DE10345669B4 - Copies with copy protection and method for generating a security code - Google Patents

Abstract

Data carrier whose carrier material is provided with a locally random structural component and an optically readable data strip of high data density mounted thereon for storing coded data, characterized in that
a) in at least one region (4a, 4b and 4c) defined with regard to the data strip, ie in the data strip itself, in its near-surface region or in its carrier material, in addition to the coded data, a security code different from data carrier to data carrier (8a, 8b and 8c) is stored,
b) this security code is uniquely related to the influence which arises when reading the data strip due to the structural component and
c) a memory for the security code is realized with such materials that it can also be subsequently superimposed on the original data strip without disturbing the reading of the original data strip.

Description

The The invention relates to a data carrier with copy protection and a Method for generating a security code according to the generic terms of the claims 1 and 5.

In the DE 199 26 194 C2 a data strip and a method for encoding and decoding is shown and described, with which it is possible to store data by printing a data carrier digitally with a high data density and then read out again. Such a data memory can be used to who, for example, to store compressed audio data on paper and then read with a simple handheld device and play acoustically. The associated method relies on printing various two-dimensional patterns on the pad, each pattern corresponding to an information bit or bit string. Two-dimensional pattern recognition is used in the decoding to reconstruct the bit sequence. In practice, it was possible to store 30 kbytes of data or up to 20 seconds of CD-quality music on an area of 18 × 55 mm.

A storage of data is then in the U.S. Patent 5,315,089 shown and described. The orientation of non-rotationally symmetrical halftone dots is systematically rotated for data storage. In the U.S. Patent 5,706,099 The technique is improved so that the perceived by the viewer gray tone pressure is optimized. For this purpose, the grid points are replaced by grid cells, in whose diagonally adjacent corners there are two 90 ° circular arcs. The decoding of such coded raster images is technically complicated and, depending on the image information, is subject to considerable decoding errors.

In the post-published DE 10217156 A1 A technique is presented and described that allows optimized gray-tone visuals while providing easy, error-free decoding. This is achieved by so-called macrocells in which a defined number of raster cells are combined with a defined rotation. The number and the rotational positions are set optimized. In addition, macro raster cells can be used well to compensate for poor imaging characteristics in the decoder. For this purpose, known, no information-bearing macro-grid cells are used.

In the post-published DE 10223019 A1 an imaging sensor is shown and described, which allows a compensation of the transfer function. Such a technique can be used in a reading device for data storage of the aforementioned type.

the Previously cited prior art is common that he no granted adequate protection, if it is - like desirable in some applications - to prevent applies that data strips are copied and duplicated by unauthorized persons. Not for the case that prints data with very fine resolution be - it enough then even higher resolving scanner, to raw data for to make a copy.

In the DE 10155832 A1 An optical data carrier with a microchip as copy protection is then shown and described. In this case, data such as a coding key or a serial number are stored in the microchip. Furthermore, a copy protection can be achieved by holographic elements, as for example in the publications EP 0420261 A2 . WO 93/18419 A1 or WO 95/04948 A1 is shown and described. The DE 10139719 A1 Finally, it contains a multilayer body or a multilayer film for attaching holograms. Again, this prior art has in common that it uses relatively complicated and expensive manufacturing technologies. In some cases, counterfeit security is achieved through the complexity and secrecy of the technology. But if the manufacturing process is known, then a forgery can no longer be ruled out.

In DE 101 31 153 A1 is an easy-to-manufacture security element, in particular described for banknotes, which consists of a magnetic and an electrically conductive material and an optically readable code. The obfuscation of the security element according to the invention is based on the fact that different security features are applied to the carrier material from the different materials, but the materials are optically indistinguishable. Unfortunately, after disclosure of the technique, a forgery is possible in principle.

In EP 0696 779 B1 a medium for falsely secure data recording is described. The medium is composed of a substrate, an analog data storage section for magnetically storing analog data, and a digital data storage section for storing digital data. The fact that the analog data storage section consists of a special, magnetic ink which is non-uniformly distributed in a binder, results in the reading of the analog data memory, a specific random signal that can be stored and used to test the authenticity of the medium.

