Embodiment
Below in conjunction with the drawings and specific embodiments, the utility model is described in further detail.
Holographic security technology combines with smart mobile phone coding techniques by the utility model, existing holographic anti-counterfeiting film is upgraded, the discrete codes point comprising hidden code is attached to holographic false proof picture and text, to provide hidden code holographic anti-counterfeiting film of the present utility model, this anti false film comprises anti-counterfeiting information layer 21.Information Level 21 is the grating picture and text formed by the version hot padding of contour forging nickel, and it comprises general holographic false proof picture and text 211 and hidden code 212.
First referring to figs. 1A to Fig. 1 D, the anti-counterfeiting information layer of four kinds of different arrangement forms is schematically shown respectively.As seen from the figure, holographic false proof picture and text 211 can be hologram, bar code, Quick Response Code etc., and also can be the forms of this several combination.Comprise holographic false proof picture and text 211 and hidden code 212 in described anti-counterfeiting information layer 21, hidden code 212 can overlapping, overlapping and/or apart arrangement with holographic false proof picture and text 211.
Holographic false proof picture and text 211 can diffraction go out rich color, dynamic change, hyperchannel and multidimensional light become effect holographic false proof image, distinguish the true and false having and dazzle beautiful diffraction light change by naked eyes, be applicable to the public false proof.Hidden code 212 is hidden in anti-counterfeiting information layer 21, and it can be such as discrete codes point, so when overlapping with holographic false proof picture and text 211, can not interfere with each other with holographic false proof picture and text 211, also not affect product attractive in appearance.
This discrete codes point can be use the programmable interface coding open to smart machine, the coded message exported, and can be shown it by the decoding program of smart machine (Handheld intelligent equipment of such as smart mobile phone) to hide Info, so realize dual antifalse effect.These hide Info can be such as word, image, voice document, video file etc., or their combination.
With reference to figure 2A and Fig. 2 B, schematically show the cross-section structure of two kinds of concrete forms of hidden code holographic anti-counterfeiting film of the present utility model, wherein Fig. 2 A is depicted as the hidden code holographic anti-counterfeiting film of gilding film form; Fig. 2 B is depicted as the hidden code holographic anti-counterfeiting film of composite membrane (or total transfer film) form.
When hidden code holographic anti-counterfeiting film of the present utility model is gilding film (Fig. 2 A), it comprises base membrane layer 1, release layer 5, information carrier layer 2, anti-counterfeiting information layer 21, film plating layer 3 and melt pressure sensitive glue-line 41 from bottom to top successively.
Base membrane layer 1 is the carrier of coating, and plastic sheeting can be adopted as base membrane layer.Be attached on base membrane layer 1 other coating uniform.Release layer 5 is attached on base membrane layer 1, and its raw material adopts fusing point at the wax of 80-90 DEG C usually.Release layer 5 plays dissociation in the use of gilding film.
Information carrier layer 2, for carrying holographic information, is transferred on base membrane layer 1 also dry by coating fluid through coating apparatus, is formed information carrier layer 2.By hot press printing technology, anti-counterfeiting information layer 21 can be transferred on information carrier layer 2.
High reverse--bias film plating layer 3 is attached on anti-counterfeiting information layer 21, it typically is the metal or medium with high reverse--bias performance, and can be formed by the method for vacuum evaporation.The existence of film plating layer 3 can improve the brightness of holographic film, conveniently utilizes reflected light directly to observe holographic false proof figure.Melt pressure sensitive glue-line 4 is outermost one deck of this hidden code holographic false proof gilding film, and it act as and adheres to stock surface when product uses.
When hidden code holographic anti-counterfeiting film of the present utility model is composite membrane (or total transfer film) (Fig. 2 B), it comprises base membrane layer 1, information carrier layer 2, Information Level 21, high reverse--bias film plating layer 3 and gum layer 42 from bottom to top successively.From Fig. 2 B, hidden code holographic false proof composite membrane (or total transfer film) of the present utility model is similar with the gilding film shown in Fig. 2 A, but does not comprise release layer, and when for composite membrane, base membrane layer 1 is not in use separated; And when for total transfer film, base membrane layer 1 is separated in the process of composite transferring, and can recycling.
