WO2004070667A2

WO2004070667A2 - Method for producing security markings

Info

Publication number: WO2004070667A2
Application number: PCT/EP2004/001087
Authority: WO
Inventors: Nils Biermann; Hilmar Rauhe
Original assignee: Informium Ag
Priority date: 2003-02-05
Filing date: 2004-02-05
Publication date: 2004-08-19
Also published as: EP1597709A2; DE10304805A1; WO2004070667A3

Abstract

The majority of prior art security markings (barcodes, holograms, etc.) are easy to counterfeit, expensive to produce, also accessible to counterfeiters or are non-machine readable. The novel method overcomes said disadvantages. A random pattern is applied to the product or the label. It is highly important that said method is that a very large number of different patterns are produced and that one such pattern can be produced at low costs. The effort required to produce a specific pattern is obviously much higher (e.g. random distribution of effect pigments). A finger print in the form of a data set is extracted from the random input pattern, said finger print containing the individual characteristics of the pattern. Said finger print is individually stored for each security marking. The finger print is extracted again during authentication and is checked to see if it matches the stored finger print. The security marking can generally be used to protect against forgery e.g. for money, documents, water marks, vehicle number plates etc.

Description

Process for the production of security labels

Description:

1. Process for the production of security labels.

2.1. Previous security labels (barcodes, holograms, etc.) are mostly easy to forge, expensive to produce, also accessible to counterfeiters, or not machine-readable. The new process is intended to overcome these disadvantages.

2.2. A random sample is applied to the product or label. It is essential that a very large number of distinguishable patterns can be generated and that the economic effort to produce such a pattern is low, while the effort to produce a very specific pattern in a targeted manner is much greater (e.g. random distribution of effect pigments). A fingerprint is extracted from the random pattern read in the form of a data record which contains the individual features of the pattern.

This fingerprint is saved individually for each security identifier. During authentication, the fingerprint is extracted again and the correspondence with the stored fingerprint is checked.

2.3. The security label can generally be used as a counterfeit protection, e.g. for means of payment, documents, tokens, license plates and much more be used.

Product piracy is a current and serious problem for industry. According to estimates by the International Chamber of Commerce, 10% of the total world trade volume is now achieved through the export and import of counterfeits. The resulting economic damage is estimated at $ 100 billion annually. The sales markets of branded goods manufacturers in Eastern Europe and Asia, which are still under construction, are particularly at risk due to the dramatic decline in sales. In addition, low-quality counterfeits can irreparably damage a brand's reputation if the

BESTATIGUNGSKOPIE Quality expectations of buyers are not met. Product piracy also has negative effects on the consumer and poses sociopolitical problems. For example, product pirates with their counterfeits violate a wealth of national and European regulations that regulate consumer protection, product liability or public health. In addition, product piracy affects industrial competitiveness and the employment situation in the respective countries. In Germany, for example, the German Ministry of Justice estimates that around 50,000 jobs will be lost each year due to product piracy. In the whole of Europe, an estimated 300,000 jobs are to be affected. In addition, it is generally believed that the profits made from the sale of counterfeit products make a direct contribution to financing organized crime. The distinction between counterfeits and original products has proven to be a significant problem for customs and police authorities, as well as for the companies themselves. In addition to the loss of sales and damage to the image of the companies concerned, there is not inconsiderable potential risk for the population, since, for example, safety-critical technical components (car components, aircraft components) and medicines are also counterfeited. The labeling of products is not only required for this reason. Product labels are also advantageous from a consumer perspective, in order to be able to assert product liability claims, for example.

Various product labels are known from the prior art. Basically, a distinction can be made between open (overt) and hidden (covert) systems for product labeling in product labeling systems. Overt systems are e.g. Labels, holograms, engravings etc. Covert systems are typically molecular markers that are inserted or attached directly in or on the product to be labeled.

A fundamental problem is that overt systems are usually easy to read out, but at the same time are quite easy to forge. The main problem is that systems like labels and holograms are easy to copy or easy to readjust. In contrast, cryptographic security is now available for Covert systems (see DE-A-19914808, page 20, from lines 65 ff. And WO0059917). In principle, Overt labeling systems used for product labeling should meet the following requirements:

- Counterfeit protection

- Machine readability (for integration in logistics processes)

- cheap price

A problem that has not yet been solved is the copy protection of security markings. This makes it easy to copy simple optical features, barcodes, labels, etc. More complicated security labels such as holograms cannot be copied directly, they can be reproduced. In order to make copying more difficult, it has so far typically been avoided to hide the security indicators (e.g. features only visible in the UV or IR range), or to encrypt information and then apply it in encrypted form to the respective product (e.g. matrix code, its Contains data in an encrypted manner). However, both strategies are relatively easy to circumvent for counterfeiters. In the first case, you perform a "scan" of different wavelengths and thereby find the wavelength in which the security label is openly visible. This can then simply be photographed and imitated or copied. In the second case, it is even easier. In this case, the - encrypted - information simply copied like any other. The forger does not need to know the actual content of the information. For this reason, all "key lock" "security" principles do not work in terms of copy protection: the copy lock fits Key of the authenticating body just like on the original.

In principle, products and goods have been tracked for a long time within logistics chains, for example with forwarding agents, postal and parcel deliverers. If you follow such chains, you can basically understand where which goods come from and for which destination. Appropriate interfaces to the product tracking system enable a certifying body (e.g. customs) to check where the goods come from and whether they are legal. One problem, however, is to check whether the labeling of goods is actually correct and whether goods have been brought into the logistics chain, removed from it or exchanged without authorization. The only thing that can help here is a security label with which goods are labeled and which can be used to identify them without any doubt. It is particularly important to protect against counterfeiting that the security label cannot be copied or reproduced. To date, however, no such feature is known from the prior art. So far, all security features used for product labeling have been imitated or counterfeited. Nor can it be guaranteed that security labels will not fall into the hands of counterfeiters.

The application US 4,218,674 describes a document security system in which magnetic fibers are incorporated into the document and are randomly distributed in the document. The magnetic fibers are detected with a magnetic head. An identification code ("fingerprint") can be calculated from the randomly distributed magnetic fibers by swiping a magnetic reading head over the marked document and logically (AND) linking the received signals with a time signal so that the signals (ordinate) versus time (abscissa) The application US4,568,936 describes a system based on the detection of random patterns of microscopic paper fragments with different light transmission, the application US4,820,912 describes a system for securing banknotes and credit cards based on randomly distributed conductive fibers can be detected with microwaves.

The application US5,354,097 describes a labeling system based on the use of spatially randomly distributed microfibers, which are characterized by an optical system. The registration is the basis of the 3DAS technology advertised by Unicate BV (Netherlands).

The application US5,521,984 describes a labeling system with the aid of video microscopes and computers.

It is striking that the actually decisive procedural steps that are absolutely necessary for the functioning of the described system are not disclosed in the applications. These include the procedural steps that reveal which patterns are recognized at all, how this is done and how it is ensured that these patterns are reproducibly recognized. Recognition in particular is a difficult problem, since on the one hand it has to be specific enough not to be “wrong” To deliver positive results, on the other hand, must be sufficiently tolerant to avoid any

To deliver "false negative" results. The process steps for recognition must take into account that there are unintentional changes in the

Target object may have come (e.g. due to scratches on the surface or the like) and that the respective detection devices themselves are never completely identical and cannot be positioned identically. With optical recording devices, e.g. Precautions against slipping or twisting the comparative image compared to the

Original recording can be considered.