In EP 0 870 269 B1 is still a method for secure data storage by means of a ma gnetischen carrier, as used for example in credit cards described. The method thereby provides a simple way of ensuring the authenticity of the card by measuring a so-called magnetic fingerprint as a reference point and measuring the relative positions of data on the magnetic medium with respect to that reference point and storing it on the data memory.

The in EP 0696 779 B1 and EP 0 870 269 B1 The common feature of the techniques described is that they rely on magnetic materials and thus verification of the authenticity requires relatively complicated or not widespread readers.

In DE 101 62 537 A1 Finally, a method and a system is described, which allows an authenticity check of documents, in particular banknotes. Before the document is released, an image of the document substrate is taken and the acquired image data is stored in a database in association with the document data. Since the substrate has individual, spatially resolved detectable features, the data record associated with the image data enables an authenticity check of the document [0006]. Preferably, the method is applied to documents in which the structural details are formed by optically distinctive small particles of the substrate or by foreign inclusions in the substrate base material. During the authenticity check, the spatial assignment of the structure details relative to the printed document data can then be used. The subject matter of the invention is furthermore a document with a marking, which represents a check code which depends on the individual substrate features [0032]. In this case, detection of counterfeiting is possible with pleasingly simple means. All it takes is a flatbed scanner or a digital camera and suitable test software, and the otherwise necessary relatively complex access to a database is eliminated. For the safety of the procedure, however, the abandonment of the database has disadvantages. For a reliable function of the method, the structural details of the substrate must be investigated in many fields. The smaller the individual structural details and the more tolerable back-up contamination and signal disturbance caused by the reader, the larger the areas must be selected, but also the larger the amount of data that needs to be processed and stored in the database. Practical examinations show that per document is typically several kilobits of data. The assignment to a relatively small check code means that you greatly reduce the amount of data. This happens in the DE 101 62 537 A1 in that preferably substrates with a pronounced individual structure are used, that one limits oneself to relatively few but distinct individual substrate features and only their arrangement and form enters into the test code. Distinctive individual substrate features, however, are the target of counterfeiting attacks as technology becomes known. In the image recording clearly visible Sustratmerkmale can be faked by a fine print and it is then no longer possible to reliably detect a fake. In this respect, substrates would therefore be desirable which do not have a pronounced individual structural component. With weak individual structural components, the method fails in DE 101 62 537 A1 but also because then it remains unclear how in the image capture these weak structural components are distinguished from disturbances that cause any reader or contamination. In summary, the practical implementation of DE 101 62 537 A1 that either a high false rejection rate of originals or a relatively high counterfeit risk are to be accepted.

Of the Invention is based on the task of designing data strips in such a way that that a copy or fake reliable and is detected by simple means as well as in a procedure for corresponding coding of the data strips. This task will according to the invention by the characterizing features of main and secondary claim solved. Here, the actual efficiency in the quality of itself resulting copying security, by the interaction a data strip of high data density achieved with the random process becomes.

What the application of the coded data and the security code, so it is advantageous to proceed according to claim 2. In a way, by the way results in a modulation component according to claim 4. Sometimes, for. However, for example, in printed data strips, supplementing the Backup codes after the printing process of the actual data as difficult to prove because, namely the disk no longer in the press. In such a case recommends it is, for the security code materials with properties according to claim 3, so that - ever as needed - once the properties later be changed can and on the other hand also the security code the original one Code be put on afterwards can.

Each real data carrier has in its material, in particular the surface, random structures to varying degrees. Some reasons for this are inhomogeneities of the materials used as well as variations in the manufacturing process of the data carrier. This is where the surplus is based tions, to use such materials or their structures and their subsequent treatment in order to win a random process according to the claims 6, 8 and 9. In order not to disturb the data storage and to increase the bit error rate (after the error protection), a procedure according to claim 7 makes sense, the random process is preferably designed to be weaker than the disturbances that occur during the reading process.

the for the Practice important disk material Paper and its preparation for use as deliberately built-in random process, the beneficial ones devote themselves Embodiments according to the claims 10 to 12. The like applies to Holographic data storage according to claim 13.

