensional image is separately encoded into according to pre-arranged code condition corresponding by S310
Angle value, the pixel characteristic attribute after being encoded.
S312, acquisition calculate separately a ring neighbours center of mass point on first kind vertex according to two kinds of normal vector calculations
The normal vector angle between two normal vectors and two normal vectors arrived.
S314, according to the pixel point feature of the normal vector angle of ring neighbours' center of mass point on first kind vertex and two dimensional image
Attribute determines the pixel for corresponding to first kind vertex in two dimensional image.
S316 adjusts corresponding first kind vertex according to the pixel characteristic attribute after coding in three-dimensional digital model
Spherical coordinates value, obtain three-dimensional digital model adjusted, that is, obtain the three-dimensional digital model of insertion the first heavy water print.
S318, screening meets the second class vertex of preset condition from three-dimensional digital model adjusted.
S320 selects the two rings neighbour for meeting the second heavy water print insertion condition from the two ring neighbours vertex on the second class vertex
Occupy vertex；The two ring neighbours vertex picked out are not belonging to first kind vertex, and are not belonging to print insertion condition sieve according to the first heavy water
Referenced vertex when selecting first kind vertex.
S322 obtains the watermark sequence of values for being used as the second heavy water print；Watermark sequence of values includes at each numerical digit
Watermark value.
Predetermined angle range is divided into multiple angle groups by S324；Corresponding to the of the two ring neighbours vertex picked out
In the triangle that two class vertex and two ring neighbours vertex on two ring neighbours vertex of connection and the second class vertex are constituted, according to
Angle group belonging to apex angle locating for second class vertex determines the numerical digit in watermark sequence of values with two ring neighbours' vertex correspondences；
According to the numerical digit of watermark sequence of values, the watermark value of two ring neighbours' vertex correspondences is determined.
S326 determines two rings according to projected position of the two ring neighbours vertex picked out on corresponding ring neighbours side
The numerical value of neighbours' vertex correspondence.
S328, when the numerical value and watermark value that correspond to two ring neighbours vertex are unequal, after determining adjustment according to watermark value
Projected position of the two ring neighbours vertex on a ring neighbours side, according to the position on two ring neighbours vertex of the corresponding adjustment of projected position
It sets, obtains the three-dimensional digital model of insertion the second heavy water print.
In above-described embodiment, double watermark is successively embedded in three-dimensional digital model, and the first heavy water print being embedded in is pre-
The second heavy water print of treated two dimensional image, insertion is watermark sequence of values, improves the robustness of watermark.Pass through multiple sieve
Condition is selected to screen the vertex for being embedded in double watermark respectively, so that the vertex of the first heavy water of insertion print and the second heavy water print is mutually only
It is vertical, further improve the robustness of watermark.
As shown in figure 4, in one embodiment, providing a kind of extracting method of dual blind watermatking of three-dimensional digital model.
The present embodiment is applied to the computer equipment in above-mentioned Fig. 1 in this way to illustrate.This method specifically comprises the following steps:
S402 obtains three-dimensional digital model.
Specifically, computer equipment obtains three-dimensional digital model, according to preset watermark extracting mode from acquired three
Watermark is extracted in dimension word model, to determine whether have been inserted into watermark in three-dimensional digital model.
S404 prints insertion conditional filtering first kind vertex according to the first heavy water from three-dimensional digital model.
Specifically, computer equipment obtains the first heavy water and prints insertion condition, prints insertion condition from being obtained according to the first heavy water
First kind vertex is screened in the three-dimensional digital model taken, the first kind vertex filtered out meets the first heavy water print insertion condition.Its
In, the first kind vertex filtered out is the vertex that the first heavy water print may be embedded in three-dimensional digital model.
In one embodiment, for acquired three-dimensional digital model, computer equipment is according to the one of each vertex of calculation
Ring neighbours vertex calculates separately ring neighbours' center of mass point on each vertex, adjacent according to a ring neighbours vertex on each vertex and a ring
Center of mass point is occupied, two normal vectors of ring neighbours' center of mass point on each vertex are calculated separately according to two kinds of normal vector calculations.
Two normal direction on a ring neighbours vertex and a ring neighbours center of mass point and ring neighbours' center of mass point further according to each vertex
Amount, selection meets the first kind vertex of the first heavy water print insertion condition from three-dimensional digital model.
Further, the first kind vertex that filters out while meeting following three conditions: first is that a ring on first kind vertex
The ratio of the maxima and minima of the distance between neighbours vertex and ring neighbours' center of mass point is less than default fractional threshold；Second is that
Normal vector angle between two normal vectors of one ring neighbours' center of mass point on first kind vertex is greater than default angle threshold value；Third is that the
A kind of vertex is not mutually neighbours vertex.Wherein, presetting fractional threshold and default angle threshold value can print with the first heavy water of insertion respectively
When preset value it is equal.
S406 determines corresponding pixel characteristic attribute according to the position on first kind vertex.
Specifically, computer equipment obtains the position on each first kind vertex filtered out, determining and each first kind top
The corresponding pixel of point, the pixel characteristic attribute of each pixel is determined according to identified corresponding relationship.
In one embodiment, computer equipment obtains ball of each first kind vertex in respective local spherical coordinate system
Coordinate value, and the two-dimensional matrix for obtaining default size determines two according to the spherical coordinates value of two-dimensional matrix and each first kind vertex
The corresponding relationship for tieing up each pixel and first kind vertex in matrix, determines pixel according to the spherical coordinates value on first kind vertex
Pixel characteristic attribute.Wherein, the default size of two-dimensional matrix and the two dimensional image for being used as the first heavy water print are in the same size.
S408 restores the two dimensional image as the first heavy water print according to pixel characteristic attribute.
Specifically, computer equipment obtains respective pixel in two-dimensional matrix according to the position on each first kind vertex respectively
The pixel characteristic attribute of point, and corresponding two dimensional image is restored according to the pixel characteristic attribute of each pixel, it is acquired
Two dimensional image as extract the first heavy water print.
In one embodiment, computer equipment will determine the X-Y scheme of the pixel characteristic attribute of each pixel
Picture restores corresponding two dimensional image according to default reduction mode, using the two dimensional image restored as insertion 3-dimensional digital mould
The first heavy water print in type.
S410, screening meets the second class vertex of preset condition from three-dimensional digital model；Preset condition is included in first
Heavy water prints insertion condition.
Specifically, computer equipment obtains three-dimensional digital model, and screening meets pre- from acquired three-dimensional digital model
If the second class vertex of condition.The second class vertex for filtering out while meeting following two conditions: first is that the one of the second class vertex
The ratio of the maxima and minima of the distance between ring neighbours vertex and ring neighbours' center of mass point is less than default fractional threshold；Two
It is the normal vector angle between two normal vectors of ring neighbours' center of mass point on the second class vertex greater than default angle threshold value.Screening
The preset condition on the second class vertex is included in the first heavy water print insertion condition, and the second class vertex filtered out includes first kind top
Point.Wherein, when preset condition and the second heavy water of insertion print, the consistent of the second class vertex is screened.
In one embodiment, the three-dimensional digital model that computer equipment screens the second class vertex, which can be, is extracted first
The three-dimensional digital model of heavy water print is also possible to not yet extract the three-dimensional digital model of the first heavy water print.
S412 selects the two rings neighbour for meeting the second heavy water print insertion condition from the two ring neighbours vertex on the second class vertex
Occupy vertex；The two ring neighbours vertex picked out are not belonging to first kind vertex, and are not belonging to print insertion condition sieve according to the first heavy water
Referenced vertex when selecting first kind vertex.
Specifically, computer equipment filters out the second class vertex for meeting preset condition from three-dimensional digital model, by
Double watermark is embedded in condition, from the two ring neighbours vertex on the second class vertex filtered out, selects and meets the print insertion of the second heavy water
Two ring neighbours vertex of condition, the two ring neighbours vertex picked out are used as the vertex of insertion the second heavy water print.Two rings selected are adjacent
It occupies vertex and is not belonging to first kind vertex, and be not belonging to the ring neighbours top on the first kind vertex of reference when screening first kind vertex
Point.
Further, the two ring neighbours vertex selected while meeting following four conditions: first is that two ring neighbours vertex only with
Two ring neighbours vertex on corresponding second class vertex are connected, wherein the two rings neighbours vertex is the second class vertex
Two ring neighbours vertex；Second is that two rings are adjacent in the triangle that two ring neighbours vertex and above-mentioned two ring neighbours vertex are constituted
Occupying base angle locating for vertex is acute angle；It is preset third is that the normal vector angle of ring neighbours' center of mass point on two ring neighbours vertex is less than
Angle threshold value, wherein ring neighbours' center of mass point on two ring neighbours vertex is the ring neighbours vertex according to the two rings neighbours vertex
The ring neighbours' center of mass point calculated；Fourth is that the normal direction of the respective ring neighbours' center of mass point in a ring neighbours vertex on two ring neighbours vertex
It measures angle and is less than default angle threshold value.
S414, according to the position on the two ring neighbours vertex picked out matched watermark value and the two ring neighbours that pick out
The corresponding numerical digit in vertex restores the watermark sequence of values as the second heavy water print.
Specifically, computer equipment is selected from the two ring neighbours vertex on the second class vertex meets the second heavy water print insertion item
Two ring neighbours vertex of part obtain the position for each of selecting two ring neighbours vertex, determine respectively in watermark sequence of values with it is every
The numerical digit of a two rings neighbours' vertex correspondence, and the watermark value of corresponding numerical digit.According to the corresponding relationship of numerical digit and watermark value, also
Former corresponding watermark sequence of values, the watermark sequence of values of reduction is as the second heavy water print extracted from three-dimensional digital model.
In one embodiment, computer equipment extracts the first heavy water print and the second heavy water print from three-dimensional digital model
Sequencing is unlimited, also can according to need any heavy in only extraction doubly time series model.
The extracting method of the above-mentioned dual blind watermatking of three-dimensional digital model prints insertion condition and preset condition according to the first heavy water
Corresponding first kind vertex and the second class vertex are filtered out respectively from three-dimensional digital model, according to the first kind vertex filtered out
Two dimensional image is restored, i.e., extracts the first heavy water print from three-dimensional digital model.It screens and meets from the second class vertex of screening
Second heavy water prints two ring neighbours vertex of insertion condition, and restores corresponding watermark numerical value according to the two ring neighbours vertex filtered out
Sequence extracts the second heavy water print from three-dimensional digital model, to realize the extraction of doubly time series model, improves three dimensions
The robustness of watermark in word model.In doubly time series model extraction process, independently of each other, and it is not related to being not embedded into the 3-dimensional digital of watermark
Model further improves the robustness that attack is resisted in embedded watermark in three-dimensional digital model.
In one embodiment, step S406 includes: to establish the respective local spherical coordinate system in first kind vertex；Obtain first
Spherical coordinates value of the class vertex in respective local spherical coordinate system；Spherical coordinates value includes two angle values and a distance value；It presses
It is separately encoded two angle values on first kind vertex according to default Gray code condition, obtains corresponding vertex attributes values；It obtains default
The two-dimensional matrix of size；Two-dimensional matrix is null matrix；According to the vertex attributes values and two-dimensional matrix for corresponding to first kind vertex
Ranks index value determines the pixel in two-dimensional matrix with first kind vertex correspondence；The pixel on first kind vertex will be corresponded to
Pixel point value be substituted for the distance value on corresponding first kind vertex, obtain the pixel point feature category of replaced two-dimensional matrix
Property.
Wherein, Gray code condition is preset to be used to angle value being encoded into corresponding vertex attributes values.Default Gray code condition
It is corresponding with pre-arranged code condition when insertion the first heavy water print, the two inverse process each other.Vertex attributes values are for indicating first
Coordinate value after class vertex encoding.Pixel is the basic unit for constituting two-dimensional matrix.Null matrix is the square that pixel is zero
Battle array.