The application US5, 354,097 indicates that the reading system recognizes polygons that arise from the superimposition of the microfibers in the room. How this detection works, however, and how, in particular, recognition is ensured, is not disclosed.

The application US5,521,984 describes in a trivial manner the idea of video microscopes and

Computer hardware for recording and storing characteristic random patterns of

To use objects. However, it is not disclosed how the functionality of the

System basically required image comparison should work at all and how an unambiguous identification of objects is guaranteed at all.

The application US5,521,984 speaks of objects without

Capture labeling. However, how this is to be done is not disclosed.

The present invention is therefore based on the object of providing a method for product identification which meets the above-mentioned requirements and overcomes the disadvantages of the prior art. In particular, the object is achieved to provide an inexpensive security identifier that cannot be copied or imitated, or can only be copied or imitated with uneconomically high expenditure.

The following terms, symbols and definitions are used for the method disclosed below:

This object is achieved according to the invention by a method for producing security labels, which is characterized in that Applies a unique, unordered random pattern to the object to be labeled (eg product or label), or uses a random pattern already on the object to be labeled for the further steps described. The random pattern is not tied to any particular material or medium, the only decisive factor is that it contains disordered information, that it can be detected with a device, that a very large number of distinguishable patterns can be generated randomly and as evenly as possible, and that the economic outlay is any Creating such a pattern is low, while the effort to create a specific pattern is much larger, or it is completely impossible. For example, small amounts of effect pigments can be added to a printing ink, and the dot pattern of these pigments at a very specific point in the print can be made visible under the microscope and used as a random pattern. Random patterns are preferably obtained with the aid of detail or signal amplification. For example, the printing of letters by printing machines appears to the normal eye regularly. However, when the detail is magnified by a microscope, random deviations and irregularities become apparent. The same applies to all signals that can be detected by suitable readers in different resolutions. For example also for electromagnetic signals. Such a pattern obtained by signal amplification can be detected by a suitable reading device, but the more difficult it is to imitate the higher the signal amplification and the more sensitive the detection device. Ideally, the random pattern is so complex that it cannot be reproduced at all. If resolution or signal amplification can be scaled up, the copying effort can also be increased later and adapted to the respective security requirements.

Reads the chaotic random pattern on the object or applied in step I into a computer with a reader.

Extract a fingerprint from the random pattern that contains the individual characteristics of the pattern. As a rule, this step is accompanied by a considerable reduction in the amount of data. IV. If desired, affix a unique identification number to the object, preferably in a machine-readable form (eg using a barcode, matrix code, transponder).

V. The fingerprint extracted in step III is stored in a data processing device or a machine data storage device.

VI. If desired, the fingerprint extracted in step III is stored directly in or on the object to be identified, preferably in machine-readable form, e.g. as a barcode, matrix code or as a transponder. In order to obtain additional security, the extracted information is preferably encrypted beforehand.

VII. If desired, the fingerprint extracted in step III, together with the unique identification number, is stored as a pair in a data structure (database).

VIII. For identification or authentication of the marked products, as in steps II and III, read the random pattern from the product, extract the fingerprint and compare it with the fingerprint originally recorded.

Process variants:

There are 3 process variants of the main process described above, which can be used depending on the respective application need:

Variant A, consisting of process steps I, II, III, V, VIII, in which a fingerprint is extracted from the random pattern and this is stored in a data structure (database) that contains the valid fingerprints. To authenticate an object, its fingerprint is determined and it is checked whether it is contained in the database. In addition to the fingerprint, other features such as batch number, article number, year of birth, etc. can be used as the key to the database.

Variant B, Consisting of process steps I, II, III, IV, VII, VIII in which, in addition to variant A, an identification number is applied to the object to be identified (e.g. as a barcode or matrix code), which is stored together with the fingerprint in a data structure (database) becomes. This simplifies and speeds up access to the respective fingerprints for the comparisons required for checking.

Variant C, consisting of process steps I, II, III, VI, VIII in which the extracted fingerprint is saved directly in or on the object to be identified, e.g. in machine-readable form as barcode or matrix code. With this variant, a reader with recognition software can subsequently extract the fingerprint for authentication and compare it directly with the fingerprint attached to or on the object. A great advantage of this method is that the authentication is also offline, i.e. works without any connection to a database. However, it is important that the fingerprint is encrypted before it is attached to or on the object. An asymmetric encryption method (e.g. RSA) is used for this, which is characterized by key pairs. One key is used for encryption and one for decrypting information. Even if a reader falls into the hands of counterfeiters and they are able to find out how the device works, they do not get the key that is used to encrypt the fingerprints because the reader only contains the key for decryption , However, the key to encryption is necessary to create labels that are authenticated as authentic. This key is only required for the production of the markings and can therefore easily be kept secret.

Variant D, consisting of variant A, B or C with additional use of the random pattern for storing pre-coded, i.e. non-random, information, eg for differentiating batches. For example, in a random pattern of two distinguishable particle types, the quantitative ratio of these particle types to one another can be used to store information. This information can be determined by appropriately mixing the particles before the production of the random samples.

Explanations to steps I and II:

An example of realizing random patterns is the production of mixtures of distinguishable microparticles. Such a mixture can e.g. made of clear lacquer

Make the base of ethyl acetate and metal microparticles (zinc, 1-50 μm) and apply it to a black background. Under side lighting takes one

Microscope at approx. 400-1000 times magnification an image of the random pattern that shows a characteristic dot pattern of the reflecting metal particles. This image is sent to a computer in electronic form.

Since the reflection behavior of the irregularly shaped metal particles differs under changed lighting conditions, one can

Random pattern a fake can be made more difficult by one

Random pattern multiple images can be taken under irradiation of light from different sides. A counterfeiter is also forced to do the exact thing

Copy the reflection behavior of the particles.

The following implementations show further exemplary configurations of the

Random pattern:

1.Random distribution of particles (e.g. colored particles, pigments, effect pigments, sand, dust, crystals (e.g. salt crystals of different colors), ferromagnetic, magnetizable, permanently magnetic, fluorescent, phosphorescent, iridescent, opalescent or radioactive particles)

2.Surface structures (e.g. cracks, scratches, bumps, notches, bubbles, solidified foam, solidified drops)

3. Random distributions of layer thicknesses

Electromagnetic random resonance frequencies (transponders)

random impedances

Patterns that arise from the crystallization of substances

Pattern of spinodal segregation

Structures of grain boundaries in crystals

Structures of phase boundaries of multi-phase material systems Random patterns can be read or detected using the following devices, for example:

1. microscopes (electron microscopes, atomic force microscopes, transmission electron microscopes (TEM), magneto-optical eddy current microscopes etc.)

2. Interferometer (laser interferometer etc.)

3. Mass spectrometer

4. Chromatographs (gas chromatographs etc.)

5. Atomic resonance spectrometers

6. Frequency analyzers

7. Tomograph

8. Spectrometer

The approaches to product labeling based on the current state of technology have in common that they are based on labeling with ordered patterns. Examples are optical patterns such as holograms and pictorial information that are only visible in a certain wavelength spectrum, barcodes, magnetic patterns on magnetic strips, etc.

Common to the previous security indicators is that information is first generated, which is then applied to an object in a targeted process. The invention described here, on the other hand, is based on the fact that the information only arises accidentally when it is applied to the object or through the production of the object itself in a process which is not targeted in relation to the information. So it is not based on the presence of regular patterns, but instead on the "scanning" of irregular, random, chaotic material arrangements, such as those found in or on the surface of an object. One example is the random arrangement of color pigments on a colored one Surface that can be seen in the microscope at high resolutions. The information actually to be recognized is therefore not “pre-coded” and applied to an object in the method according to the invention, but “post-coded”: an unordered random pattern is read (“scanned”) and information is extracted therefrom , The extracted information is then used as a fingerprint that uniquely identifies the respective object. Such Random patterns can typically be obtained by using a reader with a high resolution range.

