RU2413989C2 - Method for optical marking museum valuables - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: method for optical marking museum valuables, such as paintings, is realised using optical marks which are two-dimensional bar codes of the DATA MATRIX type, and involves depositing said marks on the back of the canvas of a painting in form of an adhesive label using a graphical method or directly onto the base of the canvas using a jet printer, as well as subsequent identification of the item using fragments of the image of the texture of the canvas itself with unique features or on the back of the canvas and/or the image itself on the front surface of the painting, wherein the position of the fragment and its orientation is given in form of angle coordinates of the said marking in form of a square two-dimensional bar code.

EFFECT: high degree of protection of the article from faking while reducing requirements for the volume of marks on the surface which are needed for identifying items.

4 cl, 3 dwg

Description

The present invention relates to methods for optical marking of products, in particular museum valuables, such as paintings, antiques with their subsequent identification and is intended to protect museum valuables from counterfeiting, forgery and falsification.

Formed the three most promising ways of marking rarities;

1. Bar coding in two-dimensional and three-dimensional modifications;

2. Radio frequency modification tags;

3. Labels for microdot technology (Australian technology).

The first technology of registration and inventory is well known to specialists, practical. It can be made directly in the museum. The tag serves for a long time, it is easy to read. The only limitation is the adaptation of the software, which binds the customer to the software manufacturer. The bar code is applied without harming the item. However, when reconciling, the label requires obligatory visual contact, which extremely fetters ministers in limited museum areas.

The second technology is interfaced with the RFID system - a wireless way of exchanging data between an electronic tag and a special electronic device. It works in automatic (active and passive) modes. At 3 square millimeters - such are the dimensions of the device - you can place much more information than in the barcode. In his electronic memory fragments of the image of the subject are held. You can enter information both in the open (not blocked) part of the chip, and in the closed (encrypted). Using this technology, the inventory problem is easily solved: at a certain frequency and volume of the mark, it is possible to scan hundreds of objects at a walking speed at the same time, which allows for a full inventory every day. With the best characteristics, these frequency labels (850-950 Mhz and 2.4-5 Ghz) work, except, of course, shielded rooms and expositions containing metal exhibits.

Marking the product by any of the above methods alone cannot provide reliable protection of the product from counterfeiting and theft.

Bar labels are easy to make but easy to copy, although counterfeiting methods can be proposed, such as using special inks that provide random spatial heterogeneity of reflective characteristics in different wavelength ranges and provide random spatial heterogeneity of the graphic image when printing.

Radio-frequency tags, apparently, are convenient for practical use, but in reality the reliability of their reading is not at all 100%, as advertising specialists and directors of the respective companies like to say. The reliability of their reading is affected by the characteristics of the material to which they are attached. For example, it is impossible to read radio frequency tags attached to blisters (aluminum foil) with drugs, and this is considered as a serious drawback of their use in medical practice. As you know, the pigments of many paints include metallic fillers. In addition, the radio frequency tag can be separated from the protected item and glued to a fake. Serious fears are caused by the ability of radio frequency tags to withstand electronic countermeasures - jammers and grabbers. As you know, a radio-frequency tag in response to the radiation of an activator emits a certain signal - which prevents it from recording this signal and repeating it in the same way as grabbers of a radio code do when they crack various car alarms, which are, in general, much more expensive and complex devices.

It seems obvious that reliable marking and identification of an expensive museum exhibit should be tied to the unique characteristics of the exhibit itself (this position can be seen in the above proposal to record image fragments in the radio frequency tag) It is obvious, however, that high-quality memory requires a large amount of memory - much more, than this can provide an RF tag. In addition, image comparison is a rather difficult task from the point of view of computational complexity, and it needs to be simplified while maintaining information about the location and orientation of the fragment.

So, there are many known methods of marking various objects in a variety of ways using, for example, fluorescent markers, inscriptions and holograms, the use of built-in electronic circuits, etc.

So known methods of applying substances that luminesce or fluoresce in the visible, ultraviolet or infrared regions of the spectrum, followed by their analysis using electronic devices, for example, French patents 2317844 and 2027429. The devices described in these patents, which implement the corresponding identification methods, only small things can be identified, for example, banknotes, credit cards, checks, stocks, documents that need to be placed in a special holder and moved inside l electronic device rather complicated design. Those. by the methods and devices described in these patents, it is impossible to detect and identify marks on large objects of large sizes, for example, antique musical instruments, and they do not provide secrecy of marking.