Hidden code holographic anti-counterfeiting film of the present utility model can utilize ripe contour forging technique to produce substantially, and cost is lower.Feature of the present utility model is the making of anti-counterfeiting information layer wherein.With reference to figure 3, the technological process for making the anti-counterfeiting information layer in this hidden code holographic anti-counterfeiting film is shown.As seen from the figure, method for making mainly comprises following step:
First in step S1, by the programmable interface coding that smart mobile phone is open, output invisible is encoded, such as discrete codes point.Afterwards in step S2, by described hidden code (discrete codes point) together with holographic false proof graphic rendering; And utilize holographic shooting or photoetching technique in step s3, make the photoetching hectograph with hidden code holographic false proof picture and text; Then utilize galvanoplastics in step s 4 which, make contour forging nickel version; Utilize hot pressing reprography afterwards, by the hidden code holographic false proof picture and text print in described contour forging nickel version on information carrying layer, to form anti-counterfeiting information layer.
It will be understood by those skilled in the art that the concrete form according to anti false film, can before above preparation process step, between and afterwards, correspondingly increase suitable operation.Such as when gilding film, before by holographic false proof figure batch imprinting and copying, need information carrying layer 2 to be coated on base membrane layer 1.Anti-counterfeiting information layer 21 utilizes vacuum coating technology, anti-counterfeiting information layer 21 plates metal film or the deielectric-coating of high reverse--bias after being formed, and forms high reverse--bias film plating layer 3; Coating melt pressure sensitive glue-line 41 or gum layer 42 are on high reverse--bias film plating layer 3.Finally anti false film can be cut, to be applied on product.
The product of making like this can diffraction go out that rich color is attractive in appearance, dynamic change, hyperchannel and multidimensional light become effect holographic false proof image, not only distinguish the true and false having and dazzle beautiful diffraction light change by naked eyes, be applicable to the public false proof; Also hide coding in Information Level, the decoding program by portable equipment (smart mobile phone) shows its information hidden, and as word, image, voice document or video file etc., has double anti-forge effect, can be applicable to the fields such as anti-false trademark.
In the concrete use of gilding film of the present utility model, because when gold stamping, gilding film is subject to high temperature and certain pressure, by the effect of temperature and pressure, first release layer 5 starts melting as the part that fusing point in each layer is minimum and is converted into liquid, under the effect of peeling force, first this layer destroys, thus is separated with base membrane layer 5 by the information carrier layer 2 of beared information layer 21.Along with the increase of temperature and pressure, melt pressure sensitive glue-line 4 starts softening and occurs viscosity, and when the effect of glue-line adhesion is greater than the wax layer adhesion of melting, bronzing material is transferred on stock, base membrane layer 1 is stripped, thus realizes complete gold stamping process.
In the concrete use of composite membrane of the present utility model (or total transfer film), utilize the adhesiveness of gum layer 4, use equipment complex will to have the base material of holographic information and another base material compound and slaking.Thus the holographic anti-counterfeiting film with hidden code is transferred on stock.As described above, when holographic anti-counterfeiting film of the present utility model is composite membrane, base membrane layer 1 is not separated in recombination process; And when holographic anti-counterfeiting film of the present utility model is total transfer film, this base membrane layer will be separated in recombination process, and can be recycled.
With reference to figure 4, illustrate according to the utility model, use portable equipment (such as smart mobile phone) to identify hidden code holographic anti-counterfeiting film of the present utility model, with the system showing and hide Info of decoding.This system mainly comprises two parts: intelligent terminal (such as smart mobile phone), comprise the device for scanning and identify hidden code holographic anti-counterfeiting film of the present utility model wherein, and the data server for retrieving hidden code holographic anti-counterfeiting film of the present utility model and mate.
Be described respectively below.
Of the present utility modelly embodiment still provides a kind of device scanning and identify above-mentioned hidden code holographic anti-counterfeiting film, this device comprises Video Capture module and hidden code analysis module, wherein Video Capture module is for scanning hidden code holographic anti-counterfeiting film, completes Video Capture; Hidden code analysis module is used for carrying out Image semantic classification to the image that Video Capture module scans hidden code holographic anti-counterfeiting film obtains, and resolve the hidden code in pretreated image, be sent to a data server by wireless network after the hidden pattern encryption that parsing is obtained and retrieve and mate.Wireless network can adopt general packet radio service technology (General Packet Radio Service, GPRS) network or 3G network.