Specifically, for each first kind vertex in three-dimensional digital model, computer equipment is respectively according to each first
The one ring neighbours vertex on class vertex obtains corresponding ring neighbours' center of mass point, according to a ring neighbours vertex on each first kind vertex
Corresponding local spherical coordinate system is established with ring neighbours' center of mass point, obtains each first kind vertex in respective local spherical coordinate system
Spherical coordinates value.Wherein, local spherical coordinate system is using a ring neighbours center of mass point on first kind vertex as origin, ring neighbours' center of mass point
A normal vector be vertical pivot, with another normal vector of ring neighbours' center of mass point in the plane determined by origin and ordinate
Be projected as horizontal axis.Acquired spherical coordinates value includes the angle of the line and vertical pivot composition between first kind vertex and origin
Value, and the angle value constituted with horizontal axis and two distances, compared to the distance ratio obtained, two distances are respectively the first kind
The distance between origin and first kind vertex are arrived to the mean value on the distance between a ring neighbours vertex in vertex.
Further, two angles in the spherical coordinates value on each first kind vertex are separately encoded according to default Gray code condition
Angle value, obtains corresponding vertex attributes values, and vertex attributes values obtained are used to obtain the pixel point feature category of respective pixel point
Property.Computer equipment obtains the two-dimensional matrix of default size, and each pixel of acquired two-dimensional matrix is zero.Computer equipment
Respectively by two angle values after each first kind vertex encoding respectively with the ranks index value ratio of pixel each in two-dimensional matrix
Compared with so that it is determined that the pixel in two-dimensional matrix with first kind vertex correspondence.It is closed according to pixel is corresponding with first kind vertex
The pixel point value of each pixel in two-dimensional matrix, is replaced with the distance value on first kind vertex by system.
In one embodiment, for each first kind vertex in three-dimensional digital model, computer equipment is by each
The angle value on a kind of vertex presses default Gray code conditional compilation, and distance value remains unchanged, and obtains corresponding vertex attributes values.It calculates
Machine equipment makes the difference two angle values after each first kind vertex encoding with the ranks index value of pixel respectively.When difference point
Not little Yu corresponding preset threshold when, the corresponding relationship on the first kind vertex and pixel is established, by the pixel point value of pixel
It is substituted for the distance value on the first kind vertex.If pixel in two-dimensional matrix not with any first kind in three-dimensional digital model
Corresponding relationship is established on vertex, then keeps zero constant the pixel point value of the pixel.
In one embodiment, for three-dimensional digital model M, the first kind vertex V of computer equipment screeningiSpherical coordinates
Value isAccording to default Gray code conditional compilationAnd θ, after being encodedAnd θ, it is denoted as Ψ and Θ, to obtain
Vertex attributes values after corresponding first kind vertex encoding.Default Gray code condition are as follows:
The two-dimensional matrix of acquisition is denoted as W, and the pixel characteristic attribute of each pixel is denoted as (w1, w2,0), and w1 and w2 divide
It Wei not the corresponding ranks index value of pixel.The Ψ on first kind vertex and Θ are compared with the w1 of pixel and w2 respectively, when
When meeting following conditions simultaneously, 0 value of pixel is replaced with to the r on first kind vertex.Condition specifically:
|w1-Ψ|<λ1&|w2-Θ|<λ2
Wherein, λ 1 and λ 2 is the preset threshold of two comparisons.
In above-described embodiment, for the first kind vertex of the first heavy water print may be embedded in, in respective local spherical coordinate system
It is middle to obtain corresponding spherical coordinates value, the pixel in first kind vertex and two-dimensional matrix is determined according to the spherical coordinates value after coding
Corresponding relationship, so that it is determined that in two-dimensional matrix each pixel pixel attributive character, according to the pixel attributive character restore
Corresponding two dimensional image, in order to extract the two dimensional image for being used as the first heavy water print from three-dimensional digital model.
In one embodiment, step S408 includes: to invert the random replaced two-dimensional matrix of condition scramble according to default；It obtains
It takes and the equal-sized two-dimensional coefficient matrix of two-dimensional matrix；It, will be with two-dimentional coefficient by two-dimensional matrix and two-dimensional coefficient matrix dot product
Two-dimensional matrix after matrix dot product is converted into airspace, obtains the two dimensional image for being used as the first heavy water print.
Wherein, the position for inverting random condition for each pixel in scramble two-dimensional matrix is preset.It is default to invert random condition
It is corresponding to the default scramble condition of two dimensional image scramble when to insertion the first heavy water print.Two-dimensional coefficient matrix can be with positive and negative system
Several unit matrixs.Two-dimensional coefficient matrix specifically can be the two-dimensional matrix being made of positive 1 and minus 1.Dot product is by two two dimensions
The pixel of corresponding position is corresponding in matrix is multiplied.
Specifically, computer equipment is according to respective pixel point in the corresponding replacement two-dimensional matrix of the distance value on first kind vertex
Pixel point value obtains replaced two-dimensional matrix, inverts the random replaced two-dimensional matrix of condition scramble by default.Computer equipment
It obtains and the equal-sized two-dimensional coefficient matrix of two-dimensional matrix.Wherein, two-dimensional coefficient matrix is embedded in during the first heavy water print,
The two dimensional image of the first heavy water print be will act as when being converted into spectrum domain, the coefficient matrix of obtained spectral matrix.Computer equipment
By after scramble two-dimensional matrix and acquired two-dimensional coefficient matrix dot product, by with the two-dimensional matrix after two-dimensional coefficient matrix dot product
It is converted into airspace, obtains the two dimensional image for being used as the first heavy water print.
In one embodiment, the two-dimensional matrix after replacement pixel point value is carried out Aronld scramble by computer equipment.Its
In, to the Aronld scrambled fashion of two dimensional image on the contrary, each to restore when the Aronld scrambled fashion and the first heavy water of insertion print
The corresponding position relationship of pixel.Computer equipment obtains the two dimensional image for being used as the first heavy water print, by acquired X-Y scheme
As being converted into spectrum domain, using the coefficient matrix of spectral matrix obtained as two-dimensional coefficient matrix, by the two-dimensional coefficient matrix
With the two-dimensional matrix dot product after scramble.Two-dimensional matrix after dot product is subjected to two-dimensional inverse discrete cosine transformation, it is corresponding to restore
Two dimensional image is printed the two dimensional image of reduction as the first heavy water extracted from three-dimensional digital model.Wherein, two after dot product
Dimension matrix is denoted as S2, and the transformed two-dimensional matrix of two-dimensional inverse discrete cosine is denoted as S0.The transformation of two-dimensional inverse discrete cosine transformation is public
Formula is as follows:
(u=0,1 ..., w1-1；V=0,1 ..., w2-1)
In above-described embodiment, according to the opposite treatment process of the two dimensional image for being used as the first heavy water print with pretreatment, processing
The two-dimensional matrix of replacement pixel point value, to restore the two dimensional image in insertion three-dimensional digital model.
In one embodiment, the extracting method of the dual blind watermatking of above-mentioned three-dimensional digital model further include: according to picking out
Projected position of the two ring neighbours vertex on corresponding ring neighbours side, determine the numerical value of two ring neighbours' vertex correspondences；It will count
Value is used as in watermark sequence of values, the watermark value with the numerical digit of two ring neighbours' vertex correspondences.
Specifically, computer equipment screens the second class vertex for meeting preset condition from three-dimensional digital model, from screening
The second class vertex in two ring neighbours vertex in select meet the second heavy water print insertion condition two ring neighbours vertex.Computer
The two ring neighbours vertex that equipment will be singled out are projected on corresponding ring neighbours side, obtain two ring neighbours vertex corresponding one
Projected position on ring neighbours side determines the numerical value of two ring neighbours' vertex correspondences according to projected position.
Further, predetermined angle range is divided into numerical digit sum phase of the quantity with watermark sequence of values by computer equipment
Deng angle group, each angle group is corresponding with each numerical digit of watermark sequence of values respectively.In three-dimensional digital model, computer
Equipment on two ring neighbours vertex by being connected between the two ring neighbours vertex selected and corresponding second class vertex, and
In the triangle that the second class vertex is constituted, the angle of apex angle locating for the second class vertex is obtained, and judge acquired angle
Angle group belonging to degree.Angle group belonging to computer equipment apex angle according to locating for the second class vertex and angle group with
The corresponding relationship of each numerical digit in watermark sequence of values determines that the second class vertex is corresponding with the numerical digit in watermark sequence of values and closes
System.Computer equipment is according to the corresponding relationship of numerical digit in the second class vertex and watermark sequence of values, respectively by each second class top
Watermark value of the numerical value of point as corresponding numerical digit, to restore watermark sequence of values.
In one embodiment, two ring neighbours vertex between the second class vertex and two ring neighbours vertex D are connected to
It is indicated respectively with A and C, the ring neighbours side being made of the line between A and C is denoted as AC, by two ring neighbours vertex D in ring neighbour
It occupies the projected position projected on the AC of side and is denoted as B.Determine in corresponding two rings neighbours vertex D whether be embedded in second according to projected position B
Heavy water print and the watermark value α of insertion.When the corresponding projected position B of two ring neighbours vertex D meets following conditions, then two ring
Neighbours vertex D is embedded in watermark, and the watermark value α being embedded in is 0.
When the corresponding projected position B of two ring neighbours vertex D meets following conditions, then the two rings neighbours vertex D is embedded in
Watermark, and the watermark value α being embedded in is 1.
For three-dimensional digital model M, the second class vertex V that computer equipment filters outiA ring neighbours vertex N1(i) divide
Corresponding ring neighbours' center of mass point Z is not calculatedi,
In one embodiment, in each of three-dimensional digital model the second class vertex and watermark sequence of values each numerical digit it
Between corresponding relationship determination, with insertion the second heavy water print when determine that the mode of corresponding relationship is identical.
In above-described embodiment, corresponding number is determined according to projected position of the second class vertex on corresponding ring neighbours side
Value, and using the numerical value of each second class vertex correspondence as the watermark value of numerical digit corresponding in watermark sequence of values, with from three dimensions
The second heavy water print is extracted in word model.
In one embodiment, step S414 includes: in watermark sequence of values, according to the two ring neighbours vertex picked out
Determine the watermark value for corresponding to the numerical digit on two ring neighbours vertex；It is true according to two ring neighbours vertex in cumulative calculation watermark sequence of values
The watermark value and the corresponding total quantity of watermark value of fixed each numerical digit；When the corresponding watermark value of numerical digit in watermark sequence of values
When more than one, watermark value of the maximum watermark value of total quantity as numerical digit is screened, the watermark number for being used as the second heavy water print is obtained
Value sequence.
Specifically, computer equipment determines each second class vertex and each number in watermark sequence of values in three-dimensional digital model
Corresponding relationship between position, each numerical digit in watermark sequence of values can be corresponding with one or extra one the second class vertex,
Based on the corresponding relationship of the second class vertex sum digit, the watermark value of corresponding numerical digit is determined according to each second class vertex, and add up
The corresponding watermark value of numerical digit and the total quantity of each watermark value determined by calculating.For each numerical digit in watermark sequence of values,
When the watermark value of the numerical digit according to determined by corresponding second class vertex is more than one, the maximum watermark value of screening total quantity is made
For the watermark value of corresponding numerical digit, the watermark sequence of values for being used as the second heavy water print is obtained.
In one embodiment, watermark sequence of values can be binary sequence, the watermark value of each numerical digit in watermark sequence of values
It can be 0 or 1, computer equipment determines the watermark value of each numerical digit according to corresponding second class vertex respectively, and calculates each numerical digit
The total quantity of the total quantity of each watermark value, the corresponding watermark value 0 of numerical digit i is denoted as sum (0), and the total quantity of watermark value 1 is denoted as sum
(1).The size for comparing sum (0) and sum (1) at the position numerical digit i takes watermark of the biggish watermark value of total quantity as numerical digit i
Value.
In above-described embodiment, determined according to the second class of each of three-dimensional digital model vertex corresponding in watermark sequence of values
The watermark value of numerical digit, when the corresponding watermark value of each numerical digit is more than one, by selecting the maximum watermark value of total quantity as phase
The watermark value for answering numerical digit improves the accuracy of restored watermark sequence of values, has correspondinglyd increase the second weight watermark extracting
Accuracy, to improve the robustness that attack is resisted in watermark.