A known method of identifying objects in which an electronic tag is placed on / or in the form of an integrated microcircuit (chip) that responds to an external signal of a given type with a specific response signal (EPO application 0563713, 10/06/1993; RF patent 2152076, 06/27/2000).

There is a method of protection and identification of material objects using hidden marking, which is applied to the surface of objects in the form of a certain pattern with invisible paint that becomes visible in ultraviolet or infrared rays (RF patent 2107945, 03/27/1998).

A disadvantage of the known methods and devices for marking is that they do not allow recording identification information, but give only an answer, an authentic document or not. All these methods have their inherent significant drawback, consisting in the fact that all the objects in this series are marked with the same code. Some methods, such as embedded microcircuits, cannot be used to protect paintings or some antiques.

Any marks applied by a person can be faked by "specialists" engaged in such activities. The way out of this situation is the development of such a system for identifying objects, which would be based on unique, unique features inherent only to this sample.

A known method of identifying cultural objects, including antiques, which consists in selecting at least one surface area of the object, capturing an image of each selected surface area, and recording information about the image of said surface area as a reference for subsequent identification, during which image of at least one of the previously selected surface areas, register information about the image of each of these areas and compare obtained from reflection of an identifiable object with a registered image of the site (RF patent 2045779, 10/10/1995). This document, however, does not say anything about how those areas of the surface of the object whose images are used for identification are determined.

From RU 2205450 C1, 10/11/2001, a method for identifying antiques is known.

In the process of identifying antiques, the latter is placed on the base, illuminated with an illuminator in the form of an IR projector. The previously selected randomly selected surface area is removed with a TV camera. The resolution of existing television systems is quite large, and therefore the image with the characteristic features of the surface structure formed in infrared radiation, the infrared range, is clearly visible on the monitor screen. If these characteristic features are detected, they are compared with the characteristic features of the surface areas of the reference antique items registered in the database, and the comparison results make the item belong to.

A known method of optical marking of the product, including the selection of the optical brand, incorporation of the named brand in the product, its packaging or in a tag attached to it. The stamp is made in the form of an embossed hologram (WO 9808691; priority 30.08.1996; publication date 03.03.1998, IPC B42D 15/10) (WO 2004029915; priority 09.29.2002; publication date 08.04.2004, IPC G09F 3/02). The process of product identification in accordance with a known method is reduced to lighting the brand, its visual inspection and identification based on the similarity of the observed picture to the images on the marks of similar products of the same manufacturer or by instrumental decoding of the image restored by the hologram. In the case of a professional examination, hidden micro-inscriptions or conventional signs on a holographic mark are additionally identified by instrumental methods.

A disadvantage of the method of marking products with embossed holograms is their insufficient degree of protection against counterfeiting, since embossed holograms can be reproduced and replicated using available and widespread technologies. In addition, the personification of a specific product sample, including an optical mark in the form of an embossed hologram, is impossible without the use of additional marking tools.

There is also known a method for optical marking of a product in which a pattern formed by macrospots with a given distribution of spots by size and spatial distribution is selected as a brand (US Patent Application 20050024626; priority 20.10.2003; publication date 02/03/2005, IPC G06K 009/74). This pattern is the same for all brands of the same series. Moreover, spots additionally contain color or polarization coding due to the presence in their composition of cholesteric liquid crystals having the property of selectively reflecting light of a specific wavelength or polarization. The use of additional polarization and color coding significantly increases the degree of protection of the brand from unauthorized copying, but does not completely exclude the possibility of falsification, since it is possible to use well-known photolithographic technologies to produce a brand with a macroscopic scale of heterogeneity.

From RU 2202127 C2, 04/10/2003 a method for recognizing a fake is known.

In accordance with this method, the brand is chosen among objects made of materials with a clearly visible random pattern. This pattern is formed by macrostructural elements of a material, for example, such as wood, stone, as well as artificial materials, such as paper, plastic, glass with a special addition of dyes, foreign inclusions, irregular grains. At the same time, the random structural pattern of the brand, which was read by the scanning device and is not repeated from product to product, as well as the coordinates of the characteristic control points of the structural pattern are stored in the database of the product manufacturer.