Video Capture module can be applicable to Android, iOS, Windows Phone(be called for short WP) etc. intelligent operating system, cover the equipment such as smart mobile phone, notebook computer, panel computer on the market with camera, its major function is set up complete Video Capture platform according to different operating system, guides user to use camera to scan interested material (comprising Quick Response Code, bar code or laser anti-counterfeit label etc.).Video Capture module, by such as under type scanning hidden code holographic anti-counterfeiting film, completes Video Capture:
A () selects video capturing device (camera as intelligent terminal), whether the parameters of definition camera as preview, picture format, preview frame number, preview size etc., then preview call back function is set, receive in call back function, analyze each frame, start previews, thus startup video capturing device, and timer is set;
B () camera focuses once every a setting-up time, this setting-up time by above-mentioned set timer, such as, can arrange and once focus every 1 second, focus successfully and take a sub-picture, then set up another a thread, image cursor is pointed to hidden code analysis module, resolved by hidden code analysis module; Thread (main thread) configuration state word (such as configuration state word be 1 or other numerical value except zero) of simultaneously taking pictures makes camera stop taking pictures, but camera proceeds Video Capture, focusing task, waits for the process of hidden code analysis module;
C () terminates when hidden code analysis module thread, may obtain analyzing successfully or failed result.If namely hidden code analysis module successfully resolved, also have successfully been obtained hidden pattern, then terminate this task of Video Capture module; Obtain hidden pattern if fail, then reminding user need be tried again, and can return and perform (b).No matter which kind of situation above-mentioned, when user turns back to Video Capture interface, main thread removes status word (making status word be 0), restarts the process of the 1 second focusing photographic analysis in interval.
Hidden code analysis module carries out pre-service to the image that Video Capture module scans obtains, obtain the self-defined random code word be distributed in image, to self-defined random code word pattern-recognition, analysis, use the encryption of double secret key self-defined code word, calculate the ciphertext of regular length and by wireless network, ciphertext be sent to the data server in high in the clouds.High in the clouds data server is then decrypted the ciphertext received, classification of service code in quick-searching server, the coding that coupling is suitable for also feeds back corresponding information, again utilize wireless network (GPRS or 3G) object information to be sent to the intelligent terminal of user, user can differentiate the true and false of commodity according to corresponding information and trace to the source.
Particularly, hidden code analysis module comprises image pre-processing unit, hidden pattern analytic unit and hidden pattern ciphering unit.Through focusing, the image of acquisition of taking pictures is subject to the impact of shadow change, weather, photographer, need carry out pre-service by image pre-processing unit to image.For feature of image, the present embodiment successively carries out homomorphic filtering, histogram equalization and Canny operator edge detection to image, and outstanding hidden pattern and noise in image are distributed.Respectively these 3 image processing process of homomorphic filtering, histogram equalization and Canny operator edge detection are described further below:
(1) homomorphic filtering:
The dynamic range of images that obtains of taking pictures is very large, and the gray scale of part interested is very dark, and image detail can not be recognized, adopts general gray-level linear transfonnations method effect bad.The homomorphic filtering of image belongs to picture frequency territory process category, and its effect adjusts gradation of image scope, by uneven problem that removal of images throws light on, strengthens the image detail of dark space, do not lose again the image detail in clear zone simultaneously.
The image f (x, y) of natural scene can by grading function f
i(x, y) and reflective function f
rthe product representation of (x, y).F
i(x, y) describes the illumination of scenery, has nothing to do with scenery; f
r(x, y) comprises the details of scenery, irrelevant with illumination.
f(x,y)=f
i(x,y)*f
r(x,y)
0<fi(x,y)<∞;0<fr(x,y)<1(1)
Because the two is multiplied, frequency domain separately process again cannot be transformed to, therefore do following process:
Formula (1) is taken the logarithm
lnf(x,y)=lnf
i(x,y)+lnf
r(x,y)(2)
Make to become addition relation in spatial domain, Fourier transform is got to formula (2)
Fln(u,v)=F
i,ln(u,v)+F
r,ln(u,v)(3)
In formula, F
i, ln: grading function spatially changes slowly, and its spectral characteristic concentrates on low-frequency range; F
r, ln: the spectrum concentration of reflective function is high band (having more details and edge) in image, the detail content of the scenery reflection image that reflective function describes, its frequency is in high-frequency region.