In one embodiment, the extracting method of the dual blind watermatking of above-mentioned three-dimensional digital model further include: obtain for testing
Demonstrate,prove the authentication image of the first heavy water print；Calculate the degree of correlation of the two dimensional image restored and authentication image；When the degree of correlation is greater than the
When one default relevance threshold, then determine to be embedded in the first heavy water print in three-dimensional digital model；And/or it obtains for verifying the
The verifying sequence of double watermark；Calculate the degree of correlation of the watermark sequence of values restored and verifying sequence；When the degree of correlation is greater than the
When two default relevance thresholds, then determine to be embedded in the second heavy water print in three-dimensional digital model.
Wherein, authentication image is used to verify the two dimensional image of the validity of extracted first heavy water print.Authentication image tool
Body can be the two dimensional image printed in insertion three-dimensional digital model as the first heavy water.Verifying sequence is for verifying extracted the
The watermark sequence of values of the validity of double watermark.Verifying sequence, which specifically can be, prints insertion 3-dimensional digital mould as the second heavy water
Verifying sequence in type.The degree of correlation is used to indicate the similarity degree of the two images or two watermark sequence of values to compare.Phase
Pass degree specifically can be any value between 0 to 1, and the degree of correlation is that 1 expression is just the same, and the degree of correlation is that 0 expression is completely different.
First default relevance threshold and the second default relevance threshold are preset for compared with the degree of correlation, to determine
Whether be embedded in three-dimensional digital model corresponding watermark and extracted watermark whether be original insertion watermark.
Specifically, two dimensional image is extracted from three-dimensional digital model and restored to computer equipment, using the two dimensional image as
The first heavy water extracted prints, and obtains the authentication image for verifying extracted first heavy water print, calculates authentication image and go back
The degree of correlation of former two dimensional image, the degree of correlation of calculating is compared with the first default relevance threshold, when the degree of correlation is greater than the
When one default relevance threshold, then determine to be embedded in the two dimensional image as the first heavy water print in the three-dimensional digital model.
Further, computer equipment extracts simultaneously reductive water printing value sequence from three-dimensional digital model, by the watermark number
Value sequence obtains the verifying sequence for verifying extracted second heavy water print as the second heavy water print extracted, and calculating is tested
The degree of correlation for demonstrate,proving the watermark sequence of values of sequence and reduction, the degree of correlation of calculating is compared with the second default relevance threshold,
When the degree of correlation is greater than the second default relevance threshold, then determine to be embedded in the three-dimensional digital model as the second heavy water print
Watermark sequence of values.
It is worth noting that computer equipment can only extract the X-Y scheme for being used as the first heavy water print from three-dimensional digital model
As or as the second heavy water print watermark sequence of values, and calculate acquired two dimensional image or watermark sequence of values and mutually come ture
Demonstrate,prove the degree of correlation between image or verifying sequence.The degree of correlation calculated and corresponding first default relevance threshold or second is pre-
If relevance threshold compares, when the degree of correlation be greater than the first default relevance threshold or the second default relevance threshold, then illustrate
The first heavy water print or the second heavy water print are embedded in three-dimensional digital model, so as to the copyright of the valid certificates three-dimensional digital model
Ownership.Through setting relevance threshold compared with the degree of correlation calculated, it can effectively avoid making because of various watermark attack methods
The watermark extracting that must be embedded in three-dimensional digital model is imperfect and leads to decision error.Computer equipment is worked as from three-dimensional digital model
It extracts any heavy in double watermark, then can determine whether to be embedded in watermark in the three-dimensional digital model, further improve three-dimensional
The robustness of attack is resisted in watermark in mathematical model.
Computer equipment can also be extracted simultaneously from three-dimensional digital model to be used as the two dimensional image of the first heavy water print and is used as
The watermark sequence of values of second heavy water print, and calculate separately acquired two dimensional image and watermark sequence of values and corresponding proof diagram
The degree of correlation between picture and verifying sequence.To calculate separately the degree of correlation obtained respectively with corresponding first default relevance threshold
It compares with the second default relevance threshold, when the corresponding degree of correlation of two dimensional image is greater than the first default relevance threshold, and water
The corresponding degree of correlation of printing value sequence is greater than the second default relevance threshold, then illustrates to be embedded in the first weight in three-dimensional digital model
Watermark and the second heavy water print, so as to the copyright ownership of the valid certificates three-dimensional digital model.When double watermark extracts,
Just determine to be embedded in the double watermark in the three-dimensional digital model, thus assert the copyright ownership of the three-dimensional digital model, it can
Ensure the accuracy of copyright ownership identification.
In above-described embodiment, by calculating the first heavy water extracted print and/or the second heavy water print and corresponding proof diagram
As or verifying sequence the degree of correlation, and three-dimensional is compared to determine with corresponding relevance threshold respectively according to the degree of correlation calculated
Whether watermark and embedded watermark information are embedded in mathematical model, and setting relevance threshold ensure that the accuracy of judgement.
As described in Figure 5, in one embodiment, a kind of extracting method of dual blind watermatking of three-dimensional digital model is provided.
Method includes the following steps:
S502 obtains three-dimensional digital model.
S504 prints insertion conditional filtering first kind vertex according to the first heavy water from three-dimensional digital model.
S506 establishes the respective local spherical coordinate system in each first kind vertex；First kind vertex is obtained in respective part
Spherical coordinates value in spherical coordinate system；Spherical coordinates value includes two angle values and a distance value；According to default Gray code condition point
Not Bian Ma first kind vertex two angle values, obtain corresponding vertex attributes values.
S508 obtains the two-dimensional matrix of default size；Acquired two-dimensional matrix is null matrix；According to corresponding to the first kind
The vertex attributes values on vertex and the ranks index value of two-dimensional matrix, determine the pixel in two-dimensional matrix with first kind vertex correspondence
Point.
The pixel point value for corresponding to the pixel on first kind vertex is substituted for the distance on corresponding first kind vertex by S510
Value, obtains the pixel characteristic attribute of replaced two-dimensional matrix.
S512 inverts the random replaced two-dimensional matrix of condition scramble according to default.
S514 is obtained and the equal-sized two-dimensional coefficient matrix of two-dimensional matrix；By two-dimensional matrix and two-dimensional coefficient matrix point
Multiply, airspace will be converted into the two-dimensional matrix after two-dimensional coefficient matrix dot product, obtain the two dimensional image for being used as the first heavy water print.
S516 obtains the authentication image for verifying the first heavy water print；Calculate the two dimensional image restored and authentication image
The degree of correlation；When the degree of correlation is greater than the first default relevance threshold, then determine to be embedded in the first heavy water in three-dimensional digital model
Print.
S518, screening meets the second class vertex of preset condition from three-dimensional digital model；Preset condition is included in first
Heavy water prints insertion condition.
S520 selects the two rings neighbour for meeting the second heavy water print insertion condition from the two ring neighbours vertex on the second class vertex
Occupy vertex；The two ring neighbours vertex picked out are not belonging to first kind vertex, and are not belonging to print insertion condition sieve according to the first heavy water
Referenced vertex when selecting first kind vertex.
S522 determines two rings according to projected position of the two ring neighbours vertex picked out on corresponding ring neighbours side
The numerical value of neighbours' vertex correspondence；Using numerical value as the watermark value in watermark sequence of values, with the numerical digit of two ring neighbours' vertex correspondences.
S524 is determined according to the two ring neighbours vertex picked out in watermark sequence of values and is corresponded to two ring neighbours vertex
Numerical digit watermark value；The watermark value of each numerical digit determined in cumulative calculation watermark sequence of values according to two ring neighbours vertex, with
And the corresponding total quantity of watermark value.
S526 screens the maximum water of total quantity when the corresponding watermark value of the numerical digit in watermark sequence of values is more than one
Watermark value of the print value as numerical digit obtains the watermark sequence of values for being used as the second heavy water print.
S528 obtains the verifying sequence for verifying the second heavy water print；Calculate the watermark sequence of values restored and verifying
The degree of correlation of sequence；When the degree of correlation is greater than the second default relevance threshold, then determine to be embedded in second in three-dimensional digital model
Heavy water print.
It is worth noting that the above-mentioned sequencing for extracting the first heavy water print and the second heavy water print from three-dimensional digital model
Do not limit, the first heavy water of extraction respectively print and the second heavy water print that can be arranged side by side, can also first wherein any one heavy water print, then
It extracts and extracts another heavy water print on the basis of a heavy water prints again.
During above-described embodiment is looked for, by extracting watermark from three-dimensional digital model, and with corresponding authentication image or verifying
Sequence calculates the degree of correlation, determines watermark and embedded watermark whether are embedded in corresponding three-dimensional digital model according to the degree of correlation
Content.
In a specific embodiment, as shown in fig. 6, providing a kind of insertion of dual blind watermatking of three-dimensional digital model
And extracting method.In this method, doubly time series model is embedded in three-dimensional digital model, then extract respectively.Wherein, a is as first
The two dimensional image of heavy water print；B is pretreated two dimensional image；C is the three-dimensional digital model for being not yet embedded in watermark；D is insertion
The three-dimensional digital model of first heavy water print；E is the watermark sequence of values printed as the second heavy water；F is that the second heavy water of insertion prints
Three-dimensional digital model is embedded in the three-dimensional digital model of doubly time series model；G is the second heavy water extracted from three-dimensional digital model
Print；H is the first heavy water print extracted from three-dimensional digital model.
As shown in fig. 7, in one embodiment, providing a kind of flush mounting of dual blind watermatking of three-dimensional digital model
700.Referring to Fig. 7, the flush mounting 700 of the dual blind watermatking of three-dimensional digital model include: obtain module 701, screening module 702,
First weight watermark embedding module 703 and the second weight watermark embedding module 704.
Module 701 is obtained, for obtaining three-dimensional digital model.
Screening module 702, for printing insertion conditional filtering first kind vertex according to the first heavy water from three-dimensional digital model.
Module 701 is obtained, is also used to obtain the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print.
First weight watermark embedding module 703, in three-dimensional digital model by pixel characteristic attribute adjustment corresponding the
The position on a kind of vertex.
Screening module 702 is also used to screen the second class top for meeting preset condition from three-dimensional digital model adjusted
Point；Preset condition is included in the first heavy water print insertion condition.
Screening module 702 is also used to from the two ring neighbours vertex on the second class vertex, is selected and is met the print insertion of the second heavy water
Two ring neighbours vertex of condition；The two ring neighbours vertex picked out are not belonging to first kind vertex, and are not belonging to according to the first heavy water
Referenced vertex when print insertion conditional filtering first kind vertex.
Module 701 is obtained, is also used to obtain the watermark sequence of values for being used as the second heavy water print；Watermark sequence of values includes place
Watermark value at each numerical digit.
Second weight watermark embedding module 704, for when the pick out, position on two ring neighbours corresponding with numerical digit vertex and
When watermark value at corresponding numerical digit mismatches, the position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value.
The flush mounting of the above-mentioned dual blind watermatking of three-dimensional digital model, by according to the two dimensional image for being used as the first heavy water print
The position on first kind vertex that filters out of the corresponding adjustment of pixel point feature, the first heavy water is printed in insertion three-dimensional digital model.
For the three-dimensional digital model after insertion the first heavy water print, the two ring neighbours top on the vertex as insertion the second heavy water print is filtered out
Point, according to the corresponding relationship on the two ring neighbours vertex filtered out and the watermark sequence of values printed as the second heavy water, corresponding adjustment
The position on the two ring neighbours vertex filtered out prints the second heavy water in insertion three-dimensional digital model, thus in 3-dimensional digital mould
It is embedded in dual blind watermatking in type, improves the robustness of watermark in three-dimensional digital model.Insertion condition is printed by the second heavy water
Restrict, ensure that the first heavy water print being embedded in three-dimensional digital model and the second heavy water print are mutually indepedent, do not interfere with each other,
Further improve the robustness of watermark.
In one embodiment, screening module 702 is also used to the ring neighbours according to vertex each in three-dimensional digital model
Vertex respectively corresponds the ring neighbours' center of mass point for calculating each vertex；According to a ring neighbours vertex on each vertex and each top
One ring neighbours' center of mass point of point, from three-dimensional digital model, screening meets preset condition and is not mutually the first kind on neighbours vertex
Vertex.
In above-described embodiment, according to a ring neighbours vertex on each vertex and corresponding ring neighbours' center of mass point, from three-dimensional
The first kind vertex for being embedded in the first heavy water print is filtered out in mathematical model, the first kind vertex filtered out meets default item
Part and be not mutually neighbours vertex, so that it is mutually indepedent between each first kind vertex of insertion the first heavy water print, to improve the first weight
The robustness of watermark.