Highest possible safety finally guarantees the Use of the statement of claim 15. If, for the security code, the structure of the carrier material and additionally the interaction of the carrier material structure used with the pressure leaves a fake almost certainly avoided. Even a photographic arbitrary exact Reproduction, which is also the influence of the aforementioned Zufallsprozesse with recorded, then can be completely unexpected as a fake be unmasked.

The following section serves to further explain the identification feature 5b:
According to claim 14, it is - by itself - well known how the modulation without the targeted random process and without unknown interference would appear on the disk. Unfortunately, however, the separate measurement of the targeted random process is not yet directly possible, as will now be explained.

m(x,y)

= ursprüngliche Modulation

z_n(x,y)

= Einfluß des gezielten Zufallsprozesses (Zufallsmodulators)

*g(x,y)

= Faltung mit der Impulsantwort der Übertragung (z.B. Druck und Lesegerät), wobei der Stern (*) für die Faltung steht

r_n(x,y)

= Rauschen beim Lesevorgang

The modulation signal i _n (x, y) measured in the nth produced data strip can be described by the following relationship: i n (x, y) = [m (x, y) + z n (x, y)] * g (x, y) + r n (x, y) Eq. 1 in which:

m (x, y)

= original modulation

_zn (x, y)

= Influence of the targeted random process (random modulator)

* G (x, y)

= Convolution with the impulse response of the transmission (eg pressure and reading device), where the asterisk (*) stands for the convolution

r _n (x, y)

= Noise during the reading process

The influence of the impulse response g (x, y) can be largely compensated, even if it is spatially or time-varying. Here, for example, one goes from the German patent application Akz. P 10223019.6 assume that the original modulation m (x, y) is composed of a finite number of known smaller modulation fundamental patterns. The sensor of the reader will thus capture several identical basic modulation patterns. As soon as a basic modulation pattern is detected, it can be fed in the correctly normalized, shifted, scaled or possibly rotated form in a reference sensor field, which enables a compensation of the impulse response g (x, y). This simplifies equation Eq. 1 to i ' n (x, y) = m (x, y) + zn (x, y) + rn (x, y) Eq. 2

With known i _'n (x, y) and m (x, y) so that the modulation deviation = dm are determined + r _n (x, y) _n (x, y) z _n (x, y).

An alternative to the above compensation is the estimation of g (x, y). Thus, [m (x, y) * g (x, y)] are known and the size [z _n (x, y) * g (x, y) + r _n (x, y)] can be used for further processing become. For estimation, known methods which minimize the quadratic error can be used.

For estimation or compensation of the impulse response g (x, y), a modulation m (x, y) can be selected, for example, the macro grid cells of the German patent application Akz. P 10217156.4-53 equivalent. In this case, a defined number of areas (with coded data and security code) are combined into a macro area, after the rotation positions of the individual patterns in the areas have been previously optimized, and finally, all patterns are rotated by the same angle for data coding. Such optimized macro raster cells can also only be incorporated locally into another modulation; they need not contain any information and then serve only to estimate the unknown impulse response.

The following section serves to explain the identification feature 5c in more detail:
In the simplest case, the modulation deviation dm _n (x, y) is directly coded and added to the data strip in previously reserved areas as a security code.

In order to detect a forgery or unwanted copy, the reader checks whether the influence of the targeted random process z _n (x, y) is present. A simple possibility for this is provided by the cross-correlation of the modulation deviation dm _n with the actual modulation deviation dm _a as follows: [dm n (x, y), dm a (x, y)] ≈ Corr [z n (x, y) + r n (x, y), [z a (x, y) + r a (x, y)] ≈ Corr [z n (x, y), z a (x, y)] ≈ const if n = a, otherwise 0

Only in the event that the current random modulation z _a correlates with the stored random modulation z _n , the data strip is also real. The noise components r _n (x, y) and r _a (x, y) as well as the random modulation z _n (x, y) and z _a (x, y) are statistically independent and therefore uncorrelated.