The difference between "pre-coded" ordered and chaotic patterns can already be seen in the amount of data: While random numbers or pseudo-random numbers are typically 32 or 64 bits, linear barcodes <100 bytes, matrix codes <100kByte contain data, the data amounts of random patterns are already with a low resolution of the detection devices used, several megabytes (about a million times more information). Even a microscope image taken with a commercially available digital camera contains several MB of data. The data volume f (m, n) increases for a resolution m and a number From n degrees of freedom to f (m, n) = m ⁿ , ie exponentially with the number of degrees of freedom. For an optical detection device with two degrees of freedom (n = 2), the data volume increases quadratically with the resolution. It can be seen that a picture with one Data volume of 1 megabyte with 1000 times higher resolution already 1000 ² times more data, in this case it already contains 1 terabyte of data. In the described method, an amount of data is extracted as a fingerprint from this vast amount of data, which is comparable to conventional identification data carriers (see above). However, there are mapping functions from random patterns to fingerprints (for example through the algorithm described below), but not vice versa from fingerprints to random patterns. This is exactly what a counterfeiter would need in order to be able to copy or copy efficiently. If the resolution is increased, the image area will generally not be expanded accordingly (from 1 megabyte to 1 terabyte in the example above), but you can now select a section from a larger image area that is not known to a third party. You can take advantage of this fact and use it for steganographic encryption. In the example, only the encryptor knows where the relevant 1 megabyte image section is located in the 1 terabyte image space. For decryption, it is therefore necessary to find the "right place" again. The information required for this is a secret key, which can be passed on to third parties as required and enables them to decrypt. The encryption strength can be increased even further more degrees of freedom can be used Another degree of freedom is, for example, the z-axis (third spatial dimension) Another degree of freedom is, for example, the number of different particles used to extract features be used. For example, particle mixtures can be used to generate chaotic patterns, in which k (ke IN, k> 1) of n (ne IN, 0 <k <n) distinguishable particle species are used for feature extraction (the decoding information would accordingly include which Particles are relevant). The encryption strength therefore increases in proportion to the size of the data volume, for which we have determined an exponential growth for the number of degrees of freedom above.

Explanations to step III:

An exemplary method for extracting a fingerprint for the above-mentioned mixture of clear lacquer and metal microparticles uses common methods of digital image processing.

1. Segmentation

The reflective metal particles create a pattern of dots (features). The dots are separated from the background by a thresholding. All pixels whose brightness lies above a threshold value are assigned to a point and colored black. The other pixels remain white. A black and white image is created from the original image (Figure 1 and Figure 2).

In the case of colored images, the color information can also be used to support segmentation and also to differentiate between different particle types.

2. erosion (erosion)

All black pixels that are not completely surrounded by black pixels are deleted. This will remove noise from the picture (Figure 3).

3.Dilation (dilation)

All pixels in the vicinity of which there is at least one black pixel are set to the color value black. With this operation, the points are restored to their original dimensions (Figure 4).

4. Detection of the points

The image is searched line by line. Whenever a black pixel is found, the associated point is filled with a filling algorithm (e.g. flood fill as in Foley, V. Dam, Feiner, Hughes, Computer Graphics Principles and Practice, 2nd Ed., Addison Wesley).

FloodFill (x, y, oldColour, newColour) {if (readPixel (x, y) == oldColour) {writePixel (x, y, newColour); FloodFill (x, y-1, oldColour, newColor) FloodFill (x, y + 1, oldColour, newColor) FloodFill (x- 1, y, oldColour, newColor) FloodFill (x + 1, y, oldColour, newColor)

}} The pixels of the point are counted as a measure of size, and an average coordinate (Centroid) is calculated from their coordinates, which is the position of the point.

From these positions, a further average position is weighted according to the size of the points. This position is used as the origin of the coordinate system, according to which the coordinates of the points together with their sizes as

Fingerprint can be saved (Figure 5).

By using steps 1-4 you get a table of the coordinates and sizes of the points. This table represents a data structure that can be further coded for data storage (e.g. in a database), e.g. as arrays of floating point numbers. This table, or the coded form of the table, represents the actual fingerprint. It should be noted that the method shown is only one way of extracting fingerprints. In principle, any feature extraction methods and any data structures can be used for this. For example, the fingerprint of a random layer thickness distribution can be saved as an array of x, y, and z coordinates. A random pattern of scratches can be saved as a line graph, a random pattern of frequency oscillations can be saved as a series of floating point numbers, etc.

Explanations to step VIII:

The most important prerequisite for the method described here to work is that the production of a fingerprint from a random pattern is reproducible, ie that one and the same fingerprint is obtained again and again from one and the same random pattern. This is made more difficult by the fact that the random patterns are random Deviations are subject. These include, for example, signs of wear, physical and chemical influences, but also deviations in the detection of the random pattern.

Most of the measurements can only be reproduced to a certain degree. For example, when using imaging methods, a scanned image can be differently illuminated, tilted, shifted or rotated in several recordings.

In order to ensure reproducibility, precautions must be taken to create deviation tolerances in order to recognize them

To guarantee random patterns and to avoid "false negatives"

Recognition must be sensitive enough to avoid "false positives". This

The reproducibility problem typically also arises in all processes

Material testing and the more sensitive the measuring system is, the more it turns out.

The following measures can be taken to ensure the best possible reproducibility:

1) The random pattern used as a characteristic must be as robust as possible against external influences. It is sufficient to use materials that are designed for this anyway. For example, Many surface coatings have already been optimized for abrasion resistance, UV resistance, etc.

2) The measurement of the random pattern must always take place under conditions that are as identical as possible. To this end, all degrees of freedom in the measurement (device parameters, illumination, etc.) are eliminated as far as possible.

3) The algorithm that compares random patterns and decides whether the compared random patterns are to be regarded as identical must contain the best possible pattern recognition / heuristic that can distinguish between insignificant (or even typical) changes and essential differences.

The method described below for comparing two fingerprints shows an example of a solution to this problem for the particle and dot patterns described above by way of example. For these, in particular, there may be points that are not recognized as such with every extraction of the fingerprint, because they are at the detection limit. There are also points that once together as one big Point are captured, and separated once as several smaller points. A slight variance of the positions of the points among one another must also be taken into account.

1) Search for pairs of corresponding points

A pair of corresponding points are two points from two fingerprints, which are probably due to the same metal microparticle of the same random pattern. A comparison of their characteristic properties determines whether two points count as such a pair. If, for example, the match of the size of the points, the distances and directions to their closest neighbors and their sizes exceeds a threshold value, the two points are considered to be a corresponding pair. Another criterion is the distance between the two points, which must not be too large. If not enough corresponding points are found, the fingerprints are not considered identical.

2) Overlaying the fingerprints

The positions of the points of one of the two fingerprints are shifted and rotated so that the sum of the distances between the points of the corresponding pairs is then minimal.