In one embodiment, screening module 702 is also used to the ring neighbours according to vertex each in three-dimensional digital model
Center of mass point and a ring neighbours vertex, two normal direction of ring neighbours' center of mass point are calculated separately according to two kinds of normal vector calculations
Amount；First kind vertex is selected from three-dimensional digital model；The one ring neighbours vertex on the first kind vertex of selection and ring neighbours' matter
The ratio of the maxima and minima of the distance between heart point is less than default fractional threshold, and a ring on the first kind vertex of selection is adjacent
Occupy normal vector angle between two normal vectors of center of mass point and be greater than default angle threshold value, and the first kind vertex selected mutually not for
Neighbours vertex.
In above-described embodiment, by the first heavy water print insertion condition be specifically divided into three conditions, based on three conditions from
Screening is used as the first kind vertex of insertion the first heavy water print in three-dimensional digital model.Based on three conditional filtering first kind tops
Point ensure that the invisibility of embedded watermark in first kind vertex, improves the accuracy of watermark insertion, effectively prevents water
The ambiguousness for printing insertion, to improve the robustness of watermark.
In one embodiment, the two ring neighbours vertex corresponding second class vertex for existing in the second class vertex and selecting,
And the two ring neighbours vertex selected only are connected with two ring neighbours vertex on corresponding second class vertex；The two ring neighbours selected
In the triangle that vertex and two ring neighbours vertex are constituted, base angle locating for two ring neighbours vertex is acute angle；It selects
The normal vector angle of ring neighbours' center of mass point on two ring neighbours vertex be less than default angle threshold value；The two ring neighbours vertex selected
The normal vector angle of the respective ring neighbours' center of mass point in a ring neighbours vertex be less than default angle threshold value.
In above-described embodiment, insertion the second heavy water print is used as based on preset condition and the second heavy water print insertion conditional filtering
Two ring neighbours vertex ensure that the part vertex in three-dimensional digital model is remained unchanged as watermark insertion invariant, improve
The robustness of embedded watermark.And the restrict that insertion condition is printed by the second heavy water, so that embedding in three-dimensional digital model
The the first heavy water print and the second heavy water print entered is mutually indepedent, does not interfere with each other, further improves the robustness of watermark.
In one embodiment, the second weight watermark embedding module 704 is also used to predetermined angle range being divided into multiple angles
Degree group；The number of angle group is equal with the numerical digit sum of watermark sequence of values；Each angle group respectively with watermark sequence of values
Each numerical digit is corresponding；Corresponding to the second class vertex on two ring neighbours vertex picked out and two ring neighbours vertex of connection and the
In the triangle that two ring neighbours vertex on two class vertex are constituted, according to angle belonging to apex angle locating for the second class vertex
Group determines the numerical digit in watermark sequence of values with two ring neighbours' vertex correspondences；According to the numerical digit of watermark sequence of values, two rings are determined
The watermark value of neighbours' vertex correspondence.
In above-described embodiment, the total equal angle group of numerical digit by obtaining quantity and watermark sequence of values, by the
Two class vertex, and connect the second class vertex and meet two of corresponding two rings neighbours vertex of the second heavy water print insertion condition
In the triangle that one ring neighbours vertex is constituted, determine that two ring neighbours push up according to angle group belonging to apex angle locating for the second class vertex
The corresponding watermark value of point, to correspond to the position for adjusting corresponding two rings neighbours vertex according to the watermark value of the determination, thus will
In the orderly insertion three-dimensional digital model of watermark sequence of values, the order and accuracy of watermark insertion ensure that.
In one embodiment, the second weight watermark embedding module 704 is also used to be existed according to the two ring neighbours vertex picked out
Projected position on corresponding ring neighbours side determines the numerical value of two ring neighbours' vertex correspondences；One ring neighbours side is corresponding to two
The line on two ring neighbours vertex on ring neighbours vertex；One ring neighbours vertex is connection two ring neighbours vertex and corresponding second
The one ring neighbours vertex on class vertex；It is true according to watermark value when the numerical value and watermark value that correspond to two ring neighbours vertex are unequal
Fixed projected position of the two rings neighbours vertex adjusted on a ring neighbours side, according to projected position two ring neighbours of corresponding adjustment top
The position of point.
In above-described embodiment, according to projected position and watermark numerical value of the two ring neighbours vertex on corresponding ring neighbours side
Corresponding to the corresponding relationship between the watermark value on two ring neighbours vertex in sequence, the two ring neighbours top for needing to adjust position is determined
The position on point and two rings neighbours vertex adjusted, and the corresponding position for adjusting corresponding two rings neighbours vertex, thus by watermark
In the orderly insertion three-dimensional digital model of sequence of values.
In one embodiment, module 701 is obtained to be also used to obtain the two dimensional image for being used as the first heavy water print；By X-Y scheme
As being converted into spectrum domain, corresponding spectral matrix is obtained；It takes absolute value to spectral matrix；It is taken absolutely according to default scramble condition scramble
To the spectral matrix after value, the pixel characteristic attribute for corresponding to the two dimensional image of spectral matrix is obtained.
In above-described embodiment, the two dimensional image for being used as the first heavy water print is pre-processed, and by pretreated two dimension
Image is embedded in three-dimensional digital model, improves the robustness of watermark, and converts each pixel in the spectral matrix of acquisition
Point includes all information of respective two-dimensional image, further improves the robustness of watermark.
In one embodiment, the first weight watermark embedding module 703 is also used to obtain the ring neighbours to first kind vertex
The normal direction between two normal vectors and two normal vectors that center of mass point is calculated separately according to two kinds of normal vector calculations
Measure angle；According to the pixel characteristic attribute of the normal vector angle of ring neighbours' center of mass point on first kind vertex and two dimensional image,
Determine the pixel for corresponding to first kind vertex in two dimensional image；It is adjacent according to a ring neighbours center of mass point on first kind vertex and a ring
Two normal vectors for occupying center of mass point establish the respective local spherical coordinate system in first kind vertex respectively；It is respective on first kind vertex
In local spherical coordinate system, the spherical coordinates value on first kind vertex is determined；The ranks index value of pixel each in two dimensional image is pressed
It is separately encoded into corresponding angle value according to pre-arranged code condition, the pixel characteristic attribute after being encoded；In 3-dimensional digital mould
In type, the spherical coordinates value on corresponding first kind vertex is adjusted according to the pixel characteristic attribute after coding.
In above-described embodiment, according in the normal vector angle and two dimensional image of ring neighbours' center of mass point on first kind vertex
Pixel characteristic attribute after each pixel coding determines the insertion sequence and each first kind of insertion the first heavy water print
Corresponding watermark information is embedded in 3-dimensional digital by the position on corresponding adjustment first kind vertex by the watermark information being embedded in vertex
In model, it ensure that the order of watermark insertion, and the invisibility of insertion watermark, improve the robustness of watermark.
As shown in figure 8, in one embodiment, providing a kind of extraction element of dual blind watermatking of three-dimensional digital model
800.Referring to Fig. 8, the extraction element 800 of the dual blind watermatking of three-dimensional digital model include: obtain module 801, screening module 802,
First heavy water prints extraction module 803 and the second heavy water prints extraction module 804.
Module 801 is obtained, for obtaining three-dimensional digital model.
Screening module 802, for printing the insertion conditional filtering first kind according to the first heavy water from the three-dimensional digital model
Vertex.
First heavy water prints extraction module 803, for determining that corresponding pixel is special according to the position on the first kind vertex
Levy attribute.
First heavy water prints extraction module 803, is also used to be restored according to the pixel characteristic attribute as the first heavy water
The two dimensional image of print.
Screening module 802 is also used to screen the second class vertex for meeting preset condition from the three-dimensional digital model；Institute
It states preset condition and is included in the first heavy water print insertion condition.
Screening module 802 is also used to from the two ring neighbours vertex on second class vertex, is selected and is met the second heavy water print
Two ring neighbours vertex of insertion condition；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and be not belonging to according to
Referenced vertex when first kind vertex described in the first heavy water print insertion conditional filtering.
Second heavy water prints extraction module 804, for the matched watermark of position institute according to the two ring neighbours vertex picked out
The two corresponding numerical digits in ring neighbours vertex for being worth and picking out restore the watermark sequence of values as the second heavy water print.
The extraction element of the above-mentioned dual blind watermatking of three-dimensional digital model prints insertion condition and preset condition according to the first heavy water
Corresponding first kind vertex and the second class vertex are filtered out respectively from three-dimensional digital model, according to the first kind vertex filtered out
Two dimensional image is restored, i.e., extracts the first heavy water print from three-dimensional digital model.It screens and meets from the second class vertex of screening
Second heavy water prints two ring neighbours vertex of insertion condition, and restores corresponding watermark numerical value according to the two ring neighbours vertex filtered out
Sequence extracts the second heavy water print from three-dimensional digital model, to realize the extraction of doubly time series model, improves three dimensions
The robustness of watermark in word model.In doubly time series model extraction process, independently of each other, and it is not related to being not embedded into the 3-dimensional digital of watermark
Model further improves the robustness that attack is resisted in embedded watermark in three-dimensional digital model.
In one embodiment, the first heavy water prints extraction module 803, is also used to establish the respective partial sphere in first kind vertex
Coordinate system；Obtain spherical coordinates value of the first kind vertex in respective local spherical coordinate system；Spherical coordinates value includes two angle values
With a distance value；It is separately encoded two angle values on first kind vertex according to default Gray code condition, obtains corresponding vertex
Attribute value；Obtain the two-dimensional matrix of default size；Two-dimensional matrix is null matrix；According to the vertex attribute for corresponding to first kind vertex
The ranks index value of value and two-dimensional matrix, determines the pixel in two-dimensional matrix with first kind vertex correspondence；First will be corresponded to
The pixel point value of the pixel on class vertex is substituted for the distance value on corresponding first kind vertex, obtains replaced two-dimensional matrix
Pixel characteristic attribute.
In above-described embodiment, for the first kind vertex of the first heavy water print may be embedded in, in respective local spherical coordinate system
It is middle to obtain corresponding spherical coordinates value, the pixel in first kind vertex and two-dimensional matrix is determined according to the spherical coordinates value after coding
Corresponding relationship, so that it is determined that in two-dimensional matrix each pixel pixel attributive character, according to the pixel attributive character restore
Corresponding two dimensional image, in order to extract the two dimensional image for being used as the first heavy water print from three-dimensional digital model.
In one embodiment, the first heavy water prints extraction module 803, is also used to invert random condition scramble replacement according to default
Two-dimensional matrix afterwards；It obtains and the equal-sized two-dimensional coefficient matrix of two-dimensional matrix；By two-dimensional matrix and two-dimensional coefficient matrix point
Multiply, airspace will be converted into the two-dimensional matrix after two-dimensional coefficient matrix dot product, obtain the two dimensional image for being used as the first heavy water print.
In above-described embodiment, according to the opposite treatment process of the two dimensional image for being used as the first heavy water print with pretreatment, processing
The two-dimensional matrix of replacement pixel point value, to restore the two dimensional image in insertion three-dimensional digital model.
In one embodiment, the second heavy water prints extraction module 804, is also used to be existed according to the two ring neighbours vertex picked out
Projected position on corresponding ring neighbours side determines the numerical value of two ring neighbours' vertex correspondences；Using numerical value as watermark numerical value sequence
Watermark value in column, with the numerical digit of two ring neighbours' vertex correspondences.
In above-described embodiment, corresponding number is determined according to projected position of the second class vertex on corresponding ring neighbours side
Value, and using the numerical value of each second class vertex correspondence as the watermark value of numerical digit corresponding in watermark sequence of values, with from three dimensions
The second heavy water print is extracted in word model.
In one embodiment, the second heavy water prints extraction module 804, is also used in watermark sequence of values, according to selecting
Two ring neighbours vertex out determine the watermark value for corresponding to the numerical digit on two ring neighbours vertex；Root in cumulative calculation watermark sequence of values
According to the watermark value and the corresponding total quantity of watermark value of each numerical digit that two ring neighbours vertex determine；When in watermark sequence of values
When the corresponding watermark value of numerical digit is more than one, watermark value of the maximum watermark value of total quantity as numerical digit is screened, is obtained as the
The watermark sequence of values of double watermark.