The identification procedure in accordance with the known method of recognizing a fake product is reduced to the following sequence of operations: 1) send a request to the database for the brand image and / or coordinates of control points of the brand’s structural pattern stored in it; 2) receive information about the structural pattern by telephone, fax or the Internet from a database; 3) perform a comparative analysis of the similarity of the control points of the displayed image of the structural pattern of the brand with the structural pattern of the brand itself or compare the coordinates of the control points printed on the product with the coordinates of the control points obtained from the database. Based on a comparison of the data, an expert opinion on the authenticity or forgery of the product is accepted.

The advantage of this method is the simplicity of labeling and procedures for identifying the authenticity of the product. In addition, a similar approach is applicable to an almost unlimited class of products. However, at the current level of development of printing and copying equipment, the production of a copy or counterfeit identification mark with a clear visible random macrostructural pattern does not present great technical complexity, as evidenced by the numerous facts of falsification of banknotes and documents.

In addition, the disadvantage of the known method of optical marking of the product is the need to use for its identification a large database that stores information about each product, which becomes a significant limitation for products of mass production.

The closest in technical essence and the achieved result with the claimed method is the known method of optical marking of a product, including selecting an optical mark, characterized by the presence of heterogeneities with random spatial distribution, reading information about the distribution of the named heterogeneities and storing the received information in the database of the identification system, while choosing optical marks are produced from a number of non-repeating, optically anisotropic objects with random spaces By distributing inhomogeneities indistinguishable to the naked eye, a light beam is directed from the radiation source to the selected brand, using a polarizing device, a spatially modulated distribution of the radiation intensity of the working spectral range is generated due to the presence of the aforementioned visually indistinguishable inhomogeneities of the optically anisotropic mark, the aforementioned radiation intensity distribution is read and stored information in the database as the main product code.

In this case, the optical brand is made of an anisotropic optical material in the form of a transparent window, for example, in the form of a glass plate or a sheet of plastic, not subjected to thermal annealing and characterized by the presence of strands, irregular undulation of the surface or spatially inhomogeneous internal stresses;

The optical brand is made of a transparent material containing randomly distributed birefringent particles, for example, encapsulated liquid crystals, particles of crushed mica, birefringent microcrystals or cellulose fibers.

The optical brand is made of perforated birefringent material with a random distribution of holes in size and spatial arrangement, for example, of a transparent perforated polymer film.

The optical brand is made in the form of a transparent window from granular microcrystalline material, for example, optical ceramics or multidomain crystals, the anisotropy of which is spatially modulated according to a random law.

When marking products, an auxiliary optical, magnetic or electronic storage medium is additionally incorporated into the product, a secondary personalized product code is received in the named carrier, obtained as a result of transcoding the main product code of one of the coding functions stored in the database of the identification system, and when identifying the authenticity of the product, comparing the named secondary personalized code with the result of transcoding the main product code read upon identification, by the encoding function used previously to generate the named secondary personified code.

The disadvantage of this method is the insufficient degree of protection of the product from counterfeiting, the complexity of the method itself, and, in particular, the need to apply a coordinate grid to the surface of the product.

The objective of the invention is to increase the degree of protection of the product from counterfeiting while significantly reducing the requirements for the volume applied to the surface, preferably on the reverse side of the product, of the markings needed to identify the products.

The stated technical problem is achieved by the method of optical marking of museum values, such as paintings, using optical marks, which are two-dimensional bar codes of the Data Matrix type, including applying them to the back of the canvas of the picture in the form of an adhesive label applied graphically or directly to the base canvas using an inkjet printer, and use for subsequent identification of the product fragments of the image texture of the canvas with its unique features or to turn tnoj side of the canvas and / or the image itself on the front surface of the painting, the position of the fragment and its orientation angles set in the form of coordinates of said marking in the form of a square two-dimensional barcode. The presence of 4 unique points (angles of a two-dimensional barcode) on each of the compared images provides a solution to the problem of determining the parameters of the projection conversion of one image to another, which is a common problem when comparing photographic images.

Moreover, the position and orientation of the image fragment on the back of the canvas and / or the front side of the picture is determined in the process of decoding the specified two-dimensional bar marking using software. In addition, the database address and the key identification parameter of the image fragment are entered in the indicated two-dimensional marking.