By becoming aware of the procedures described above, the tamper resistance of the data strips would no longer be present. Counterfeit security despite disclosure of the method can be achieved by using the random modulation z _n only at defined but not publicly known positions (x, y) of the data strip. Typically, high density data strips contain well over 100,000 different pressure cells. If you choose z. For example, if 100 out of 100,000 pressure cells randomly, there are about 10 ⁵⁰⁰ possibilities. Without knowing the position (x, y) a forgery is practically impossible.

• – die Größe des Datenstreifens ändert sich aufgrund der Temperatur, Alterung oder mechanischen Beanspruchung,
• – unterschiedliche geometrische Verzeichnungen des Lesegerätes und
• – Rotation des Lesegerätes in Bezug zum Datenstreifen.

When examining authenticity, difficulties arise in practice in that it is difficult to find positions (x, y). Some reasons for this are:

• The size of the data strip changes due to temperature, aging or mechanical stress,
• - different geometrical distortions of the reader and
• - Rotation of the reader relative to the data strip.

The The consequence is that these are expensive readers. The data strip according to the invention allows here a surprise simple solution: The positions to be checked (x, y) are relative to that applied to the data strip Modulation selected. The modulation is thus the reference grid for the measurement of the random process.

A further increase in the security against forgery is achieved by determining not the modulation deviation dm _m (x, y), but its effects on the decoding process and stored on the data strip. In this case, probabilities for the code bits are calculated in the decoder and the difference to probability 1 is stored (coding with soft decision). Without knowledge of the decoding algorithm a forgery is not possible.

in the The following are with reference to a drawing embodiments of the invention explained in more detail, wherein the corresponding parts in the individual figures the same Have reference numerals. The latter are different in the drawing from "logical ONE" and "logical zero" with a perimeter Mistake. It shows

1 the preparation of code bits for printing on a z. B. paper strips,

2 the influence of structured carrier material on code formation and modulation,

3 the data strip after applying the security code in different areas,

4 the effects of different random processes on a data strip and

5 the security code as an overlay in known matrix codes.

1 now shows in detail how code bits 1 through a modulator 2 be prepared for printing. In addition to the actual data field with the modulation signal 3 Here are three save areas 4a . 4b and 4c provided in which the security code is written. The position of the areas can be kept secret and / or from data strips 9 vary to data strips. Where the security code is located may be encoded in the data strip before the security code is added.

Particularly favorable is the following procedure: The area 4a lies in the actual data area. The code bit to be stored in this area is simply not printed, but is replaced by a bit of the security code. The decoder of the reader becomes the possibly "wrong" code bit of the area 4a recognize as faulty and correct by its error protection. After decoding the code bits, the save areas are known and the security code can be read.

2 shows the influence of the structured carrier material 5 , The boundaries of the individual areas for the code bits and the backup bits are drawn for clarity, although they are not printed. The intended modulation signal is due to the structure of the carrier material 5 and the printing process at random in the area 6 disturbed. The systematic running of the color between adjacent pressure points 7 is not a random process and therefore not suitable for the security code. This effect is compensated by digital processing or the areas are not used for the security code. All systematic influences, in particular those caused by the reader, must be compensated.

3 shows the data strip 9 after the Bits of the security code in the domains 8a . 8b and 8c (previously 4a . 4b and 4c ) are added. The backup bit in the area 8a (previously 4a ) is a logical ZERO and not a logical ONE like the code bit of the corresponding cell. By an error correction (channel coding) the error in the code bits can be corrected.

4 illustrates the effect of different random processes on a data strip. The printed data strip 9a contains the same code bits as the data strip 9 , What remained the same is the bleeding of the ink at neighboring pressure points 7 , On the other hand, the structure of the carrier material has changed 5a and the random nature of the color due to the substrate 6a , The influence of the two random processes is recorded and as a security code in the fields 10a . 10b and 10c stored. This backup code is, of course, opposite the code in the fields 8a . 8b and 8c of the 3 changed. Since the fuse code uses the structure of the carrier material and, in addition, the interaction of the carrier material structure with the pressure, counterfeiting is virtually eliminated. Even a photographically arbitrarily accurate reproduction, which also accurately detects the influence of the aforementioned random processes, can be completely unexpectedly recognized as a forgery. This is done by analyzing the data strip both in reflected light and in transmitted light. The transmitted light captures the inner paper structure.