In above-described embodiment, determined according to the second class of each of three-dimensional digital model vertex corresponding in watermark sequence of values
The watermark value of numerical digit, when the corresponding watermark value of each numerical digit is more than one, by selecting the maximum watermark value of total quantity as phase
The watermark value for answering numerical digit improves the accuracy of restored watermark sequence of values, has correspondinglyd increase the second weight watermark extracting
Accuracy, to improve the robustness that attack is resisted in watermark.
As shown in figure 9, in one embodiment, the extraction element 800 of the dual blind watermatking of three-dimensional digital model further include: test
Demonstrate,prove module 805.
Authentication module 805, for obtaining the authentication image for being used for verifying the first heavy water print；Calculate the two dimensional image restored
With the degree of correlation of authentication image；When the degree of correlation is greater than the first default relevance threshold, then determine to be embedded in three-dimensional digital model
First heavy water print；And/or obtain the verifying sequence printed for verifying the second heavy water；Calculate the watermark sequence of values restored
With the degree of correlation of verifying sequence；When the degree of correlation is greater than the second default relevance threshold, then determine to be embedded in three-dimensional digital model
Second heavy water print.
In above-described embodiment, by calculating the first heavy water extracted print and/or the second heavy water print and corresponding proof diagram
As or verifying sequence the degree of correlation, and three-dimensional is compared to determine with corresponding relevance threshold respectively according to the degree of correlation calculated
Whether watermark and embedded watermark information are embedded in mathematical model, and setting relevance threshold ensure that the accuracy of judgement.
A kind of computer equipment, the computer equipment include memory and processor, and memory is stored with computer program,
When computer program is executed by processor, so that processor executes following steps: obtaining three-dimensional digital model；From 3-dimensional digital mould
Insertion conditional filtering first kind vertex is printed according to the first heavy water in type；Obtain the pixel for being used as the two dimensional image of the first heavy water print
Characteristic attribute；The position on corresponding first kind vertex is adjusted by pixel characteristic attribute in three-dimensional digital model；From adjusted
Screening meets the second class vertex of preset condition in three-dimensional digital model；Preset condition is included in the first heavy water print insertion condition；
From the two ring neighbours vertex on the second class vertex, the two ring neighbours vertex for meeting the second heavy water print insertion condition are selected；It picks out
Two ring neighbours vertex be not belonging to first kind vertex, and be not belonging to according to the first heavy water print insertion the conditional filtering first kind vertex when
Referenced vertex；Obtain the watermark sequence of values for being used as the second heavy water print；Watermark sequence of values includes at each numerical digit
Watermark value；When watermark value at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit mismatches,
The position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value.
In one embodiment, insertion conditional filtering first kind vertex packet is printed according to the first heavy water from three-dimensional digital model
It includes: according to a ring neighbours vertex on vertex each in three-dimensional digital model, respectively corresponding the ring neighbours' matter for calculating each vertex
Heart point；According to ring neighbours' center of mass point on a ring neighbours vertex on each vertex and each vertex, from three-dimensional digital model, sieve
Choosing meets preset condition and is not mutually the first kind vertex on neighbours vertex.
In one embodiment, screening meets preset condition and is not mutually the first kind vertex on neighbours vertex, comprising: root
A ring neighbours center of mass point and a ring neighbours vertex according to vertex each in three-dimensional digital model, according to two kinds of normal vector calculations
Calculate separately two normal vectors of ring neighbours' center of mass point；First kind vertex is selected from three-dimensional digital model；The first of selection
The ratio of the maxima and minima of the distance between the one ring neighbours vertex on class vertex and ring neighbours' center of mass point is less than default
Fractional threshold, the normal vector angle between two normal vectors of ring neighbours' center of mass point on the first kind vertex of selection are greater than default
Angle threshold value, and the first kind vertex selected is not mutually neighbours vertex.
In one embodiment, the two ring neighbours vertex corresponding second class vertex for existing in the second class vertex and selecting,
And the two ring neighbours vertex selected only are connected with two ring neighbours vertex on corresponding second class vertex；The two ring neighbours selected
In the triangle that vertex and two ring neighbours vertex are constituted, base angle locating for two ring neighbours vertex is acute angle；It selects
The normal vector angle of ring neighbours' center of mass point on two ring neighbours vertex be less than default angle threshold value；The two ring neighbours vertex selected
The normal vector angle of the respective ring neighbours' center of mass point in a ring neighbours vertex be less than default angle threshold value.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: will preset
Angular range is divided into multiple angle groups；The number of angle group is equal with the numerical digit sum of watermark sequence of values；Each angle group
It is corresponding with each numerical digit of watermark sequence of values respectively；Correspond to the second class vertex on two ring neighbours vertex picked out and
In the triangle for connecting two ring neighbours vertex composition on two ring neighbours vertex and the second class vertex, according to the second class vertex institute
Angle group belonging to the apex angle at place determines the numerical digit in watermark sequence of values with two ring neighbours' vertex correspondences；According to watermark numerical value
The numerical digit of sequence determines the watermark value of two ring neighbours' vertex correspondences.
In one embodiment, at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit
Watermark value when mismatching, the position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value, comprising: according to choosing
Projected position of the two ring neighbours vertex selected on corresponding ring neighbours side, determines the numerical value of two ring neighbours' vertex correspondences；
One ring neighbours side is the line on two ring neighbours vertex corresponding to two ring neighbours vertex；One ring neighbours vertex is two rings of connection
The one ring neighbours vertex on neighbours vertex and corresponding second class vertex；When corresponding to numerical value and the numerical value on two ring neighbours vertex not phase
Whens equal, projected position of the two rings neighbours vertex adjusted on a ring neighbours side is determined according to watermark value, according to projected position
The position on two ring neighbours vertex of corresponding adjustment.
In one embodiment, the step of obtaining the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print, packet
It includes: obtaining the two dimensional image for being used as the first heavy water print；Two dimensional image is converted into spectrum domain, obtains corresponding spectral matrix；It is right
Spectral matrix takes absolute value；Spectral matrix after taking absolute value according to default scramble condition scramble obtains and corresponds to spectral matrix
Two dimensional image pixel characteristic attribute.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: acquisition pair
Two normal vectors that the one ring neighbours center of mass point on first kind vertex is calculated separately according to two kinds of normal vector calculations and
Normal vector angle between two normal vectors；According to the normal vector angle and X-Y scheme of ring neighbours' center of mass point on first kind vertex
The pixel characteristic attribute of picture determines the pixel for corresponding to first kind vertex in two dimensional image；According to the one of first kind vertex
Two normal vectors of ring neighbours center of mass point and ring neighbours' center of mass point establish the respective local spherical coordinates in first kind vertex respectively
System；In the respective local spherical coordinate system in first kind vertex, the spherical coordinates value on first kind vertex is determined；Pixel characteristic attribute packet
Include ranks numerical value and corresponding pixel point value；Corresponding first kind vertex is adjusted by pixel characteristic attribute in three-dimensional digital model
Position include: that the ranks numerical value of pixel each in two dimensional image is separately encoded into corresponding angle according to pre-arranged code condition
Angle value, the pixel characteristic attribute after being encoded；In three-dimensional digital model, according to the pixel characteristic attribute tune after coding
The spherical coordinates value on whole corresponding first kind vertex.
Above-mentioned computer equipment passes through the corresponding adjustment sieve of pixel point feature according to the two dimensional image for being used as the first heavy water print
The position on the first kind vertex selected prints the first heavy water in insertion three-dimensional digital model.After insertion the first heavy water print
Three-dimensional digital model filters out the two ring neighbours vertex on the vertex as insertion the second heavy water print, adjacent according to two rings filtered out
The corresponding relationship on vertex and the watermark sequence of values as the second heavy water print is occupied, the two ring neighbours vertex that corresponding adjustment filters out
Second heavy water is printed in insertion three-dimensional digital model, to be embedded in dual blind watermatking in three-dimensional digital model, is mentioned by position
The robustness of watermark in high three-dimensional digital model.The restrict that insertion condition is printed by the second heavy water, ensure that three dimensions
The the first heavy water print and the second heavy water print being embedded in word model are mutually indepedent, do not interfere with each other, further improve the robust of watermark
Property.
A kind of computer equipment, the computer equipment include memory and processor, and memory is stored with computer program,
When computer program is executed by processor, so that processor executes following steps: obtaining three-dimensional digital model；From 3-dimensional digital mould
Insertion conditional filtering first kind vertex is printed according to the first heavy water in type；Obtain the pixel for being used as the two dimensional image of the first heavy water print
Characteristic attribute；The position on corresponding first kind vertex is adjusted by pixel characteristic attribute in three-dimensional digital model；From adjusted
Screening meets the second class vertex of preset condition in three-dimensional digital model；Preset condition is included in the first heavy water print insertion condition；
From the two ring neighbours vertex on the second class vertex, the two ring neighbours vertex for meeting the second heavy water print insertion condition are selected；It picks out
Two ring neighbours vertex be not belonging to first kind vertex, and be not belonging to according to the first heavy water print insertion the conditional filtering first kind vertex when
Referenced vertex；Obtain the watermark sequence of values for being used as the second heavy water print；Watermark sequence of values includes at each numerical digit
Watermark value；When watermark value at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit mismatches,
The position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value.
In one embodiment, insertion conditional filtering first kind vertex packet is printed according to the first heavy water from three-dimensional digital model
It includes: according to a ring neighbours vertex on vertex each in three-dimensional digital model, respectively corresponding the ring neighbours' matter for calculating each vertex
Heart point；According to ring neighbours' center of mass point on a ring neighbours vertex on each vertex and each vertex, from three-dimensional digital model, sieve
Choosing meets preset condition and is not mutually the first kind vertex on neighbours vertex.
In one embodiment, screening meets preset condition and is not mutually the first kind vertex on neighbours vertex, comprising: root
A ring neighbours center of mass point and a ring neighbours vertex according to vertex each in three-dimensional digital model, according to two kinds of normal vector calculations
Calculate separately two normal vectors of ring neighbours' center of mass point；First kind vertex is selected from three-dimensional digital model；The first of selection
The ratio of the maxima and minima of the distance between the one ring neighbours vertex on class vertex and ring neighbours' center of mass point is less than default
Fractional threshold, the normal vector angle between two normal vectors of ring neighbours' center of mass point on the first kind vertex of selection are greater than default
Angle threshold value, and the first kind vertex selected is not mutually neighbours vertex.
In one embodiment, the two ring neighbours vertex corresponding second class vertex for existing in the second class vertex and selecting,
And the two ring neighbours vertex selected only are connected with two ring neighbours vertex on corresponding second class vertex；The two ring neighbours selected
In the triangle that vertex and two ring neighbours vertex are constituted, base angle locating for two ring neighbours vertex is acute angle；It selects
The normal vector angle of ring neighbours' center of mass point on two ring neighbours vertex be less than default angle threshold value；The two ring neighbours vertex selected
The normal vector angle of the respective ring neighbours' center of mass point in a ring neighbours vertex be less than default angle threshold value.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: will preset
Angular range is divided into multiple angle groups；The number of angle group is equal with the numerical digit sum of watermark sequence of values；Each angle group
It is corresponding with each numerical digit of watermark sequence of values respectively；Correspond to the second class vertex on two ring neighbours vertex picked out and
In the triangle for connecting two ring neighbours vertex composition on two ring neighbours vertex and the second class vertex, according to the second class vertex institute
Angle group belonging to the apex angle at place determines the numerical digit in watermark sequence of values with two ring neighbours' vertex correspondences；According to watermark numerical value
The numerical digit of sequence determines the watermark value of two ring neighbours' vertex correspondences.
In one embodiment, at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit
Watermark value when mismatching, the position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value, comprising: according to choosing
Projected position of the two ring neighbours vertex selected on corresponding ring neighbours side, determines the numerical value of two ring neighbours' vertex correspondences；
One ring neighbours side is the line on two ring neighbours vertex corresponding to two ring neighbours vertex；One ring neighbours vertex is two rings of connection
The one ring neighbours vertex on neighbours vertex and corresponding second class vertex；When corresponding to numerical value and the numerical value on two ring neighbours vertex not phase
Whens equal, projected position of the two rings neighbours vertex adjusted on a ring neighbours side is determined according to watermark value, according to projected position
The position on two ring neighbours vertex of corresponding adjustment.