If image illumination is uneven, then on image, the mean flow rate of each several part has fluctuating.Corresponding to the image detail structure just more difficult resolution of dark space, need to eliminate this unevenness.The tonal range of grading function can be compressed, on frequency domain, namely weaken the composition of grading function, strengthen the spectrum component of reflective function simultaneously, just can increase the contrast of the reflective function of reflection picture contrast.As a result, dark space image detail on image is increased, and keeps the image detail in clear zone as wide as possible.Be multiplied by transfer function H (u, v) (homomorphic filter), low-frequency range is compressed, and high band extends.
G
ln(u,v)=F
ln(u,v)*H(u,v)
=G
i,ln(u,v)+G
r,ln(u,v)(4)
Then Fourier inversion is asked, must at corresponding spatial domain representation
g(x,y)=exp{f
-1{G
ln(u.v)}}(5)
According to different picture characteristics and needs, select different H (u, v), poor to detail contrast, differentiate unclear image homomorphic filter process after, image frame brightness ratio is comparatively even, and details is strengthened.
(2) histogram equalization:
After homomorphic filtering, image is converted to gray-scale map, due to the gray-scale value non-uniform Distribution of many images, wherein gray-scale value concentrates on the image in a minizone is very common.Histogram equalization is the method being strengthened picture contrast by each gray-scale value that again distributes equably.Image through histogram equalization is chosen very favourable to binary-state threshold.
The present embodiment uses image scale transform: the pixel between gray area [a, b] is mapped to [z
1, z
k] interval.Generally, between original image gray area, [a, b] is usually space [z
1, z
k] subspace, now, the pixel z in former interval being mapped to the function representation of pixel z ' in new district is
Above-mentioned mapping relations in fact by [a, b] interval extension to interval [z
1, z
k] on, make the more black of image black, white is whiter.
(3) rim detection:
Need the position of extracting hidden code and noise in image, namely in image there is zone boundary jumpy in gray scale.In fact, this is a typical rim detection problem.The essence of rim detection adopts certain algorithm to extract the boundary line between objects in images and background, Canny operator be neighborhood centered by pending pixel as the basis of carrying out gray analysis, realize the extraction of image border and achieved good treatment effect.Its basic step is:
1, filtering.Rim detection is mainly based on derivative calculations, but affected by noise.Wave filter also causes the loss of edge strength while reducing noise.
2, strengthen.The point that gray scale in neighborhood has significant change by enhancing algorithm highlights.Generally completed by compute gradient amplitude.
3, detect.What in some image, gradient magnitude was larger is not marginal point.The simplest rim detection is gradient magnitude threshold determination.
4, locate.Accurately determine the position at edge.
Canny operator asks marginal point specific algorithm step as follows:
1, Gaussian filter smoothed image is used.
2, with single order local derviation finite difference formulations gradient magnitude and direction.
3, non-maxima suppression is carried out to gradient magnitude.
4, detect with dual threshold algorithm and be connected edge.
Can give prominence to hidden pattern and noise distributes through pretreated image, and the final purpose of the present embodiment is extracted from noise by hidden pattern and identified.Hidden code code word in the present embodiment contains 10 radixes, and every 30 form a hidden pattern, and the quantity that can calculate all hidden patterns is 10
30individual.For obtaining the hidden pattern containing noise, first needing to carry out denoising to hidden code image, then using above-mentioned 10 radixes of the classification function identification obtained by machine learning method, finally image being scanned, differentiate and whether contain corresponding hidden pattern.