5 Finally, it shows how an existing matrix code is provided with a security code without the additional domains 4c and 4b of the data strip are printed with the security code. The data strip 9b contains the same code bits as data strips 9 in 2 , It should also be printed on the same paper, and therefore contain, for example, the areas 4c and 4b the same paper structure. The protection code is now superimposed on the normal code bits by slightly decreasing the pressure points for a logical ONE and leaving them untouched for a logical ZERO. In the area 11a Thus, a NULL is coded in the upper left corner and an ONE of the security code is coded in the lower right corner. The code bits remain readable by the small changes with the existing readers because such fluctuations also occur during printing. Leaving the security code with high redundancy, it can be decoded despite the said fluctuations. A data strip with such a superimposed security code can not practically be distinguished from a data strip without security code.

Claims (16)

A data carrier whose carrier material is provided with a locally random structural component and an optically readable data strip of high data density mounted thereon for storing coded data, characterized in that a) in at least one area defined with respect to the data strip ( 4a . 4b and 4c ), that is, in the data strip itself, in its near-surface region or in its carrier material, in addition to the encoded data, a different backup code from disk to disk ( 8a . 8b and 8c b) this security code is uniquely related to the influence which arises when reading the data strip due to the structural component and c) a memory for the security code is realized with such materials that it can also be superimposed on the original data strip without disturbing the reading of the original data strip. Data strip according to claim 1, characterized in that within a region ( 4a . 4b and 4c ) overlaps the area for the actual data (original code) with that for the security code. Data strip according to claim 1 or 2, characterized in that the superimposed security code in the form of a transparent coating of paint, ink, foil or film - for example by bleaching - is applied, either its properties (reflection behavior, spectral intensity, polarization) by a subsequent Treatment - eg (laser) exposure of inks - changes or the entire security code ( 8a . 8b or 8c ) is only invisible afterwards. Data strip according to one of the preceding claims, characterized in that the superimposed security code when applying the data on the carrier material ( 5 ) happened accidentally, incidentally incidentally due to the varying from data strip to data strip modulation component. Method for generating a security code for a data-dense optical data strip which is mounted on a carrier material with a locally random structural component, characterized in that a) the data strip ( 9 ) is read with an optoelectronic reader, b) thereby the effects are determined and quantified, which are due to the locally random structural component of the substrate, c) encoded the data thus determined and as a security code in at least one, with respect to the data strip defined area ( 4a . 4b and 4c ), that is, in the data strip itself, in the oberflä near the original area or in its substrate, in addition to the originally coded data. Method according to claim 5, characterized in that that the carrier material and / or the data strip are selectively subjected to a random process. Method according to claims 5 and 6, characterized in that that a weakly laid out, which applied to the disk Modulation slightly changing Random process is used. Method according to Claim 6, characterized that as a disk Paper, metal, plastic or holographic material use takes place, where each one random Structure component - at Paper whose wood content - evaluated becomes. Method according to claim 8, characterized in that that one (pseudo-) random Structure by subsequent Sandblasting or subsequent embossing of the material is achieved. A method according to claim 8, characterized in that a) the data carrier is produced in a manner known per se by printing on paper, b) the - due to the paper structure - resulting variation of the amount of ink within a data field and / or data strips ( 9 ) is exploited to data strip. Method according to claim 10, characterized in that that a disk made of paper with a much finer structure than the size of the pressure points is used. Method according to claim 10 or 11, characterized that the running of the color amount by locally different occurrence of the paper medium before printing and / or by the nature of the backing structure is increased. Method according to claim 8, characterized in that that in the manufacture of holographic data storage by laser exposure exploits the fluctuations of the laser power for the intentional generation of random disturbances becomes. Method according to claim 5, characterized in that on the data strip ( 9 ) either a known modulation is applied or the modulation is determined in case of need by decoding, error correction and re-modulation. Method according to one of the preceding claims, characterized in that the data strip ( 9 ) to detect a forgery or unwanted copy - eg by means of a retroreflector - is analyzed both in reflected light and transmitted light. Method according to one of the preceding claims, characterized in that the data strip ( 9 ) is coated with a protective film.