In one embodiment, the step of obtaining the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print, packet
It includes: obtaining the two dimensional image for being used as the first heavy water print；Two dimensional image is converted into spectrum domain, obtains corresponding spectral matrix；It is right
Spectral matrix takes absolute value；Spectral matrix after taking absolute value according to default scramble condition scramble obtains and corresponds to spectral matrix
Two dimensional image pixel characteristic attribute.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: acquisition pair
Two normal vectors that the one ring neighbours center of mass point on first kind vertex is calculated separately according to two kinds of normal vector calculations and
Normal vector angle between two normal vectors；According to the normal vector angle and X-Y scheme of ring neighbours' center of mass point on first kind vertex
The pixel characteristic attribute of picture determines the pixel for corresponding to first kind vertex in two dimensional image；According to the one of first kind vertex
Two normal vectors of ring neighbours center of mass point and ring neighbours' center of mass point establish the respective local spherical coordinates in first kind vertex respectively
System；In the respective local spherical coordinate system in first kind vertex, the spherical coordinates value on first kind vertex is determined；Pixel characteristic attribute packet
Include ranks numerical value and corresponding pixel point value；Corresponding first kind vertex is adjusted by pixel characteristic attribute in three-dimensional digital model
Position include: that the ranks numerical value of pixel each in two dimensional image is separately encoded into corresponding angle according to pre-arranged code condition
Angle value, the pixel characteristic attribute after being encoded；In three-dimensional digital model, according to the pixel characteristic attribute tune after coding
The spherical coordinates value on whole corresponding first kind vertex.
Above-mentioned computer equipment prints insertion condition according to the first heavy water and preset condition is sieved respectively from three-dimensional digital model
Corresponding first kind vertex and the second class vertex are selected, two dimensional image is restored according to the first kind vertex filtered out, i.e., from three-dimensional
The first heavy water print is extracted in mathematical model.Screening meets the two of the second heavy water print insertion condition from the second class vertex of screening
Ring neighbours vertex, and restore corresponding watermark sequence of values according to the two ring neighbours vertex filtered out, i.e., from three-dimensional digital model
In extract the second heavy water print, to realize the extraction of doubly time series model, improve the robustness of watermark in three-dimensional digital model.
In doubly time series model extraction process, independently of each other, and it is not related to being not embedded into the three-dimensional digital model of watermark, further improves three-dimensional
The robustness of attack is resisted in embedded watermark in mathematical model.
A kind of computer readable storage medium is stored with computer program, when the computer program is executed by processor,
So that the processor executes following steps: obtaining three-dimensional digital model；It is embedding according to the first heavy water print from three-dimensional digital model
Enter conditional filtering first kind vertex；Obtain the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print；In 3-dimensional digital
The position on corresponding first kind vertex is adjusted in model by pixel characteristic attribute；Symbol is screened from three-dimensional digital model adjusted
Close the second class vertex of preset condition；Preset condition is included in the first heavy water print insertion condition；Two rings from the second class vertex are adjacent
It occupies in vertex, selects the two ring neighbours vertex for meeting the second heavy water print insertion condition；The two ring neighbours vertex picked out are not belonging to
First kind vertex, and it is not belonging to print vertex referenced when insertion conditional filtering first kind vertex according to the first heavy water；It obtains and uses
Make the watermark sequence of values of the second heavy water print；Watermark sequence of values includes the watermark value at each numerical digit；When it is picking out, with
When the position on the corresponding two rings neighbours vertex of numerical digit and the watermark value at corresponding numerical digit mismatch, by the corresponding two rings neighbours of numerical digit
The position on vertex is adjusted to match with watermark value.
In one embodiment, insertion conditional filtering first kind vertex packet is printed according to the first heavy water from three-dimensional digital model
It includes: according to a ring neighbours vertex on vertex each in three-dimensional digital model, respectively corresponding the ring neighbours' matter for calculating each vertex
Heart point；According to ring neighbours' center of mass point on a ring neighbours vertex on each vertex and each vertex, from three-dimensional digital model, sieve
Choosing meets preset condition and is not mutually the first kind vertex on neighbours vertex.
In one embodiment, screening meets preset condition and is not mutually the first kind vertex on neighbours vertex, comprising: root
A ring neighbours center of mass point and a ring neighbours vertex according to vertex each in three-dimensional digital model, according to two kinds of normal vector calculations
Calculate separately two normal vectors of ring neighbours' center of mass point；First kind vertex is selected from three-dimensional digital model；The first of selection
The ratio of the maxima and minima of the distance between the one ring neighbours vertex on class vertex and ring neighbours' center of mass point is less than default
Fractional threshold, the normal vector angle between two normal vectors of ring neighbours' center of mass point on the first kind vertex of selection are greater than default
Angle threshold value, and the first kind vertex selected is not mutually neighbours vertex.
In one embodiment, the two ring neighbours vertex corresponding second class vertex for existing in the second class vertex and selecting,
And the two ring neighbours vertex selected only are connected with two ring neighbours vertex on corresponding second class vertex；The two ring neighbours selected
In the triangle that vertex and two ring neighbours vertex are constituted, base angle locating for two ring neighbours vertex is acute angle；It selects
The normal vector angle of ring neighbours' center of mass point on two ring neighbours vertex be less than default angle threshold value；The two ring neighbours vertex selected
The normal vector angle of the respective ring neighbours' center of mass point in a ring neighbours vertex be less than default angle threshold value.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: will preset
Angular range is divided into multiple angle groups；The number of angle group is equal with the numerical digit sum of watermark sequence of values；Each angle group
It is corresponding with each numerical digit of watermark sequence of values respectively；Correspond to the second class vertex on two ring neighbours vertex picked out and
In the triangle for connecting two ring neighbours vertex composition on two ring neighbours vertex and the second class vertex, according to the second class vertex institute
Angle group belonging to the apex angle at place determines the numerical digit in watermark sequence of values with two ring neighbours' vertex correspondences；According to watermark numerical value
The numerical digit of sequence determines the watermark value of two ring neighbours' vertex correspondences.
In one embodiment, at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit
Watermark value when mismatching, the position on the corresponding two rings neighbours vertex of numerical digit is adjusted to match with watermark value, comprising: according to choosing
Projected position of the two ring neighbours vertex selected on corresponding ring neighbours side, determines the numerical value of two ring neighbours' vertex correspondences；
One ring neighbours side is the line on two ring neighbours vertex corresponding to two ring neighbours vertex；One ring neighbours vertex is two rings of connection
The one ring neighbours vertex on neighbours vertex and corresponding second class vertex；When corresponding to numerical value and the numerical value on two ring neighbours vertex not phase
Whens equal, projected position of the two rings neighbours vertex adjusted on a ring neighbours side is determined according to watermark value, according to projected position
The position on two ring neighbours vertex of corresponding adjustment.
In one embodiment, the step of obtaining the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print, packet
It includes: obtaining the two dimensional image for being used as the first heavy water print；Two dimensional image is converted into spectrum domain, obtains corresponding spectral matrix；It is right
Spectral matrix takes absolute value；Spectral matrix after taking absolute value according to default scramble condition scramble obtains and corresponds to spectral matrix
Two dimensional image pixel characteristic attribute.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: acquisition pair
Two normal vectors that the one ring neighbours center of mass point on first kind vertex is calculated separately according to two kinds of normal vector calculations and
Normal vector angle between two normal vectors；According to the normal vector angle and X-Y scheme of ring neighbours' center of mass point on first kind vertex
The pixel characteristic attribute of picture determines the pixel for corresponding to first kind vertex in two dimensional image；According to the one of first kind vertex
Two normal vectors of ring neighbours center of mass point and ring neighbours' center of mass point establish the respective local spherical coordinates in first kind vertex respectively
System；In the respective local spherical coordinate system in first kind vertex, the spherical coordinates value on first kind vertex is determined；Pixel characteristic attribute packet
Include ranks numerical value and corresponding pixel point value；Corresponding first kind vertex is adjusted by pixel characteristic attribute in three-dimensional digital model
Position include: that the ranks numerical value of pixel each in two dimensional image is separately encoded into corresponding angle according to pre-arranged code condition
Angle value, the pixel characteristic attribute after being encoded；In three-dimensional digital model, according to the pixel characteristic attribute tune after coding
The spherical coordinates value on whole corresponding first kind vertex.
Above-mentioned computer readable storage medium passes through the pixel point feature pair according to the two dimensional image for being used as the first heavy water print
The position that the first kind vertex filtered out should be adjusted prints the first heavy water in insertion three-dimensional digital model.For being embedded in the first weight
Three-dimensional digital model after watermark filters out the two ring neighbours vertex on the vertex as insertion the second heavy water print, according to filtering out
Two ring neighbours vertex and the watermark sequence of values as the second heavy water print corresponding relationship, two rings that corresponding adjustment filters out are adjacent
The position for occupying vertex prints the second heavy water in insertion three-dimensional digital model, to be embedded in three-dimensional digital model dual
Blind watermatking improves the robustness of watermark in three-dimensional digital model.The restrict of insertion condition is printed by the second heavy water, is guaranteed
The the first heavy water print and the second heavy water print being embedded in three-dimensional digital model are mutually indepedent, do not interfere with each other, further improve water
The robustness of print.
A kind of computer readable storage medium is stored with computer program, when the computer program is executed by processor,
So that the processor executes following steps: obtaining three-dimensional digital model；According to the first heavy water from the three-dimensional digital model
Print insertion conditional filtering first kind vertex；Corresponding pixel characteristic attribute is determined according to the position on the first kind vertex；Root
The two dimensional image as the first heavy water print is restored according to the pixel characteristic attribute；Symbol is screened from the three-dimensional digital model
Close the second class vertex of preset condition；The preset condition is included in the first heavy water print insertion condition；From second class
In the two ring neighbours vertex on vertex, the two ring neighbours vertex for meeting the second heavy water print insertion condition are selected；Two rings picked out are adjacent
It occupies vertex and is not belonging to the first kind vertex, and be not belonging to print first kind top described in insertion conditional filtering according to first heavy water
Referenced vertex when point；According to the position on the two ring neighbours vertex picked out matched watermark value and two rings picked out
The corresponding numerical digit in neighbours vertex restores the watermark sequence of values as the second heavy water print.
In one embodiment, the position according to the first kind vertex determines corresponding pixel characteristic attribute packet
It includes: establishing the respective local spherical coordinate system in the first kind vertex；The first kind vertex is obtained in respective local spherical coordinates
Spherical coordinates value in system；The spherical coordinates value includes two angle values and a distance value；Distinguish according to default Gray code condition
Two angle values for encoding the first kind vertex, obtain corresponding vertex attributes values；Obtain the two-dimensional matrix of default size；Institute
Stating two-dimensional matrix is null matrix；According to the ranks number for the vertex attributes values and the two-dimensional matrix for corresponding to the first kind vertex
Value, determines the pixel in the two-dimensional matrix with the first kind vertex correspondence；The institute on the first kind vertex will be corresponded to
The pixel point value for stating pixel is substituted for the distance value on the corresponding first kind vertex, obtains the picture of replaced two-dimensional matrix
Vegetarian refreshments characteristic attribute.
In one embodiment, described that the X-Y scheme as the first heavy water print is restored according to the pixel characteristic attribute
Picture, comprising: invert the replaced two-dimensional matrix of random condition scramble according to default；It obtains equal in magnitude with the two-dimensional matrix
Two-dimensional coefficient matrix；It, will be with the two-dimensional coefficient matrix dot product by the two-dimensional matrix and the two-dimensional coefficient matrix dot product
Two-dimensional matrix afterwards is converted into airspace, obtains the two dimensional image for being used as the first heavy water print.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: according to choosing
Projected position of the two rings neighbours vertex selected on corresponding ring neighbours side, determines two rings neighbours' vertex correspondence
Numerical value；Using the numerical value as the watermark value in watermark sequence of values, with the numerical digit of two rings neighbours' vertex correspondence.