To speed up the identification, the compressed graphic information of the fragments used is recorded in one or more two-dimensional line marks applied to the back of the picture. When using several two-dimensional marks, their relative position will also be an additional protective factor during verification. In addition, in this case, the accuracy of the projection conversion between photographic images increases, since the number of comparable points is increased.

The following is a detailed description of the claimed method of optical marking of museum property.

1) Marking is carried out using two-dimensional bar marks of the Data Matrix type (Figure 1).

2) Marking can be carried out using an adhesive label with a graphic image applied to it - Data Matrix, using an inkjet printer using both ordinary and “invisible” inks.

3) The software for decoding two-dimensional barcodes must determine with high accuracy (preferably with subpixel, i.e. in fractions of a pixel) the coordinates of the angles of the found and decoded barcode.

4) The found coordinates of the 4 angles (x, y) in the image obtained using a high-resolution digital camera provide a coordinate system on the image of the surface of the product, taking into account the above-mentioned projection transformation.

5) The obtained image of the back of the product marked with a square barcode (side size of the square is 5-10 mm) allows you to uniquely identify the product by unique texture features of the reverse surface, such as the pitch and direction of the threads, random features of the texture of the threads (knots, etc.) item), because due to random features of the texture the image of the back is similar to a fingerprint (Fig 1). A two-dimensional bar code is used to set the coordinate system on the image of the back of the picture.

6) To obtain a high-quality image of a two-dimensional ShK with revealing the texture of the surrounding reverse surface of the picture, a contrast lighting device is used (Figure 2)

7) Two-dimensional ShK, deposited on the reverse surface of the picture, is used to coordinate a fragment of the image of the front side of the picture using a device consisting of two digital cameras that simultaneously shoot the reverse with the applied ShK, and the front side of the picture (Figure 3).

8) The address of the database and the key account number in the database, where information about the fragment of the image of the picture will be stored, are written to the 2D barcode. It is possible that a fragment of the image can be compressed in a barcode format (for example, in the form of Jpeg2000, etc.).

A single label of the Data Matrix type can contain up to 1500 bytes of information, in addition, the amount of recorded information can be significantly increased through the use of the so-called Structured Addition, when information is recorded in several (up to 16 labels). Moreover, each label of the supplement contains the number of additions and the serial number of the labels in the supplement, which makes it possible to decode information in a single block in the correct sequence during decoding.

Thus, the inventive method for marking museum valuables, such as paintings, is not difficult to implement and provides 100% protection of museum valuables from falsification.

To prove the compliance of this invention with the condition of patentability "industrial applicability" we give examples of specific performance. These examples do not exhaust the essence of the proposal.

Example 1

On the back of the picture, made on canvas, a label with a two-dimensional bar code type Data Matrix is glued (Figure 1.). The canvas is always characterized by the presence of irregular (random) waviness of the threads, the presence of nodules on the threads, irregular thickness and pitch of the threads.

To identify and subsequently use the indicated features of real canvas textures, a high-quality photograph of the surface of the canvas back surface, including a glued (or printed using an inkjet printer) two-dimensional mark, is required. A high-quality photograph revealing features of the canvas texture requires a combination of the following factors:

- the sensor of the digital camera must have a sufficiently large number of pixels, currently professional digital cameras have 60-100 Mgp;

- the optical circuit of the camera should provide macro mode, i.e. shooting at close range, at which the characteristic dimensional element of the canvas (thread thickness, spacing between the threads) should correspond to a sufficiently large number of sensor pixels, currently most professional or semi-professional cameras have macro mode;

- Lighting of the photographed surface should emphasize (reveal) the characteristic features of the canvas. This problem is solved using a contrast lighting device made according to scheme 2. Experience and calculations show that this scheme provides high uniformity of illumination of the photographed surface, while maintaining the principle of contrast lighting - lighting from a low angle.

The resulting photograph is converted for further calculations into an image "in gray scale", ie 256 shades of gray. Next, a special digital processing of this image is carried out - the identification of the edges. The image of the edges contains almost all the characteristic features of the canvas and, in addition, it is easy to compress with a high degree of compression, for example, according to the JBIG algorithm (Joint Bi-Level Experts Group). Thus, the image of the edges in compressed form is written to the database.

During reconciliation, a photograph of a canvas element with a label with a two-dimensional bar code is compared with the image of edges recorded in the database. At the same time, when decoding the Data Matrix bar mark, the coordinates (x, y) of the 4 corners of the Data Matrix are found. Note that the coordinates of 4 angles (8 values) make it possible to find the parameters of the projection transformation (typical of a photographic image) for each image and superimpose one image onto another. In addition to calculating the projection transformation parameters, the so-called Distance Transform can be used to accurately calculate image alignment. The Transformation of Distances for a certain image of the edges is an image (or a two-dimensional array), where for each pixel with x, y coordinates the image (array) contains the Cartesian distance from the nearest edge. The advantage of using Distance Transformation is that to compare the image of the edges recorded in the database and the image of the edges of the fragment obtained during reconciliation, well-known methods are used to minimize the function of several variables.

Example 2

Identification of a picture by its fragment. Since nobody will allow sticking a label with a two-dimensional barcode on the front side, we stick the label on the back. To coordinate the fragment relative to the label, we use a device implemented according to the scheme in figure 3. Accordingly, with this scheme we get two photos - a photograph of a fragment of the picture and a photograph of the label. Similarly to Example 1, we select the edges in the photograph of a fragment of the picture and obtain an image of the edges, which we write in compressed form to the database. When decoding a digital photo of a label, we determine the coordinates (x, y) of the four corners of the Data Matrix symbol. Note that each corner of a two-dimensional bar code is unique in its location to the search pattern of the Data Matrix symbolism - the letter "L", so they can not be confused. As you know, when photographing a square on a digital sensor, an arbitrary rectangle is obtained in the general case. The relationship between the space of the subject (picture) and the space of images (on a digital sensor) is determined by the projection transformation, characterized by 8 parameters, which are determined precisely from the presence of two coordinates (x, y) of the four corners of the label. Thus, for coordination of a fragment of a picture, these 8 values of the coordinates of the angles or the parameters of the projection transformation calculated from them are also stored in the database, together with the image of the edges of the fragment of the picture.

If the picture is identical, we also get two photographs of the front and back sides of the picture using the same device, made according to the scheme in Fig. 3. We determine the coordinates of the 4 corners of the two-dimensional bar code of the Data Matrix type and, based on their correspondence to the coordinates of the same 4 corners selected from the database, determine the projection transformation between the image of the edges stored in the database and the image of the edges obtained from the fragment photograph during reconciliation . By the coincidence of these two images (taking into account the projection transformation), we judge the identity of the picture.

In figure 2 is indicated:

1 - Lighting LEDs;

2 - Housing of a contrast lighting device;

3 - Stops;

4 - The screen is protective;

5 - Digital camera with macro mode;

A - white scattering surfaces;

B - black absorbing surfaces;

The protective screen eliminates direct radiation from the LEDs on the photographed surface.

A lateral cylindrical surface A having a white scattering surface (realized, for example, based on beryllium white), provides illumination of the photographed surface from a low angle with uniform illumination relative to the vertical axis.

In figure 3 is indicated:

6 - Picture;

7 - Digital camera front side

8 - Back side digital camera

9 - device bed

10 - Technology emphasis

11 - Two-dimensional bar code type Data Matrix

A ¹ - the front side of the picture

B ¹ - the reverse side of the picture.

Claims (4)

1. A method for optical marking of museum values, such as paintings, using optical marks, which are two-dimensional barcodes of the DATA MATRIX type, including applying them to the back of the canvas of the painting in the form of an adhesive label applied graphically or directly to the canvas base with using an inkjet printer and using for subsequent product identification fragments of the image texture of the canvas with its unique features either on the back of the canvas and / or the image itself and the front surface of the picture, while the position of the fragment and its orientation are set in the form of coordinates of the angles of the indicated marking in the form of a square two-dimensional barcode. 2. The method of optical marking of museum values according to claim 1, characterized in that the position and orientation of the image fragment on the back of the canvas and / or front side of the picture is determined in the process of decoding the specified two-dimensional bar marking using software. 3. The method of optical marking of museum values according to any one of claims 1 or 2, characterized in that the database address and the key identification parameter of the image fragment are entered in the indicated two-dimensional marking. 4. The method of optical marking of museum values according to any one of claims 1 or 2, characterized in that to expedite the identification, the compressed graphic information of the fragments used is entered into one or more two-dimensional line marks applied to the back of the canvas.

Images