Therefore, in the present embodiment, by hidden pattern analytic unit hidden pattern extracted from noise and identified, comprise: the hidden pattern denoising containing noise in the image adopting the Markov random field model for bianry image to obtain Video Capture module scans hidden code holographic anti-counterfeiting film, use the machine learning method of linear multi-class support vector machine to carry out model training and obtain classification function, use 10 radixes of the hidden code image of classification function identification, scan to obtain hidden pattern sequence to the image after denoising.In addition, by hidden pattern ciphering unit to resolving the hidden pattern encryption obtained, such as, personal key algorithm is used the character string of hidden pattern to be encrypted.Respectively the processing procedure of hidden code image denoising and hidden code base identification is described further below:
(1) hidden code image denoising:
Image denoising herein for be hidden pattern containing noise.Consider efficiency, the present embodiment adopts the Markov random field model for bianry image.Suppose that the quantity of noise is less than real hidden pattern, can think any point x in muting hidden pattern
iy is put with on the relevant position in the hidden pattern of Noise
ithere is stronger correlativity, equally also can think consecutive point x in the hidden pattern of noiseless
iwith x
jthere is stronger correlativity.We by such point with strong correlation to being defined as one group of neighborhood, so just have the neighborhood of two types: (x
i, y
i) and (x
i, x
j).These two kinds of prioris are programmed in energy function, for (x
i, y
i), energy function gets-η x
ix
j, η is a positive constant, to (x
i, x
j), energy function gets-γ x
ix
j, γ is a positive constant.In addition, for the hidden pattern of the present embodiment, need to add an energy function and be conducive to independent appearance of putting, so complete energy function form is:
Energy function define have the joint probability distribution of a vector x and y:
In formula (8), Z is normalized factor.
Because y is the point observed, formula (8) is in fact equivalent to and defines conditional probability p (x|y), and the object of image denoising finds the probable value of maximum possible.The algorithm that native system adopts is graph-cuts, and this algorithm can ensure global convergence, has higher efficiency simultaneously.
(2) hidden code base identification
In the present embodiment, hidden pattern is made up of 10 primary image patterns, for implicit pattern in the picture effectively can be extracted, model training need be carried out in advance, obtain discriminant function, thereafter detection primary image mode sequences and coding mode sequence be carried out to the image clapped.
The machine learning algorithm that in the present embodiment, model training adopts is multi-class support vector machine (Support Vector Machine, SVM).Support vector machine is that V.Vipnik etc. is according to Statistical Learning Theory (Statistical Learning Theory, SLT) a kind of new machine learning method proposed, in the non-linear and high dimensional pattern identification problem of solution, show many distinctive advantages, in pattern-recognition, approximation of function and Multilayer networks etc., achieve good effect.Support vector machine is a kind of feedforward neural network in essence, according to empirical risk minimization, under the prerequisite making training sample error in classification minimization, improves the extensive Generalization Ability of sorter as far as possible.From the angle implemented, the core concept of Training Support Vector Machines is equivalent to the quadratic programming problem solving a linear restriction, thus a structure lineoid is as decision plane, makes the distance in feature space between two quasi-modes maximum, and can ensure that the solution obtained is globally optimal solution.
SVM method proposes from the optimal classification surface (Optimal Hyperplane) linear separability situation.Optimal classification surface requires that sorting track not only can separate faultless for two class samples exactly, and the distance between two classes will be made maximum.
If linear separability sample set is (x
i, y
i), i=1,2 ..., n, x ∈ R
d, {+1 ,-1}, y is category label to y ∈.The general type of d dimension space neutral line discriminant function is: g (x)=w
tx+b, classifying face equation is:
w
Tx+b=o(9)
Discriminant function is normalized, make all samples of two classes all satisfied | g (x) |>=1, that is, make from the nearest sample of classifying face | and g (x) |=1, such class interval just equals 2/||w||, and therefore maximum being equivalent in interval makes || w||(or || w||
2) minimum; And require sorting track correctly to classify to all samples, require that it meets exactly:
y
i[w
Tx+b]-1≥0,(i=1,2...,n)(10)
Therefore, meet above-mentioned condition and make || w||
2minimum classifying face is exactly optimal classification surface.From the nearest point of classifying face and the training sample be parallel on the lineoid of optimal classification surface makes those samples that in formula (9), equal sign is set up exactly in this two classes sample, they are called support vector (Support Vectors).According to the above discussion, optimal classification surface problem can be expressed as following constrained optimization problem, namely under the constraint of formula (10), finds a function:
Minimum value.This is a quadratic programming problem, can obtain optimum solution w
*, b
*.Then obtaining optimal classification function is:
Wherein: sgh (.) is sign function, for given unknown sample x, only need the classification that calculating formula (12) can judge belonging to x.
In sum, the present embodiment uses linear multi-class support vector machine, after the training stage completes, obtains classification function.To image-region scanning, obtain hidden pattern sequence, use key the character string of hidden for correspondence pattern to be encrypted (such as use personal key algorithm encryption), and ciphertext is sent to data service module coupling, searches for Item Information corresponding to this pattern.
In addition, this device also comprises a display execution module, for receiving the Item Information after retrieval and coupling that data server returned by wireless network, and is shown on a display screen by Item Information.
Those skilled in the art should know, this device is by programming formation one application program, such as be applicable to the application program of intelligent terminal, after this application program is installed on intelligent terminal, the video capturing device of combined with intelligent terminal device can realize the function scanning and identify above-mentioned hidden code holographic anti-counterfeiting film, is identifying the commodity true and false and other information that can also superpose layer display data server feedback on video successfully.
Of the present utility modelly embodiment still provides a kind of intelligent terminal scanning and identify above-mentioned hidden code holographic anti-counterfeiting film, this intelligent terminal comprise device as above (when by this device by programming formation one application program, be understood in this intelligent terminal and this application program is installed), intelligent terminal is provided with video capturing device and possesses wireless networking capabilities, and the mode of wireless Internet access can adopt GPRS or 3G network.Intelligent terminal can be the intelligent movable equipment such as smart mobile phone, notebook computer or PDA, and it generally all possesses this video capturing device of camera.
Of the present utility modelly embodiment still provides a kind of data server that above-mentioned hidden code holographic anti-counterfeiting film is retrieved and mated, this data server comprises database module and data service module, and wherein database module is used for stores service classification code and the Item Information corresponding with classification of service code; Data service module is used for the hidden pattern of the encryption being received certain device or equipment transmission by wireless network, hidden pattern is deciphered after carrying out error checking and correction by cyclic redundancy check (CRC) code (Cyclic Redundancy Check), classification of service code in retrieve data library module, match from classification of service code with the coding of hidden pattern adaptation and obtain corresponding Item Information of encoding, finally by wireless network, the object information after retrieval and coupling being fed back to device or the equipment of scanning recognition.Along with product increases, the record in database will rise to the order of magnitude having a strong impact on retrieval rate, adopts level indexing structure to improve query performance in the present embodiment.
Of the present utility modelly embodiment still provides a kind of system identified above-mentioned hidden code holographic anti-counterfeiting film, this system comprises scanning as above and identifies intelligent terminal and the data server retrieved hidden code holographic anti-counterfeiting film and mate as above of hidden code holographic anti-counterfeiting film.Wherein, the image of intelligent terminal shooting hidden code holographic anti-counterfeiting film, the hidden pattern in analysis diagram picture, is sent to data server by wireless network after the hidden pattern encryption that parsing is obtained and retrieves and mate; Data server receives the hidden pattern of the encryption that intelligent terminal sends by wireless network, retrieval service classification code from database module, match from classification of service code with the coding of hidden pattern adaptation and obtain corresponding Item Information of encoding, by wireless network, the Item Information after retrieval and coupling being fed back to device or equipment.Same, wireless network can be GPRS or 3G network.
During embody rule, at Quick Response Code, the hidden code of additional random on bar code or laser anti-counterfeit label, by the smart mobile phone of user, notebook computer, panel computers etc. have the intelligent terminal of camera to Quick Response Code, bar code or laser anti-counterfeit tag scan also take specimen page, the picture obtained identifies on intelligent terminal, hidden pattern corresponding to wherein article resolved by intelligent terminal, the coding of hidden pattern is sent to data server by wireless network, the corresponding informance of article is obtained after String matching, again by wireless network, Item Information (true and false information of such as commodity or additional information) is fed back to the intelligent terminal of user from data server, thus complete an identification process.
Native system can with APP(Application) form download is provided, iOS, Android and WP version can be provided.During use, videoscanning is carried out to the carrier containing hidden pattern, system can complete auto-focusing, takes pictures, pattern analysis, feedback merchandise news process.Such as: in holographic laser label, add microcode anti-counterfeiting technology, with intelligent terminal, label is scanned, APP decodes to scan image, form the interval code of a string encryption, the data server being sent to high in the clouds by wireless network again carries out Data Matching, returns corresponding data on the intelligent terminal of user after the match is successful.
The above embodiment of the present utility model, does not form the restriction to the utility model protection domain.Any various other done by technical conceive of the present utility model change and distortion accordingly, all should be included in the protection domain of the utility model claim.