In one embodiment, it the position matched watermark value of institute on the two ring neighbours vertex that the basis is picked out and chooses
The two corresponding numerical digits in ring neighbours vertex selected restore the watermark sequence of values as the second heavy water print, comprising: in the water
In printing value sequence, the water for corresponding to the numerical digit on the two rings neighbours vertex is determined according to the two rings neighbours vertex picked out
Print value；The watermark value of each numerical digit determined in watermark sequence of values described in cumulative calculation according to the two rings neighbours vertex, and
The corresponding total quantity of the numerical value；When the corresponding watermark value of the numerical digit in the watermark sequence of values is more than one
When, watermark value of the maximum watermark value of total quantity as the numerical digit is screened, the watermark numerical value sequence for being used as the second heavy water print is obtained
Column.
In one embodiment, when computer program is executed by processor, so that processor executes following steps: obtaining and use
In the authentication image for verifying the first heavy water print；It is related to the authentication image to calculate the two dimensional image restored
Degree；When the degree of correlation is greater than the first default relevance threshold, then determine to be embedded in described the in the three-dimensional digital model
One heavy water print；And/or obtain the verifying sequence printed for verifying second heavy water；Calculate the watermark numerical value restored
The degree of correlation of sequence and the verifying sequence；When the degree of correlation is greater than the second default relevance threshold, then described three are determined
The second heavy water print is embedded in dimension word model.
Above-mentioned computer readable storage medium prints insertion condition and preset condition from three-dimensional digital model according to the first heavy water
It is middle to filter out corresponding first kind vertex and the second class vertex respectively, two dimensional image is restored according to the first kind vertex filtered out,
The first heavy water print is extracted from three-dimensional digital model.Screening meets the print insertion of the second heavy water from the second class vertex of screening
Two ring neighbours vertex of condition, and restore corresponding watermark sequence of values according to the two ring neighbours vertex filtered out, i.e., from three-dimensional
The second heavy water print is extracted in mathematical model, to realize the extraction of doubly time series model, improves watermark in three-dimensional digital model
Robustness.In doubly time series model extraction process, independently of each other, and it is not related to being not embedded into the three-dimensional digital model of watermark, further
Improve the robustness that attack is resisted in embedded watermark in three-dimensional digital model.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a non-volatile computer and can be read
In storage medium, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, provided herein
Each embodiment used in any reference to memory, storage, database or other media, may each comprise non-volatile
And/or volatile memory.Nonvolatile memory may include that read-only memory (ROM), programming ROM (PROM), electricity can be compiled
Journey ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include random access memory
(RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms, such as static state RAM
(SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhanced SDRAM
(ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) directly RAM (RDRAM), straight
Connect memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each step in above method embodiment, which does not limit, executes sequence, can carry out any adjustment, to make description letter
It is clean, all possible permutation and combination of step each in above method embodiment are not all described, as long as however, all these
Contradiction is not present in the permutation and combination of step, all should be considered as described in this specification.
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
Limitations on the scope of the patent of the present invention therefore cannot be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to guarantor of the invention
Protect range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (18)
1. a kind of embedding grammar of the dual blind watermatking of three-dimensional digital model, which comprises
Obtain three-dimensional digital model；
Insertion conditional filtering first kind vertex is printed according to the first heavy water from the three-dimensional digital model；
Obtain the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print；
The position on corresponding first kind vertex is adjusted by the pixel characteristic attribute in the three-dimensional digital model；
Screening meets the second class vertex of preset condition from three-dimensional digital model adjusted；The preset condition is included in institute
State the first heavy water print insertion condition；
From the two ring neighbours vertex on second class vertex, the two ring neighbours top for meeting the second heavy water print insertion condition is selected
Point；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to print insertion item according to first heavy water
Part screens vertex referenced when the first kind vertex；
Obtain the watermark sequence of values for being used as the second heavy water print；The watermark sequence of values includes the watermark at each numerical digit
Value；
Watermark value at position pick out, two ring neighbours corresponding with numerical digit vertex and corresponding numerical digit mismatches
When, the position on the corresponding two rings neighbours vertex of the numerical digit is adjusted to match with the watermark value.
2. the method according to claim 1, wherein it is described from the three-dimensional digital model according to the first heavy water
The step of print insertion conditional filtering first kind vertex, comprising:
According to a ring neighbours vertex on vertex each in the three-dimensional digital model, the ring neighbour for calculating each vertex is respectively corresponded
Occupy center of mass point；
According to ring neighbours' center of mass point on the ring neighbours vertex on each vertex and each vertex, from the 3-dimensional digital mould
In type, screening meets preset condition and is not mutually the first kind vertex on neighbours vertex.
3. according to the method described in claim 2, it is characterized in that, the screening meets preset condition and does not push up for neighbours mutually
The first kind vertex of point, comprising:
According to the ring neighbours center of mass point on vertex each in the three-dimensional digital model and the ring neighbours vertex, according to
Two kinds of normal vector calculations calculate separately two normal vectors of ring neighbours' center of mass point；
First kind vertex is selected from the three-dimensional digital model；The one ring neighbours vertex on the first kind vertex of selection and a ring are adjacent
The ratio for occupying the maxima and minima of the distance between center of mass point, which is less than, presets fractional threshold, and the one of the first kind vertex of selection
Normal vector angle between two normal vectors of ring neighbours' center of mass point is greater than default angle threshold value, and the first kind vertex selected is mutual
It is not neighbours vertex.
4. according to the method described in claim 3, it is characterized in that, existing and the two ring neighbours that select in second class vertex
Vertex corresponding second class vertex, and the two ring neighbours vertex selected only two ring neighbours vertex with corresponding second class vertex
It is connected；
In the triangle that the two ring neighbours vertex and described two ring neighbours vertex selected are constituted, described two ring neighbours tops
The locating base angle of point is acute angle；
The normal vector angle of one ring neighbours' center of mass point on the two ring neighbours vertex selected is less than the default angle threshold value；
The normal vector angle of the respective ring neighbours' center of mass point in one ring neighbours vertex on the two ring neighbours vertex selected is less than described
Default angle threshold value.
5. according to the method described in claim 4, it is characterized in that, the method also includes:
Predetermined angle range is divided into multiple angle groups；The number of the angle group and the numerical digit of the watermark sequence of values are total
Number is equal；Each angle group is corresponding with each numerical digit of the watermark sequence of values respectively；
Correspond to second class vertex on the two rings neighbours vertex picked out and the connection two rings neighbours vertex and
In the triangle that two ring neighbours vertex on second class vertex are constituted, according to apex angle institute locating for second class vertex
The angle group of category determines the numerical digit in the watermark sequence of values with two rings neighbours' vertex correspondence；
According to the numerical digit of the watermark sequence of values, the watermark value of two rings neighbours' vertex correspondence is determined.
6. according to the method described in claim 5, it is characterized in that, described when two rings pick out, corresponding with the numerical digit
When watermark value at the position on neighbours vertex and corresponding numerical digit mismatches, by the position on the corresponding two rings neighbours vertex of the numerical digit
It is adjusted to match with the watermark value, comprising:
According to projected position of the two rings neighbours vertex picked out on corresponding ring neighbours side, determine that two ring is adjacent
Occupy the numerical value of vertex correspondence；The one ring neighbours side is the company on two ring neighbours vertex corresponding to the two rings neighbours vertex
Line；The one ring neighbours vertex is the ring neighbours top on the connection two rings neighbours vertex and corresponding second class vertex
Point；
When the numerical value and the watermark value that correspond to the two rings neighbours vertex are unequal, determined according to the watermark value
Projected position of the two rings neighbours vertex adjusted on the ring neighbours side, according to the corresponding adjustment of the projected position
The position on two ring neighbours vertex.
7. method according to any one of claim 1 to 6, which is characterized in that described obtain is used as the first heavy water print
The step of pixel characteristic attribute of two dimensional image, comprising:
Obtain the two dimensional image for being used as the first heavy water print；
The two dimensional image is converted into spectrum domain, obtains corresponding spectral matrix；
It takes absolute value to the spectral matrix；
Spectral matrix after taking absolute value according to default scramble condition scramble obtains the two dimension for corresponding to the spectral matrix
The pixel characteristic attribute of image.
8. method according to any one of claim 1 to 6, which is characterized in that further include:
What acquisition calculated separately a ring neighbours center of mass point on the first kind vertex according to two kinds of normal vector calculations
Normal vector angle between two normal vectors and two normal vectors；
According to the picture of the normal vector angle and the two dimensional image of ring neighbours' center of mass point on the first kind vertex
Vegetarian refreshments characteristic attribute determines the pixel for corresponding to the first kind vertex in the two dimensional image；
According to two normal vectors of a ring neighbours center of mass point on the first kind vertex and ring neighbours' center of mass point, build respectively
Found the respective local spherical coordinate system in the first kind vertex；
In the respective local spherical coordinate system in the first kind vertex, the spherical coordinates value on the first kind vertex is determined；
The pixel characteristic attribute includes ranks index value and corresponding pixel point value；It is described in the three-dimensional digital model
The step of adjusting the position on corresponding first kind vertex by the pixel characteristic attribute, comprising:
The ranks index value of pixel each in the two dimensional image is separately encoded into corresponding angle according to pre-arranged code condition
Angle value, the pixel characteristic attribute after being encoded；
In the three-dimensional digital model, corresponding first kind vertex is adjusted according to the pixel characteristic attribute after coding described in
Spherical coordinates value.
9. a kind of extracting method of the dual blind watermatking of three-dimensional digital model, which comprises
Obtain three-dimensional digital model；
Insertion conditional filtering first kind vertex is printed according to the first heavy water from the three-dimensional digital model；
Corresponding pixel characteristic attribute is determined according to the position on the first kind vertex；
The two dimensional image as the first heavy water print is restored according to the pixel characteristic attribute；
Screening meets the second class vertex of preset condition from the three-dimensional digital model；The preset condition is included in described
One heavy water prints insertion condition；
From the two ring neighbours vertex on second class vertex, the two ring neighbours top for meeting the second heavy water print insertion condition is selected
Point；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to print insertion item according to first heavy water
Part screens vertex referenced when the first kind vertex；
According to the position on the two ring neighbours vertex picked out matched watermark value and the two ring neighbours vertex picked out it is corresponding
Numerical digit, restore as the second heavy water print watermark sequence of values.
10. according to the method described in claim 9, it is characterized in that, the position according to the first kind vertex determines phase
The step of pixel characteristic attribute answered, comprising:
Establish the respective local spherical coordinate system in the first kind vertex；
Obtain spherical coordinates value of the first kind vertex in respective local spherical coordinate system；The spherical coordinates value includes two angles
Angle value and a distance value；It is separately encoded two angle values on the first kind vertex according to default Gray code condition, obtains phase
The vertex attributes values answered；
Obtain the two-dimensional matrix of default size；The two-dimensional matrix is null matrix；
According to the ranks index value for the vertex attributes values and the two-dimensional matrix for corresponding to the first kind vertex, described two are determined
Tie up the pixel in matrix with the first kind vertex correspondence；
The pixel point value for corresponding to the pixel on the first kind vertex is substituted for the corresponding first kind vertex
Distance value obtains the pixel characteristic attribute of replaced two-dimensional matrix.
11. according to the method described in claim 10, it is characterized in that, described restore use according to the pixel characteristic attribute
The step of making the two dimensional image of the first heavy water print, comprising:
The replaced two-dimensional matrix of random condition scramble is inverted according to default；
It obtains and the equal-sized two-dimensional coefficient matrix of the two-dimensional matrix；
By the two-dimensional matrix and the two-dimensional coefficient matrix dot product, by with the two-dimensional matrix after the two-dimensional coefficient matrix dot product
It is converted into airspace, obtains the two dimensional image for being used as the first heavy water print.
12. according to the method described in claim 9, it is characterized in that, the method also includes:
According to projected position of the two rings neighbours vertex picked out on corresponding ring neighbours side, determine that two ring is adjacent
Occupy the numerical value of vertex correspondence；
Using the numerical value as the watermark value in watermark sequence of values, with the numerical digit of two rings neighbours' vertex correspondence.