3) Assessment of similarity

As a final assessment of the similarity of the fingerprints, all corresponding pairs of points are searched again. The search for a corresponding point in the other fingerprint is limited to a very narrow spatial area. The final assessment of similarity is based on the ratio of the number of points that are part of a corresponding pair to the total number of points of both fingerprints. If the ratio exceeds an adjustable limit (e.g. 90%), the fingerprints are considered identical. The average distance between the points of the corresponding pairs can be used as a further criterion. Advantages of the procedure:

a) The random sample can be produced easily, cheaply and in high throughput. A simple, low-quality printing device without self-intelligence is sufficient for this, which can print random distributions of mixtures of particles. b) In the simplest case, no random pattern has to be applied at all, but there is already a usable pattern on the object. c) Random samples and test code (variant C) can be applied separately from each other, so that e.g. B. a packaging manufacturer applies random samples in high throughput while a second instance determines and applies the test codes. d) The security label can be produced in high throughput as a "token", each containing a random sample and test code (variant C) and can be applied to any object at any time. e) The security of the security label is scalable. The security can thereby be increased continuously That encoders and test devices are made more sensitive. For the determination of the fingerprint / test code ever higher resolutions of the random pattern can be used, whereby the counterfeiting effort increases more and more. f) The validity can only be checked on the object with the test device The connection to a database is not required for variant C. g) The validation of the security identifier is very simple, for example, a security identifier can be devalued a) by destruction b) by blocking or deleting the corresponding database entry. H) The method described according to the invention f functions according to the principle of a fingerprint comparison of physical features or physical authentication (see the applications US4,218,674, US4,568,936, US4,820,912, US5,354,097, US5,521,984). An embodiment of this principle was described by SR Pappu (Pappu, Srinivasa Ravikanth, Physical One-Way functions, 2001, MIT Press and Pappu, R., et al. 2002. Physical one-way functions. Science 297 (Sept. 20): 2026-2030). Pappu's method is based on so-called "physical one-way functions", the extraction of 2400-bit identifiers ("digital fingerprint") from the radiation of an unordered, transparent, light-scattering medium (randomly distributed Glass beads in epoxy resin) with laser light and the digital compression (Gabor hash) of the speckle patterns ("speckle pattems") obtained by this radiation to 2400-bit identifiers. SR Pappu deserves to have shown on the basis of his method variant that this Principle "physical authentication" follows cryptographic principles and thus security features based on the principle of secrecy ("security by obscurity") are far superior, since it has the advantage of higher security and mathematically quantifiable security. Registrations before SR Pappu must, if they work , intrinsic hash functions and identification algorithms. Unfortunately, these are not disclosed in the applications mentioned without exception. Apparently, the authors were not aware that each of them is a real cryptographic process or that a real cryptographic process is based on the secrecy of the respective key ( wavelength used , Positions of radiation source and detector, etc.), not based on the secrecy of the method itself. i) Compared to the "physical authentication" methods listed under h), to which reference is hereby made in full, the method described according to the invention has the following advantages or combinations of advantages:

- The method described according to the invention also works for non-transparent objects and in incident light;

- The method described in the invention does not require coherent light (laser);

- The method described according to the invention does not require a micrometer-precise positioning mechanism and therefore no complex, error-prone and, moreover, expensive hardware; In particular, the method described according to the invention can achieve at least a similarly high level of security as that of the SR Pappu method, but with extremely reduced hardware and time expenditure. On the one hand, this is due to the fact that an amplifier, preferably optical magnification (lens system), is used. The number of distinguishable patterns (the measure of security or the amount of counterfeiting) increases quadratically with the magnification factor. On the other hand, this is due to the more complex image processing in the method according to the invention, which provides better amplification of the actual signal (signal / noise ratio) compared to the Gabor hash (SR Pappu). In the case of SR Pappu's procedure, the problem with the author is particularly problematic "Avalanche effect" (avalanche effect) which is particularly preferred. This has the effect that even the smallest deviations in the positioning of the probe (eg shift by approx. 60 μm) or minimal changes in the random pattern lead to completely different identifiers, so that the random pattern is not recognized It is precisely this effect, however, that makes the identification system extremely susceptible to actually irrelevant disturbances. A particular advantage of the method according to the invention is that the deviations can be several orders of magnitude higher, which saves considerable positioning effort choose different optical magnifications from commercially available devices, for example, deviations of 600 μm and 6 mm can also be achieved, if desired. A further advantage of the method according to the invention is that the magnification factor is described with most according to the invention level probes can be parameterized (scalability); the method described according to the invention does not require any identification for certain objects. For example, euro notes can be authenticated without any additional labeling and thus without any change to the previous printing process (see Figure 17 and comment). Process variants A and B described according to the invention are particularly suitable, process variant B being particularly preferred. At the same time, the “retrofitting” with the method according to the invention would result in a considerably higher level of security against counterfeiting than is the case today

The use of particles, for example the use of platelet-shaped interference pigments with a thickness of, for example, 1-3 μm, allows the doping of printing inks. By adding or dispersing the marker, these become security printing inks. A doped printing ink can also be used to produce very thin coatings, e.g. the layer thicknesses <= 10μm typical for containers and packaging, even layer thicknesses <= 5μm, e.g. layer thicknesses between 1 and 3μm. For comparison: The 3DAS process uses fibers with a thickness of 38μm, the process from RS Pappu derives its strength precisely from the combinatorics of many superimposed layers, which easily reach a thickness> = 80μm and are only suitable for transmitted light anyway. In contrast, the interference pigments described according to the invention are not only for particularly thin ones Layers are suitable, but also have the advantage of generating 3-dimensional random patterns by randomly tilted particles after application of a doped printing ink in the room planes, which result in different patterns when the radiation source or probe is translated or rotated (see Figure 9 and Figure for comparison 10). The method described according to the invention is therefore particularly suitable for coating objects by means of common printing methods, in particular screen, gravure and offset printing, and is therefore particularly suitable for bulk articles, packaging and containers;

- Method variant B has the particular advantage of requiring only one (instead of n) comparisons in a database with n stored fingerprints, since only a comparison with the fingerprint stored in the database under the identification number is necessary, which is extremely time-consuming and time-consuming for authentication saves;

- Method variant C has the particular advantage of allowing authentication without database access (offline) and therefore also works in environments that do not allow a connection to a database or would make it very complex;

- For the reasons mentioned, the method described according to the invention is particularly suitable for the labeling of bulk articles, the advantages being in particular the combination of high (cryptographic) security and comparatively little effort and thus low costs.

Examples of use:

1) Counterfeit-proof means of payment

Counterfeit-proof means of payment (eg banknotes) can be produced as follows: a) The means of payment are provided with a random sample at a precisely defined point. This random pattern can be optical (eg random distribution of pigments), magnetic (random distribution of magnetic particles) or electromagnetic (random distribution of frequency resonators). This random pattern serves as a certificate of authenticity and copy protection. b) The certificates are "scanned" with the detection device corresponding to the marking (microscope, eddy current probe, frequency scanner) and stored in a database. c) The means of payment equipped with a random sample are copy-protected because the certificates are described in a way as described above d) The means of payment protected in this way can be deactivated and tracked by simply locking the data records in question in the database or transmitting the last reading locations using suitable readers Test devices that check the authenticity of the means of payment only need access to the database.

For simplification, machine-readable serial numbers (e.g. barcodes) can be attached to the means of payment, which are saved together with the certificate.

For means of payment such as banknotes, for example euro notes, the method described according to the invention even works without any identification. The source of random patterns is the effect field formed from interference pigments (see Figure 17 a), which shows a random distribution of reflecting particles under the microscope. The effect field is recorded with a video microscope used as a probe at a defined magnification (for example, but not exclusively, magnifications such as 20, 50 or 100 times magnification are suitable) at an optional, but defined, position and, as described according to the invention, processed into a fingerprint , In one embodiment as process variant B, at least one random pattern is extracted from the effect field and the serial number of the bill is used as the identification number under which the fingerprint extracted from the effect field is stored in this database (see Figure 17 and comment on this). Particular advantages of the method described according to the invention are that a) the banknote has an additional security feature with considerably higher (cryptographic) security than the previous features b) for this additional security feature, the actual printing process of the banknotes does not have to be changed, which makes the security feature considerably less expensive than other security features and c) if desired, banknotes can be traced and, if desired (e.g. by locking the data record), can be deactivated without these banknotes to move in.

2) Fingerprint for painted objects

Another application example is individual labeling for painted objects (e.g. machines and vehicles). It is sufficient to spray an object to be marked in at least one predetermined location with an uneven layer of lacquer (for example, simply by applying “any” random layer with a spray gun). Such a layer will have irregularities in the layer thickness If so, the sequence of measured layer thicknesses results in a sequence of floating point numbers, which can be saved as a unique identifier and subsequently identified again.

3) Deposit return system

A deposit redemption system can be implemented on the basis of the method described here. The deposit return system is designed to ensure that objects remain in a defined economic and recovery cycle and are not removed from it. In order to ensure that objects are returned to the economic and recovery cycle even when they are temporarily removed, a deposit is levied on the objects before removal, which is only repaid when they are returned to the cycle. The requirements for the deposit system include: a) No false objects should be placed in the deposit system, b) The deposit system must be as simple as possible for everyone involved, c) The deposit system should be as inexpensive as possible.

This problem is solved using the method described above as follows: a) A marking is applied / applied to the objects, which is tamper-proof (copy-protected) and can be easily validated. b) The identification consists of a random pattern (eg a random distribution of color particles) and a corresponding test code, which is preferably machine-readable (eg as a bar code) (variant C). The marking can be designed, for example, as printing (screen printing, gravure printing, digital printing, wax printing, etc.) or as a label. c) The marking functions as an authentication feature, the validity of which can be checked by a test device by comparing the random pattern and the test code. The random pattern acts as a copy protection, which cannot be cracked or only cracked at an uneconomically high cost. The check code in turn is obtained from the random pattern by extracting features from the random pattern as described above and encoding it as a data record (fingerprint). d) The marking can be applied / applied at any time after the production of the object to be marked. e) The verification of labels is carried out using test devices which have a detector for the random pattern and software which extracts features from the random pattern, calculates a fingerprint and compares this with the test code. f) Objects without or without complete identification or without random patterns and test code matching are invalidated. The devaluation can be carried out in several ways, for example tearing off (in the case of a label), scraping, brief heating (in the case of a wax print), etc.

4) Individual identification of containers, packaging and barcodes. Slight irregularities which arise when printing packaging and containers "on their own" are sufficient (irregularities in the layer thickness, irregularities in the distribution of the printing ink) in order to use the method described in accordance with the invention to otherwise distinguishable packaging and make containers individually identifiable and tamper-proof (copy-proof). A particular advantage here is that the printing process does not have to be changed. Particularly suitable is the identification of the area of containers or packaging which contains printed barcodes or matrix codes, as a result of this Using the method according to the invention, additional individual data can be extracted from the respective bar code (matrix code). Figure 21 shows a container from which fingerprints can be taken from the surface of the container for individual identification using the method described according to the invention. Figure 22 and Figure 23 show the barcode of the container, from which, as described according to the invention, a unique fingerprint was obtained, on the basis of which the individual container can be recognized. The illustrations in Figure 29 and Figure 30 show the identification of the container and the recognition of the fingerprint which is achieved by comparing the original fingerprint with a subsequently recorded fingerprint. The example illustrates how the method described according to the invention can also be used to individually identify unmarked bulk items, which is used, for example, to detect counterfeits or is useful, for example, within a deposit return system. The functionality of the deposit return system is based on the fact that the containers to be pledged are read into the database of the return system. When returning empty containers, the containers are then identified by a reading device according to the method described in accordance with the invention and deleted or blocked in the database so that they cannot be returned several times or illegal copies of containers can be produced and returned. The method according to the invention has the advantage over deposit return systems that are based on the individualization of containers by producing random numbers that are applied to containers as barcodes or matrix codes, that neither random numbers have to be generated, nor that containers with barcode / matrix code have to be printed or labeled , which entails not inconsiderable and costly changes in the manufacturing process of the containers (in the Federal Republic there are about 15 billion disposable packaging annually).

The identification of barcodes in configurations as variant A is particularly advantageous if there are several objects with the same barcode and variant B if there are several objects with different barcodes. In the latter case, the barcode, as described in accordance with the invention, serves as a check code under which the respective fingerprint is stored in the database. Objects of the present invention are thus:

- A method for producing security labels, characterized in that the security label contains a random pattern.

- A method according to the invention, characterized in that the .security mark is formed by the reflecting surface of the object to be marked in each case.

- A method according to the invention characterized in that the safety mark of particles, for. B. colored particles, pigments, effect pigments, sand, dust, crystals (e.g. salt crystals of different colors), ferromagnetic, magnetizable, permanent magnetic, fluorescent, phosphorescent, iridescent, opalescent or radioactive particles.

A method according to the invention, characterized in that the security identifier is formed by fluorescent, phosphorizing, iridescent, opalescent or reflective particles. For example, organic and inorganic, particularly inorganic fluorescent particles such as those offered by the company Leuchtstoffwerk Breitungen GmbH (98597 Breitungen, www.leuchtstoffwerk.com) are suitable. Crystals, platelet-shaped crystals, liquid crystals, reflective pigments from at least two layers with different refractory indexes, effect pigments (interference pigments, pearlescent and metallic luster pigments) are also very suitable. Effect pigments are available, for example, under the trade names IriodinΘ / Afflair® and Colourstream® from Merck KGaA, Darmstadt, under the trade name Helicone® from Wacker Chemie, Burghausen. A particular advantage of reflecting particles is the fact that a) unlike fluorescent particles, they are subject to almost no wear from light, and b) different random patterns arise depending on the viewing and lighting angle. Such changes arise when looking at the effect pigments, for example, in that each individual pigment acts like a microscopic mirror for light of a certain wavelength, which reflects this light in its random spatial orientation, which creates 3-dimensional random patterns. A method according to the invention, characterized in that the particles used have a diameter between 0.1 μm and 1 μm, between 1 μm and 10 μm, between 10 μm and 80 μm, between 80 μm and 150 μm and between 150 μm and 2000 μm, and a thickness of up to 50 μm, preferably to 10μm, particularly preferably up to 3μm.

Use of fluorescent particles, crystals, platelet-shaped crystals, liquid crystals, reflective pigments from at least two layers with different refractory index, effect pigments (interference pigments, pearlescent and metallic luster pigments) for the identification of objects according to the principle of individual random patterns.

A method according to the invention, characterized in that the particles are introduced into a matrix (carrier) and the mixture formed in this way is used for coating objects. Suitable as a matrix are paints and lacquers, preferably water, solvent, powder, UV lacquers, epoxy resins, plastics (e.g. polyethylene), ethyl acetate and comparable materials, paraffins, waxes and wax-like coatings (e.g. Flexane).

Coating obtainable by a process according to the invention.

Coating obtainable by a process according to the invention, characterized in that the coating thickness is <= 20 μm, preferably <= 10 μm, particularly preferably <= 5 μm, very particularly preferably <= 3 μm.

A method according to the invention, characterized in that the particles are introduced into printing inks. Printing inks for screen, gravure, flexo and offset printing are preferred.

Printing ink, obtainable by a process according to the invention.

Printing ink, obtainable by a process according to the invention, characterized in that the coating thickness is <= 20 μm, preferably <= 10 μm, particularly preferably <= 5 μm, very particularly preferably <= 3 μm.

A method according to the invention, characterized in that at least two distinguishable particle types are used for coding additional information; A method according to the invention, characterized in that the marker is applied by spraying, galvanizing or printing, preferably by ink-jet, wax, screen, gravure, letterpress or offset printing;

A method according to the invention, characterized in that the concentration in percent by weight of the particles in the matrix is between 0.01% and 30%, preferably between 0.01% and 1%, between 1% and 10% and between 10% and 30% , is particularly preferably between 0.01% and 1% and between 10% and 15%;

Reading or scanning of individual, chaotic or random features of an object, preferably with the help of optical and electromagnetic detection devices (microscope, interferometer, NMR, AFM etc.).

Reading or scanning of individual, chaotic or random features of an object, preferably with the aid of optical and electromagnetic detection devices (microscope, interferometer, NMR, AFM etc.), characterized in that the random pattern used for identification can be obtained by a method according to the invention.

A method according to the invention, characterized in that at least one radiation source is used to induce a random pattern detectable by the probe and at least one probe is used to detect this random pattern, each of which contains a signal amplifier (e.g. optical or electronic lenses) in front of the actual detector.

A method according to the invention, characterized in that at least one radiation source is used to induce a random pattern detectable by the probe and at least one probe, which preferably contains a signal amplifier (e.g. optical or electronic lenses) in front of the actual detector, is used to detect this random pattern.

A method according to the invention, characterized in that the object or a localized partial area (target area) of the object is recorded in incident light. For this purpose, at least one radiation source and one probe focus in the same target area on the object. The radiation source and probe are preferably located at a defined distance and defined solid angles from the target area of the object. The position of the radiation source (s) and probe (s) in relation to the object is preferred for first scanning of the object is stored and preferably kept secret, so that precise knowledge of the positioning data in the steganographic sense represents a secret key which, if desired, increases the security of the identification.

A method according to the invention, characterized in that no coherent radiation or coherent light (laser) is required for the identification of the random pattern, or the radiation source does not have to emit coherent radiation or coherent light. Suitable sources of radiation are, for example, but not exclusively, diodes, halogen lamps, UV lamps, IR lamps.

A method according to the invention, characterized in that the radiation source and probe work in the spectral range of X-rays, UV, visible light, infrared, microwaves, particularly preferably in the spectral range of UV, visible light or infrared.

A method according to the invention, characterized in that the reading device contains a CMOS or CCD with a digital output and at least 640x480 pixels, preferably 1024x768 pixels resolution.

A method according to the invention, characterized in that the probe contains, as an amplifier in front of the detector, a lens or a lens system for focusing electromagnetic waves. It is preferably an enlarging optical lens system with a transmittance for wavelengths between 100nm and 950nm, particularly preferably a lens system with a transmittance for wavelengths between 300nm and 850nm. The optical magnification of the signal amplifier is at least 2x the original image, preferably between 10x and 200x, 200x and 400x, 400x and 750x and 750x and 1000x, particularly preferably between 25x and 50x, 70x and 100x, 120x and 150x, 170x and 21 Ox, 220x and 250x, 280x and 325x, 350x and 400x, 500x and 600x, 650x and 750x, 800x and 850x, 900x and 1000x.

A method according to the invention for the purpose of identifying or authenticating objects.

A method according to the invention, characterized in that the security identifier is different for each object. A method according to the invention, characterized in that the security identifier is different for each object and is unique in each case (individual identifier).

A method according to the invention, characterized in that packaging and containers are identified individually.

A method according to the invention, characterized in that barcodes and matrix codes are individually identified by random patterns.

A method according to the invention, characterized in that there is a second identifier on the object to be identified, which is preferably machine-readable (e.g. as a barcode or matrix code).

A method according to the invention, characterized in that there is a second identifier on the object to be identified, which is preferably machine-readable (for example as a barcode or matrix code), characterized in that the fingerprint extracted by this object is stored in a database and all accesses run on this extracted fingerprint via the corresponding second identifier.

A method according to the invention, characterized in that there is a second identifier on the object to be identified, which is preferably machine-readable (for example as a barcode or matrix code), characterized in that the second identifier is obtained by extracting a fingerprint from this object is encrypted, is preferably applied using asymmetrical encryption methods and is again applied to the same object as a second identifier, preferably in machine-readable form (for example as a barcode or matrix code).

Use of method variant D for the additional storage of (precoded) data on an object, e.g. of manufacturer and product data.

A method according to the invention, characterized in that the security identifier is machine-readable.

A method according to the invention for the extraction of features from the random pattern and translation into a data structure (fingerprint). A method according to the invention for storing the fingerprint in a database.

Saving the fingerprint as a certificate / verification code on the same object from which it was extracted, preferably encrypted, particularly preferably encrypted using asymmetrical encryption methods.

A method according to the invention, characterized in that the fingerprint can be read electronically.

A method according to the invention, characterized in that the fingerprint cannot be copied or reproduced using optical methods.

A method according to the invention, characterized in that the fingerprint cannot be copied or imitated efficiently, i.e. that counterfeiting the labeled product is too expensive for counterfeiting to be economically worthwhile.

A method according to the invention, characterized in that the label has an identifier in addition to the security identifier (e.g. serial number), machine-readable identifiers (e.g. barcodes, labels, transponders) are preferred.

The detection of labels according to the invention in a database in which fingerprints and identifiers which are associated with one another in each case, or pairs of fingerprints and identifiers which are associated with one another, are stored in a database for subsequent identification.

A method according to the invention, characterized in that probes are used to identify electrical and magnetic fields in order to identify the security indicators.

A method according to the invention, characterized in that the reading device recognizes electrical or magnetic fields or field lines.

A method according to the invention, characterized in that, for identification of the safety indicators, probes and devices from material testing, particularly preferably Hall detectors, magneto-resistive detectors, detectors using the eddy current method, capacitive, inductive detectors, X-ray fluorescence devices, Interferometer and laser interferometer, electron microscope and

Scanning electron microscopes (AFMs, STMs) are used.

A method according to the invention, characterized in that high-resolution detection devices are used, i.e. Devices that can read structures in the μm and nm range.

Use of methods according to the invention as copy protection.

Use of methods according to the invention for product security (counterfeit protection).

Use of methods according to the invention as theft protection.

Use of the method according to the invention as counterfeit protection for means of payment.

Use of the method according to the invention for marking painted objects.

Manufacture of the security label by a method according to the invention by printing (color printing, screen printing, gravure printing, digital printing, flexographic printing, etc.).

Design of security signs according to a method according to the invention as labels.

Production of security identifier, characterized in that the security identifier contains a random pattern obtainable by a method according to the invention as copy protection and the fingerprint extracted from the random pattern as a certificate.

Design of security labels according to a method according to the invention as a label.

Device (generator) for producing security signs comprising device for producing random patterns.

Device (encoder) for producing security labels, comprising o detection probe o computer o Software for recording random patterns, extracting features, encoding features as a fingerprint and, if necessary, encrypting fingerprints o Device for applying the (encrypted) fingerprints to objects.

Device (certification device) comprehensive, o detection probe o computer o software for recording random patterns, extracting features and encoding features as a fingerprint o software for decrypting fingerprints if necessary o software for comparing fingerprints.

Particle mixtures obtainable by a method according to the invention for the production of random patterns.

Use of the method according to the invention within a deposit redemption system.

List of pictures:

Figure 1: Random pattern

Figure 2: Image after threshold operation

Figure 3: Image after erosion

Figure 4: Image after dilation

Figure 5: Resulting dot pattern. The point marked with a marks the

Origin of the coordinate system.

Figure 6: Example of data obtained from a feature extraction from a random pattern. This data table is the actual fingerprint. The data table can now be encoded into a data representation by mapping (e.g. as Arrav or String) which is particularly well suited for storage (e.g. in databases).

Figure 7: Data pipeline, part 1. Excitation of a target area S _t on the object S with a radiation source Q. The excitation of S _t is carried out by Q with a defined one

Wavelength spectrum; Q = source of electromagnetic radiation, Q is in a precisely defined position with respect to S _t ; S physical system; S _t = target area; E = amplifier, E is in a precisely defined position with respect to S _t ; D = detector, D is in a precisely defined position with respect to S _t or E. The emission induced by Q (preferably reflection or reflection, but also transmission, refraction or diffraction) of radiation from S _t is amplified by E and D is detected (digitized).

Figure 8: Data pipeline, part 2. E = amplifier, D = detector, D transforms the radiation sent by E to digital data packets; I = amount of digitized data

(Image data); U = unique identification number. 1 = irradiation of S by Q; 2 =

Emission of radiation from S induced by 1; 3 = The radiation amplified by E falls on D; 4 = the data digitized by D are passed on as a quantity of data in a defined data format; 5 = transformation from I to a unique one

Identifier U as described according to the invention.

Figure 9: The particles described according to the invention form 3-dimensional

Random pattern. When rotating the target object (Figure 9) by approx. 60 ° in the

The random pattern changes completely. The registration marks in the picture (white

Crosses) show the corresponding points of the same object before (Figure 9) and after (Figure 10) the rotation.

Figure 10: (see Figure 9) Figure 11: Reading device consisting of video microscope and computer with software. Figure 12: Reading device consisting of video microscope and computer with software. Figure 13: Random sample obtained by printing Helicone® HC XL effect pigments from Wacker Chemie (Jade XL / Scarabeus on RAL 9005 jet black). The XL pigments have a particle size of approx. 500 μm, which is why the random patterns can be seen with the naked eye.

Figure 14: Random sample obtained by printing Helicone® HC effect pigments from Wacker Chemie (Maple on RAL 9005 jet black). The apparently uniform distribution actually shows distinctive random patterns that become visible when magnified with a microscope (see also Figure 1). Figure 15: Security label according to process variant B. a: Effect field with random pattern (Merck Colourstream®, Merck KGaA, Darmstadt), b: Barcode; According to method variant B, a fingerprint of the effect field is stored in a database under a unique identification code (b) as described, so that the fingerprint can be checked in constant runtime (O (1)) by database access to the identification code.

Figure 16: Design of a security label according to process variant C. a contains a random pattern that is encrypted and saved as a check code b, a and b can subsequently be checked for correspondence without database access. Figure 17: The individual fingerprint is taken from the effects field (a) and used for authentication according to one of the process variants. In the case of method variant B, the fingerprint is stored together with the serial number (b) in a database. The example shows a 50 euro note. After reading the fingerprint and serial number into a database, a single bill is subsequently tested for identity by checking the correspondence between features a and b.

Figure 18: Fingerprinting vs. Chip: a) Conventional authentication strategy using a magnetic stripe / chip, (b) “fingerprint card. In the case of b), the card is authenticated according to one of the method variants described according to the invention. A particular advantage of the variant described according to the invention over other fingerprint methods (such as the 3DAS method, see Figure 19 and Figure 20) is that the method described according to the invention a) also works in incident light, b) none coherent light (laser) required) c) works for standard printing inks and printing processes.

Figure 19: 3DAS marking on plastic card. (Image from: R.L. van Renesse, Optical

Document Security, 2nd Edition, Artech House Boston / London).

Figure 20: 3DAS marking: detailed view of Figure 19. You can see the

Polymer fibers with a diameter of 40μm. (Image from: R.L. van Renesse, Optical Document

Security, 2nd Edition, Artech House Boston / London).

Figure 24: Unprocessed image of a barcode.

Figure 25: Image after threshold operation.

Figure 26: Image after erosion and dilation.

Figure 27: Resulting dot pattern.

Figure 28: Resulting data table (fingerprint) of the coordinates and sizes of the points.

Figure 29: Comparison of two dot patterns that were read from the same barcode.

The lines drawn between corresponding points indicate an inaccurate positioning of the probe.

Figure 30: The overlay of the two dot patterns shows the agreement of the

Fingerprints. (Dots of both fingerprints in white).

Claims

claims:

1. A method for producing security labels, characterized in that the security label contains a random pattern.

2. The method according to claim 1, characterized in that the security label is formed by the reflective surface of the object to be labeled.

3. The method according to claim 1, characterized in that the security indicator of particles, for. B. colored particles, pigments, effect pigments, sand, dust, crystals (e.g. salt crystals of different colors), ferromagnetic, magnetizable, permanent magnetic, fluorescent, phosphorescent, iridescent, opalescent or radioactive particles.

4. The method according to claim 3, characterized in that the security identifier is formed by fluorescent, phosphorizing, iridescent, opalescent or reflective particles. For example, organic and inorganic, particularly inorganic fluorescent particles such as those offered by the company Leuchtstoffwerk Breitungen GmbH (98597 Breitungen, www.leuchtstoffwerk.com) are suitable. Crystals, platelet-shaped crystals, liquid crystals, reflective pigments from at least two layers with different refractory indexes, effect pigments (interference pigments, pearlescent and metallic luster pigments) are also very suitable. Effect pigments are offered, for example, under the trade names lriodin® / Afflair® and Colourstream® by Merck KGaA, Darmstadt, under the trade name Helicone® by Wacker Chemie, Burghausen. A particular advantage of reflecting particles is the fact that a) unlike fluorescent particles, they are subject to almost no wear from light, and b) different random patterns arise depending on the viewing and lighting angle. Such changes arise when looking at the effect pigments, for example, because each individual pigment acts like a microscopic mirror for light of a certain wavelength this light reflects in its random spatial orientation, creating 3-dimensional random patterns.

5. The method according to claim 3 or 4, characterized in that the particles used have a diameter between 0.1 microns and 1 microns, between 1 microns and 10 microns, between 10 microns and 80 microns, between 80 microns and 150 microns and between 150 microns and 2000 microns, and a thickness up to 50 μm, preferably up to 10 μm, particularly preferably up to 3 μm.

6. Use of fluorescent particles, crystals, platelet-shaped crystals, liquid crystals, reflective pigments from at least two layers with different refractory index, effect pigments (interference pigments, pearlescent and metallic luster pigments) for the identification of objects according to the principle of individual random patterns.

7. The method according to claim 3 or 4, characterized in that the particles are introduced into a matrix (carrier) and the resulting mixture is used for coating objects. Suitable as a matrix are paints and lacquers, preferably water, solvent, powder, UV lacquers, epoxy resins, plastics (e.g. polyethylene), ethyl acetate and comparable materials, paraffins, waxes and wax-like coatings (e.g. Flexane).

8. Coating obtainable according to claim 7.

9. Coating obtainable according to claim 7, characterized in that the coating thickness is <= 20 μm, preferably <= 10 μm, particularly preferably <= 5 μm, very particularly preferably <= 3 μm.

10. The method according to claim 7, characterized in that the particles are introduced into printing inks. Printing inks for screen, gravure, flexo and offset printing are preferred.

11. Printing ink, obtainable according to claim 10.

12. Printing ink, obtainable according to claim 10, characterized in that the coating thickness is <= 20 μm, preferably <= 10 μm, particularly preferably <= 5 μm, very particularly preferably <= 3 μm.

13. The method according to claim 1, characterized in that at least two distinguishable particle types are used for coding additional information;

14. The method according to claim 1, characterized in that the application of the marker is carried out by spraying, electroplating or printing, preferably by inkjet, wax, screen, gravure, letterpress or offset printing;

15. The method according to claim 7, characterized in that the concentration in percent by weight of the particles in the matrix between 0.01% and 30%, preferably between 0.01% and 1%, between 1% and 10% and between 10% and 30%, particularly preferably between 0.01% and 1% and between 10% and 15%;

16. Reading or scanning of individual, chaotic or random features of an object, preferably with the aid of optical and electromagnetic detection devices (microscope, interferometer, NMR, AFM etc.).

17. The method according to claim 16, characterized in that the random pattern used for identification is available according to one of claims 2 to 15.

18. The method according to claim 16, characterized in that at least one radiation source is used to induce a random pattern detectable by the probe and at least one probe is used to detect this random pattern, each of which contains a signal amplifier (for example optical or electronic lenses) in front of the actual detector ,

19. The method according to claim 17, characterized in that at least one radiation source is used to induce a random pattern detectable by the probe and at least one probe is used to detect this random pattern, which preferably contains a signal amplifier (for example optical or electronic lenses) in front of the actual detector becomes.

20. The method according to claim 16 to 19, characterized in that the object, or a localized partial area (target area) of the object is recorded in incident light. For this purpose, at least one radiation source and one probe focus in the same target area on the object. The radiation source and probe are preferably located at a defined distance and defined solid angles from the target area of the object. The The position of the radiation source (s) and probe (s) relative to the object is preferably stored when the object is first scanned and preferably kept secret, so that precise knowledge of the positioning data in the steganographic sense represents a secret key which, if desired, increases the security of the identification.

21. The method according to any one of claims 16 to 20, characterized in that no coherent radiation or coherent light (laser) is required for the identification of the random pattern, or the radiation source does not have to emit coherent radiation or coherent light. Suitable sources of radiation are, for example, but not exclusively, diodes, halogen lamps, UV lamps, IR lamps.

22. The method according to any one of claims 17 to 21, characterized in that the radiation source and probe work in the spectral range of X-rays, UV, visible light, infrared, microwaves, particularly preferably in the spectral range of UV, visible light or infrared.

23. The method according to any one of claims 16 to 22, characterized in that the reading device contains a CMOS or CCD with a digital output and at least 640x480 pixels, preferably 1024x768 pixels resolution.

24. The method according to any one of claims 16 to 23, characterized in that the probe as an amplifier in front of the detector contains a lens or a lens system for focusing electromagnetic waves. It is preferably an enlarging optical lens system with a transmittance for wavelengths between 100nm and 950nm, particularly preferably a lens system with a transmittance for wavelengths between 300nm and 850nm. The optical magnification of the signal amplifier is at least 2x the original image, preferably between 10x and 200x, 200x and 400x, 400x and 750x and 750x and 1000x, particularly preferably between 25x and 50x, 70x and 100x, 120x and 150x, 170x and 21 Ox, 220x and 250x, 280x and 325x, 350x and 400x, 500x and 600x, 650x and 750x, 800x and 850x, 900x and 1000x.

25. The method according to any one of claims 17 to 24 for the purpose of identifying or authenticating objects.

26. The method according to claim 1, characterized in that the security identifier is different for each object.

27. The method according to any one of the preceding claims, characterized in that the security identifier is different and unique for each object (individual identification).

28. The method according to any one of the preceding claims, characterized in that packaging and containers are identified individually.

29. The method according to any one of the preceding claims, characterized in that barcodes and matrix codes are individually identified by random patterns.

30. The method according to any one of the preceding claims, characterized in that there is a second identifier on the object to be identified, which is preferably machine-readable (e.g. as a barcode or matrix code).

31. The method according to claim 30, characterized in that the fingerprint extracted from this object is stored in a database and all accesses to this extracted fingerprint take place via the corresponding second identifier.

32. The method according to claim 30, characterized in that the second identifier is obtained by extracting a fingerprint from this object, encrypting it, preferably using asymmetric encryption methods and applying it again as the second identifier on the same object, preferably in machine-readable form Form (e.g. as a barcode or matrix code).

33. Use of method variant D for the additional storage of (precoded) data on an object, e.g. of manufacturer and product data.

34. The method according to claim 1, characterized in that the security identifier is machine-readable.

35. Extraction of features from the random pattern and translation into a data structure (fingerprint).

36. Store the fingerprint in a database.

37. Saving the fingerprint as a certificate / verification code on the same object from which it was extracted, preferably encrypted, particularly preferably encrypted using asymmetrical encryption methods.

38. The method according to any one of the preceding claims, characterized in that the fingerprint can be read electronically.

39. Method according to one of the preceding claims, characterized in that the fingerprint cannot be copied or reproduced using optical methods.

40. Method according to one of the preceding claims, characterized in that the fingerprint cannot be copied or imitated efficiently, i.e. that counterfeiting the labeled product is too expensive for counterfeiting to be economically worthwhile.

41. The method according to any one of the preceding claims, characterized in that the label has an identifier in addition to the security identifier (e.g. serial number), machine-readable identifiers (e.g. barcodes, labels, transponders) are preferred.

42. The detection of labels in a database in which, for subsequent identification, respectively associated fingerprints and identifiers, or pairs of mutually associated fingerprints and identifiers, are stored in a database.

43. The method according to any one of the preceding claims, characterized in that probes for detecting electrical and magnetic fields are used to identify the security identifier.

44. The method according to any one of the preceding claims, characterized in that the reading device detects electrical or magnetic fields or field lines.

45. The method according to any one of the preceding claims, characterized in that for identification of the safety indicators probes and devices from material testing, particularly preferably Hall detectors, magnetoresistive detectors, detectors using the eddy current method, capacitive, inductive detectors, X-ray fluorescence devices, interferometers and laser interferometer, Electron microscopes and scanning electron microscopes (AFMs, STMs) are used.

46. Method according to one of the preceding claims, characterized in that high-resolution detection devices are used, i.e. Devices that can read structures in the μm and nm range.

47. Use of a method according to one of the preceding claims as copy protection.

48. Use of methods according to one of the preceding claims for product security (protection against counterfeiting).

49. Use of a method according to one of the preceding claims as protection against theft.

50. Use of the method described above as counterfeit protection for means of payment.

51. Using the procedure described above for marking painted objects.

52. Manufacture of the security label according to one of the preceding claims by printing (color printing, screen printing, gravure printing, digital printing, flexographic printing, etc.).

53. Design of security labels according to one of the preceding claims as labels.

54. Production of security identifier, characterized in that the security identifier contains a random pattern according to claim 1 as copy protection and the fingerprint extracted from the random pattern as a certificate.

55. Design of security indicator according to claim 54 as a label.

56. Device (generator) for producing security labels comprising device for producing random patterns.

57. Device (encoder) for producing security labels, comprising a) detection probe b) computer c) software for recording random patterns, extraction of features, coding of features as a fingerprint and, if appropriate, encryption of fingerprints d) device for applying the (encrypted) fingerprints to objects.

58. Device (certification device) comprising, a) detection probe b) computer c) software for recording random patterns, extraction of features and coding of features as a fingerprint d) software for decrypting fingerprints if necessary e) software for comparing fingerprints.

59. Particle mixtures for the production of random patterns.

60. Use of the described method within a deposit redemption system.