13. according to the method for claim 12, which is characterized in that
The position on the two ring neighbours vertex that the basis is picked out matched watermark value and the two ring neighbours vertex picked out
Corresponding numerical digit restores the watermark sequence of values as the second heavy water print, comprising:
In the watermark sequence of values, determine that corresponding to the two rings neighbours pushes up according to the two rings neighbours vertex picked out
The watermark value of the numerical digit of point；
The watermark value of each numerical digit determined in watermark sequence of values described in cumulative calculation according to the two rings neighbours vertex, Yi Jisuo
State the corresponding total quantity of watermark value；
When the corresponding watermark value of the numerical digit in the watermark sequence of values is more than one, screening total quantity is maximum
Watermark value of the watermark value as the numerical digit obtains the watermark sequence of values for being used as the second heavy water print.
14. according to the method described in claim 9, it is characterized in that, the method also includes:
Obtain the authentication image for verifying the first heavy water print；
Calculate the degree of correlation of the two dimensional image and the authentication image that restore；
When the degree of correlation is greater than the first default relevance threshold, then determine to be embedded in described the in the three-dimensional digital model
One heavy water print；
And/or
Obtain the verifying sequence for verifying the second heavy water print；
Calculate the degree of correlation of the watermark sequence of values restored and the verifying sequence；
When the degree of correlation is greater than the second default relevance threshold, then determine to be embedded in described the in the three-dimensional digital model
Double watermark.
15. a kind of flush mounting of the dual blind watermatking of three-dimensional digital model, which is characterized in that described device includes:
Module is obtained, for obtaining three-dimensional digital model；
Screening module, for printing insertion conditional filtering first kind vertex according to the first heavy water from the three-dimensional digital model；
Module is obtained, is also used to obtain the pixel characteristic attribute for being used as the two dimensional image of the first heavy water print；
First heavy watermark embedding module, in the three-dimensional digital model by pixel characteristic attribute adjustment corresponding the
The position on a kind of vertex；
Screening module is also used to screen the second class vertex for meeting preset condition from three-dimensional digital model adjusted；It is described
Preset condition is included in the first heavy water print insertion condition；
Screening module is also used to from the two ring neighbours vertex on second class vertex, is selected and is met the second heavy water print insertion item
Two ring neighbours vertex of part；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to according to described
Referenced vertex when first kind vertex described in one heavy water print insertion conditional filtering；
Module is obtained, is also used to obtain the watermark sequence of values for being used as the second heavy water print；The watermark sequence of values includes being in
Watermark value at each numerical digit；
Second heavy watermark embedding module, for working as position pick out, two ring neighbours corresponding with numerical digit vertex and phase
When the watermark value at numerical digit being answered to mismatch, the position on the corresponding two rings neighbours vertex of the numerical digit is adjusted to and the watermark value
Matching.
16. a kind of extraction element of the dual blind watermatking of three-dimensional digital model, which is characterized in that described device includes:
Module is obtained, for obtaining three-dimensional digital model；
Screening module, for printing insertion conditional filtering first kind vertex according to the first heavy water from the three-dimensional digital model；
First heavy water prints extraction module, for determining corresponding pixel characteristic attribute according to the position on the first kind vertex；
First heavy water prints extraction module, is also used to restore the two dimension as the first heavy water print according to the pixel characteristic attribute
Image；
Screening module is also used to screen the second class vertex for meeting preset condition from the three-dimensional digital model；It is described default
Condition is included in the first heavy water print insertion condition；
Screening module is also used to from the two ring neighbours vertex on second class vertex, is selected and is met the second heavy water print insertion item
Two ring neighbours vertex of part；The two ring neighbours vertex picked out are not belonging to the first kind vertex, and are not belonging to according to described
Referenced vertex when first kind vertex described in one heavy water print insertion conditional filtering；
Second heavy water prints extraction module, for according to the matched watermark value of position institute on the two ring neighbours vertex picked out and choosing
The two corresponding numerical digits in ring neighbours vertex selected restore the watermark sequence of values as the second heavy water print.
17. a kind of computer equipment, including memory and processor, the memory is stored with computer program, the calculating
When machine program is executed by the processor, so that the processor perform claim requires the step of any one of 1 to 14 the method
Suddenly.
18. a kind of computer readable storage medium is stored with computer program, when the computer program is executed by processor,
So that the processor perform claim requires the step of any one of 1 to 14 the method.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711322951.0A CN109919821B (en) | 2017-12-12 | 2017-12-12 | Embedding and extracting method of three-dimensional digital model double blind watermark and storage medium |
| PCT/CN2018/099080 WO2019114290A1 (en) | 2017-12-12 | 2018-08-07 | Embedding method for three-dimensional digital model dual layer blind watermark, extracting method for three-dimensional digital model dual layer blind watermark, and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711322951.0A CN109919821B (en) | 2017-12-12 | 2017-12-12 | Embedding and extracting method of three-dimensional digital model double blind watermark and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109919821A true CN109919821A (en) | 2019-06-21 |
| CN109919821B CN109919821B (en) | 2020-12-25 |
Family
ID=66819518
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711322951.0A Active CN109919821B (en) | 2017-12-12 | 2017-12-12 | Embedding and extracting method of three-dimensional digital model double blind watermark and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109919821B (en) |
| WO (1) | WO2019114290A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117575880A (en) * | 2024-01-16 | 2024-02-20 | 四川数盾科技有限公司 | Blind watermark embedding/extracting method and system |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101393635A (en) * | 2008-10-27 | 2009-03-25 | 华为技术有限公司 | Watermark information embedding, extracting method and processing system |
| CN101533508A (en) * | 2009-03-23 | 2009-09-16 | 福建师范大学 | Dual digital watermark method of three-dimensional grid model |
| CN101833743A (en) * | 2010-04-01 | 2010-09-15 | 杭州电子科技大学 | Image based three-dimensional grid model blind watermarking method |
| CN101853484A (en) * | 2010-05-21 | 2010-10-06 | 天津大学 | Construction method and detection method for three-dimensional mesh non-embedded robust watermark |
| CN102521825A (en) * | 2011-11-16 | 2012-06-27 | 宁波大学 | Three-dimensional image quality objective evaluation method based on zero watermark |
| CN102622721A (en) * | 2012-03-06 | 2012-08-01 | 福建师范大学 | Three-dimensional grid model blind-watermarking method based on depth image mapping |
| US20120230396A1 (en) * | 2011-03-11 | 2012-09-13 | Mitsubishi Electric Research Laboratories, Inc. | Method for Embedding Decoding Information in Quantized Transform Coefficients |
| CN103116871A (en) * | 2013-01-29 | 2013-05-22 | 福建师范大学 | Embedding method and extracting method for double fragile digital watermarks with robustness |
| CN103996163A (en) * | 2014-05-29 | 2014-08-20 | 深圳大学 | Digital watermark embedding and extraction method and device and system |
| CN104318505A (en) * | 2014-09-30 | 2015-01-28 | 杭州电子科技大学 | Three-dimensional mesh model blind watermarking method based on image discrete cosine transformation |
| CN106709855A (en) * | 2017-03-28 | 2017-05-24 | 西安理工大学 | 3D model watermarking method based on Beamlet line feature positioning |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100940451B1 (en) * | 2008-02-25 | 2010-02-04 | 주식회사 마크애니 | Apparatus for watermarking 3d mesh model using spherical surface parameter technique and method thereof |
| CN104036448B (en) * | 2014-06-12 | 2017-02-15 | 上海理工大学 | Digital watermark algorithm for three-dimensional model |
| CN106485642B (en) * | 2016-09-30 | 2019-05-24 | 北京交通大学 | The method of visible watermark is embedded in three-dimensional grid model |
-
2017
- 2017-12-12 CN CN201711322951.0A patent/CN109919821B/en active Active
-
2018
- 2018-08-07 WO PCT/CN2018/099080 patent/WO2019114290A1/en active Application Filing
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101393635A (en) * | 2008-10-27 | 2009-03-25 | 华为技术有限公司 | Watermark information embedding, extracting method and processing system |
| CN101533508A (en) * | 2009-03-23 | 2009-09-16 | 福建师范大学 | Dual digital watermark method of three-dimensional grid model |
| CN101833743A (en) * | 2010-04-01 | 2010-09-15 | 杭州电子科技大学 | Image based three-dimensional grid model blind watermarking method |
| CN101853484A (en) * | 2010-05-21 | 2010-10-06 | 天津大学 | Construction method and detection method for three-dimensional mesh non-embedded robust watermark |
| US20120230396A1 (en) * | 2011-03-11 | 2012-09-13 | Mitsubishi Electric Research Laboratories, Inc. | Method for Embedding Decoding Information in Quantized Transform Coefficients |
| CN102521825A (en) * | 2011-11-16 | 2012-06-27 | 宁波大学 | Three-dimensional image quality objective evaluation method based on zero watermark |
| CN102622721A (en) * | 2012-03-06 | 2012-08-01 | 福建师范大学 | Three-dimensional grid model blind-watermarking method based on depth image mapping |
| CN103116871A (en) * | 2013-01-29 | 2013-05-22 | 福建师范大学 | Embedding method and extracting method for double fragile digital watermarks with robustness |
| CN103996163A (en) * | 2014-05-29 | 2014-08-20 | 深圳大学 | Digital watermark embedding and extraction method and device and system |
| CN104318505A (en) * | 2014-09-30 | 2015-01-28 | 杭州电子科技大学 | Three-dimensional mesh model blind watermarking method based on image discrete cosine transformation |
| CN106709855A (en) * | 2017-03-28 | 2017-05-24 | 西安理工大学 | 3D model watermarking method based on Beamlet line feature positioning |
Non-Patent Citations (2)
| Title |
|---|
| 唐斌 等: "基于三维网格模型的双重数字盲水印算法", 《计算机工程》 * |
| 钱逸 等: "基于顶点统计特征的三维网格模型盲水印算法", 《计算机辅助设计与图形学学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117575880A (en) * | 2024-01-16 | 2024-02-20 | 四川数盾科技有限公司 | Blind watermark embedding/extracting method and system |
| CN117575880B (en) * | 2024-01-16 | 2024-04-12 | 四川数盾科技有限公司 | Blind watermark embedding/extracting method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019114290A1 (en) | 2019-06-20 |
| CN109919821B (en) | 2020-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110472623A (en) | Image detecting method, equipment and system | |
| Lee et al. | Vector watermarking scheme for GIS vector map management | |
| CN109029381A (en) | A kind of detection method of tunnel slot, system and terminal device | |
| CN110766594B (en) | Information hiding method and device, detection method and device and anti-counterfeiting tracing method | |
| JP7257631B2 (en) | Methods of Extended Authentication of Material Subjects | |
| CN103974144B (en) | A kind of video digital watermark method of feature based change of scale invariant point and micro- scene detection | |
| CN111968027B (en) | Robust color image zero watermarking method based on SURF and DCT features | |
| CN110942076B (en) | Method and system for generating anti-counterfeiting mark of ceramic product | |
| Yan et al. | Multi-scale difference map fusion for tamper localization using binary ranking hashing | |
| CN104636764B (en) | A kind of image latent writing analysis method and its device | |
| Gou et al. | Data hiding in curves with application to fingerprinting maps | |
| Hou et al. | Detection of hue modification using photo response nonuniformity | |
| CN110659651B (en) | Tampering positioning reversible image watermarking method based on SIFT | |
| CN107305556A (en) | Device and method for 3D printing | |
| Chen et al. | Screen-cam robust and blind watermarking for tile satellite images | |
| CN114445256A (en) | Training method, device, equipment and storage medium for digital watermark | |
| CN109919821A (en) | Insertion, extracting method and the storage medium of the dual blind watermatking of three-dimensional digital model | |
| CN110349072B (en) | Watermark synchronization method in vector geographic data watermark embedding and detecting process | |
| CN116757904A (en) | Information hiding and verifying method and device based on digital image watermark | |
| Tralic et al. | Copy-move forgery detection using cellular automata | |
| CN107358072A (en) | Map vector digital finger-print copy-right protection method based on I codes and CFF codes | |
| Wu et al. | Detecting image forgeries using metrology | |
| Zhang et al. | A semi-fragile digital watermarking algorithm for 2D vector graphics tamper localization | |
| Iuliani et al. | Reliability assessment of principal point estimates for forensic applications | |
| KR20090030454A (en) | Watermark inserting method to vector data for using azimuth information and watermark detecting method to the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |