EP3084686A1 - Two-dimensional identification pattern, article including such a pattern, and methods for marking and identifying such a pattern - Google Patents

Abstract

Marking comprising on a surface: a first pattern for coding a first numerical information item on said surface, -a second pattern for coding a second numerical information item on said surface, wherein the first and second pattern are adapted to be read under corresponding distinct first and second reading conditions.

Description

Two-dimensional identification pattern, article including such a pattern, and methods for marking and identifying such a pattern

The invention relates to two-dimensional identification patterns, in particular patterns used for the identification of articles.

The invention also relates to the articles comprising such patterns, as well as the methods for marking and identifying these patterns on such articles.

Currently, articles are identified by a mark or a logo inscribed on a package o f the article or on the article itself. These inscriptions are visible and make it possible for the article to be identified by all the users.

It is also possible to use other visible identifiers which contain encrypted information so that the content of the identifier is not recognized by all the users .

Identifiers which can be cited as examples include one- dimensional identifiers o f the bar-code type or two-dimensional identifiers of the data matrix type which are the most widely used.

Bar codes represent information using a pattern consisting o f bars of different thicknesses and spacings . However, these patterns can easily be decrypted.

Moreover, the identifiers o f "data matrix" type use white and black pixels forming a symbol and which are arranged within a pattern having a rectangular or quadratic form. The pattern o f such a data matrix consists o f two adjacent "L" shaped so lid edges, which as a who le are called "identification pattern", and two edges formed by alternate white and black pixels, called "reference edges". The identification pattern is used to locate and orient the symbo l and the reference edges make it possible to count the number of co lumns and rows in the symbo l. Furthermore, the data matrix also uses a detection area, sometimes called "quiet zone", to detect the identification pattern. This detection area is used to clearly iso late the identification pattern from any other element o f the article with which it could be confused.

However, these identifiers, which are visible to the user, notably through their identification pattern and their detection area, have to be marked in a chosen place on the article in order not to damage its general appearance. Moreover, the identification pattern is a known and constant pattern regardless of the article to be identified. This identification pattern does not contain information and is used purely to detect the symbol.

Furthermore, these data matrices, although provided with error correction mechanisms, are not sufficiently robust because they depend on the substrate on which they are deposited. In practice, depending on the unevennesses of the substrate, the latter may degrade the pattern during its marking on the article. The robustness of such a pattern also depends on the print quality, because a printing defect, for example an ink which spreads or an absence o f printing o f a pixel o f the symbo l, may greatly affect the pattern. In certain cases, a pattern in the form o f a data matrix with dimensions o f 16 pixels by 16 pixels may thus be rendered illegible by the destruction o f more than 6 pixels.

There is proposed, according to the present invention, a two- dimensional marking which enhances the robustness o f the identification of conventional articles .

There is also proposed a marking which can be invisible to the user while being detectable by a suitable detection system.

There is also proposed a marking which is harder to reproduce and read.

Also proposed are a method for marking, a system for marking, a method for reading and a system for reading.

According to one aspect of the invention, there is proposed a marking comprising on a surface :

- a first pattern for coding a first numerical information item on said surface, - a second pattern for coding a second numerical information item on said surface,

wherein the first and second pattern are adapted to be read under corresponding distinct first and second reading conditions .

"Reading conditions" are in particular physical, lighting and optical conditions that are required for the reading of a marking. Reading conditions may be associated with specific reading means that are used to make readable by adjusting said physical, lighting and optical conditions a pattern and to extract the information coded in the pattern.

By "distinct first and second reading conditions", it should be understood that under the first reading conditions, the second pattern is unreadable or even undiscernable or invisible . Similarly, under the second reading conditions, the first pattern is unreadable or even undiscernable or invisible.

By way o f example, the second pattern may be readable to the naked eye and the first pattern may be invisible to the naked eye.

It should therefore be noted that in order to read both patterns, the distinct reading conditions may lead to the use o f distinct reading means or apparatus .

The second pattern is used to make more complex the reproduction and the reading o f the marking comprising the first and second pattern. A user with knowledge of the presence o f only one pattern will only adapt the reading conditions for one pattern and thus will miss the other pattern. The user will then obtain incomplete information, or may be enticed into reproducing only one o f the patterns for counterfeiting purposes .

The first and second patterns may encode the same or different information.

In an embodiment, the first pattern is associated with a first reso lution so that it can be read under the first reading condition and the second pattern is associated with a second resolution so that it can be read under the second reading conditions, the first resolution being distinct from the second resolution. The reso lution o f a pattern should be understood as the smallest dimension o f an element o f said pattern. This resolution may be measured in dots per inch (DPI) .

The first reso lution may be at least five times finer than the second reso lution or at least five times coarser than the second reso lution.

This difference o f reso lution allows the overlapping either partial or complete of the two patterns . Furthermore, this difference of reso lution can make it impossible for a device adapted for one of the reso lution to read the pattern with the other resolution.

Alternatively and in order to increase said difficulty, in a variant, the first reso lution is at least ten times finer than the second reso lution or at least ten times coarser than the second resolution.

For examp le, one of the resolutions may be of the order of 30 to 40 DPI and the other reso lution may be o f the order o f 300 to 600 DPI . Such a difference o f reso lution allows overlapping o f the patterns, either partial or total, without any loss o f information. In these embo diments, the surface area (the total area used by the pattern) of the pattern may be sufficiently distinct high to allow total overlapping.

An advantage o f the use o f two distinct reso lution is that an counterfeiter can be deceived by a pattern visible to the naked eye (the pattern with the lowest resolution) and fail to recognize and/or decode the other pattern having a higher resolution.

In an embodiment, the first pattern is marked using a first ink so that it can be read under the first reading condition and the second pattern is marked using a second ink so that it can be read under the second condition, the first ink being distinct from the second ink.

The first ink may be chosen in the group formed by black or co lored ink and the second ink may be chosen in the group formed by black or colored IR (Infrared) transparent ink. An advantage o f using different inks is the ability to ensure no information is lost due to partial or total overlap, which is independent of the ratio between the reso lutions o f each pattern which can then be the same or distinct. In these embodiments a device is used to distinguish different wavelengths from the first and second ink, and to extract the information encoded by the patterns .

An advantage o f the use o f two distinct inks is that an counterfeiter can be deceived by a pattern visible to the naked eye (using an ink visible to the naked eye) and fail to recognize and/or decode the other pattern marked with an ink that requires specific illumination.

In an embo diment, the first pattern comprises a specific arrangement of a plurality o f symbo ls belonging to an alphabet, each symbo l o f the alphabet comprising at least three states, including at least two valid states and at least one invalid state, each representation of a symbo l in said arrangement corresponding to a valid state of said symbo l.

Thus, an first pattern is provided which is particularly robust because it comprises a number of symbols and offers a high number o f possibilities o f organization for a determined reference pattern. The expression "reference pattern" should be understood to mean an arrangement of symbo ls, and the expression "identification pattern" should be understood to mean one possible representation out of all the representations of the symbols of the reference pattern.

Furthermore, each symbo l o f the reference pattern may have a number of possible representations, which increases the number o f combinations of representations of one and the same reference pattern. In practice, a pattern that has a determined arrangement of symbo ls may have a number of possible representations corresponding to the possible valid states of each symbol.

Thus, the first pattern is rendered illegible to a user who does not already know the alphabet used to create the pattern, that is to say, who does not know the valid and invalid states of the symbo ls . Even if a user could detect the states inscribed on an article, not knowing the alphabet it is very difficult, or almo st impossible, for him or her to determine the symbols from a reconstruction of all the combinations o f valid and invalid states to obtain the alphabet o f symbo ls. The fact that the symbo ls inscribed on the article cannot be determined makes the pattern impossible to find. The invalid state or states of the symbo ls in particular make the reading practically impossible without the knowledge of the alphabet.

In particular, if the first pattern comprises different symbo ls, the number o f combinations increases and makes its recognition even more difficult.

Furthermore, the first pattern does not require the use o f a particular identifier in its subsequent identification. Compared to a pattern in the form o f a data matrix, the first pattern according to this embodiment does not need a specific detection area to identify the pattern.

The second pattern may comprise a specific arrangement o f a plurality o f symbo ls belonging to an alphabet, each symbo l o f the alphabet comprising at least three states, including at least two valid states and at least one invalid state, each representation o f a symbol in said arrangement corresponding to a valid state of said symbol.

Thus, according to this specific embo diment, the first and second patterns are of a similar type. The difficulty to read the pattern without full knowledge o f the states o f symbo ls therefore also applies to the second pattern.

In an alternative embodiment, the second pattern is a one- dimensional barcode or a two-dimensional barcode.

This choice of pattern may mislead a person trying to reproduce the code and will therefore increase the security o f the marking.

The first pattern may be embedded in at least one black portion of the barcode. By way o f examp le, the reso lution o f the first pattern or its ink may be adapted in order to allow the overlapping.

The first pattern may be embedded in at least one white portion of the barcode. The first pattern and second pattern may overlap fully or partially.

According to another aspect of the invention, there is proposed a method for applying a marking on a surface, comprising the fo llowing steps :

- defining first reading conditions,

- defining a first pattern to be applied for coding a first numerical information item,

- defining second reading conditions distinct from the first reading conditions,

- defining a second pattern to be applied for coding a second numerical information item,

- providing a surface,

- applying said first pattern on said surface using a first application process so that the first pattern can be read under the first reading conditions,

- applying said second pattern on said surface using a second application process so that the second pattern can be read under the second reading conditions .

In an embodiment, the first application process applies the first pattern with a first reso lution and the second application process applies the second pattern with a second resolution, the first resolution being distinct from the second resolution.

The first reso lution may be at least five times finer than the second reso lution or at least five times coarser than the second reso lution.

In an embodiment, the first application process uses a first ink and the second application process uses a second ink, the first ink being distinct from the second ink.

The first ink may be chosen in the group formed by black or co lored ink and the second ink is chosen in the group formed by black or colored IR transparent ink. According to another aspect of the invention, there is proposed a process for marking obj ects by inkj et printing, wherein projected ink forms the above-defined patterns .

According to another aspect of the invention, there is proposed a system for applying a marking as defined above on a surface, comprising :

- processing means configured for defining a first and a second pattern,

- application means for applicating on a surface said first pattern and said second pattern adapted to be read under corresponding distinct first and second reading conditions .

The application means may comprise first application means adapted for the first reading conditions and second application means adapted for the first reading conditions, the first and second application means being distinct.

According to another aspect of the invention, there is proposed a method for reading a marking as defined above on a surface, comprising :

- reading said marking using a first reading process adapted for first reading conditions for reading the first pattern,

- reading said marking using a second reading process adapted for second reading conditions for reading the second pattern.

According to yet another aspect of the invention, there is proposed system for reading a marking as defined above on a surface, comprising :

- first reading means adapted for first reading conditions,

- second reading means adapted for second reading conditions . According to one embodiment, each valid state comprises at least two elements having a first possible representation and a second possible representation distinct from the first representation.

According to an advantageous embodiment, each state of the symbo ls o f the alphabet comprises a group of elements. Each o f said groups of elements comprises at least two elements. Each element comprises a possible representation out of a set of representations comprising a first representation and a second representation distinct from the first representation.

Furthermore, each symbo l comprises a first valid state comprising a first group of at least one element represented according to the first representation and at least one element represented according to the second representation, and a second valid state comprising a second group of at least one element represented according to the second representation and at least one element represented according to the first representation.

Thus, an identification pattern is provided for which the symbo ls have two complementary valid states since the first element of a group for one valid state is different from the first element o f the group corresponding to the other valid state. Similarly, the second element of one group is different from the second element of the other group, each group corresponding to a valid state. Valid states are thus provided which have representations arranged differently. When, for example, the same ink surface area is used to represent these complementary valid states on the article, the identification pattern is made uniform on a macroscopic scale, which makes it more difficult to detect. In practice, if each element is produced, for example, by a dot or a spot of ink of determined surface area, the fact that the valid states all have the same number o f dots or spots of ink makes it possible to obtain the same surface area of ink and therefore the same surface density o f ink for all the valid states and therefore also for the identification pattern consisting o f the arrangement of the different symbo ls in one or other of their valid states .

According to another embodiment, each o f said groups o f elements comprises more than two elements. Each symbo l comprises a first valid state comprising a first group of a number o f elements represented according to a first representation and o f a number o f elements represented according to a second representation. Each symbo l also comprises a second valid state comprising a second group of a number o f elements represented according to the second representation and o f a number o f elements represented according to the first representation. The number of elements in each group may be, for example, four, without this number being limiting .

Thus, an identification pattern is provided which comprises complex symbo ls, which increases the difficulty in recognizing the pattern in a subsequent unauthorized identification attempt.

The abovementioned groups o f elements may preferably comprise one and the same number o f elements, which enhances the uniformity o f the identification pattern to make it even more difficult to detect when said elements are produced by a dot or a spot of ink o f determined surface area.

According to a preferred embodiment, at least one valid state of at least one symbol of said identification pattern represents a digital information item. Thus, it is possible to encrypt a digital information item using the valid states of the symbo ls of the identification pattern.

The invention thus provides a very robust encryption means because, firstly, the identification pattern has to be recognized, then its content has to be identified, before the information that it contains can be accessed.

The valid states of each symbo l o f the pattern may respectively represent binary values o f a digital information item. The pattern is therefore particularly suited to encrypting binary information.

According to another aspect of the invention, an article is proposed which comprises at least one identification pattern defined above.

This article may be any type o f medium for the identification pattern, such as a box, a sheet, for example made o f aluminum, a capsule or a cartridge which may, for examp le, contain an edible substance such as co ffee or tea, a bottle stopper, a bank note, a passport, a security document, a label, a card, a trade voucher, a pharmaceutical pill, a watch, any medium having a metal or ceramic surface, a package, a semiconductor, peripheral devices for computers (mice, keyboards, ...), etc.

According to one embodiment, at least one identification pattern is present on the article, for example on a layer of said article. The article may comprise a number o f identification patterns corresponding to one or more representations o f one and the same reference pattern.

By virtue of the repetition o f the identification pattern, the robustness with respect to its subsequent identification is enhanced. This repetition prevents any impairment of the detection of the pattern, since the impairment of one identification pattern does not prevent the detection o f the presence o f at least one pattern out of all the identification patterns marked on the article.

The identification patterns may be all identical or all different in their representation.

The article may also comprise identical identification patterns and different identification patterns in their representation, that is to say, for which the valid states used for the representation of the different symbo ls are not the same for all the symbols of the reference pattern used.

According to another aspect, the invention also proposes a method for marking an article with an identification pattern which may include an information item relating to the article and/or its use . The method comprises the following steps :

- an alphabet is defined which consists of a number o f symbo ls;

- for each symbo l, at least two valid states and at least one invalid state are defined;

- a reference pattern is created which comprises a set of symbo ls o f the abovementioned alphabet;

- at least one representation o f said reference pattern is created by assigning a specific valid state to each symbo l of the reference pattern;

- finally, a marking reproducing said representation o f the reference pattern is applied to the article.

By virtue of the structure of the identification pattern used, the unevennesses o f the medium or the print quality cannot impair it because a number of symbo ls o f the identification patterns may be affected without in any way preventing a subsequent identification o f the pattern.

Each symbo l can be represented by at least two elements having a first possible representation and a second possible representation.

Preferably, each valid state comprises a group of at least one element represented according to the first representation and o f at least one element represented according to the second representation, each valid state o f one symbo l being different from a valid state o f another symbol.

According to another imp lementation, each invalid state of a symbo l is different from the valid states of any one o f the symbo ls o f the alphabet.

The valid states of all the symbo ls may be represented by a group of elements comprising the same number of elements.

In another aspect, the invention also proposes a method for identifying an identification pattern applied to an article and which may include an information item relating to the article and/or its use . The method comprises the fo llowing steps :

- an image is captured of an area o f the article which contains at least a portion o f a representation o f at least one stored reference pattern comprising an arrangement of a number of symbo ls belonging to a stored alphabet, each symbo l comprising at least two valid states and at least one invalid state, each representation o f a symbo l in said arrangement corresponding to a valid state of said symbo l, and the captured image is stored;

- the symbo ls o f the stored alphabet of symbo ls are identified by comparing the captured image with the arrangement o f representations of symbols of the stored reference pattern;

- the number o f identified symbo ls represented in a valid state and arranged according to the arrangement of said stored reference pattern is computed;

- the result of the computation is compared to a probability thresho ld; - and an identification signal is sent if the result reaches said thresho ld.

According to yet another implementation, a first symbo l is defined in a first position o f the arrangement o f the stored reference pattern; the captured image is searched for a representation of at least one valid state of said first symbo l; if this representation is found, it is considered that the selected symbo l is identified, then another symbo l is defined in another position o f the arrangement of the stored reference pattern, and the preceding search step is reiterated for this other symbo l in this other position, and so on until all the symbo ls o f the stored reference pattern are identified or not.

In another imp lementation, the method comprises a step o f decrypting at least one digital information item represented by at least one valid state of at least one symbo l of the identification pattern comprising an identification o f said digital information item based on at least one valid state of at least one identified symbol.

In all cases, the identification method is particularly robust because it requires both the recognition of the states of a plurality o f stored symbo ls and the recognition o f a specific arrangement o f a certain number of these symbo ls in a valid state, this arrangement forming a stored reference pattern.

Other advantages and features o f the invention will become apparent from studying the detailed description of a few exemplary implementations and embodiments, which are no way limiting, and the appended drawings, in which:

- figure 1 schematically illustrates one implementation o f a creation o f an identification pattern according to the invention;

- figure 2 schematically illustrates an implementation o f a creation o f another identification pattern;

- figure 3 schematically illustrates an implementation o f a method for marking the identification pattern of figure 1 on an article;

- figure 4 schematically illustrates an implementation o f a method for identifying an identification pattern applied to an article : Figure 5 schematically illustrates a marking comprising two patterns,

- Figure 6 schematically illustrates another marking comprising two patterns .

In the figures, the same references correspond to the same elements. The elements o f the figures are not necessarily drawn to scale .

Figure 1 schematically represents one implementation o f a creation S I of a two-dimensional reference pattern P I according to the invention which may be used for the creation o f a first pattern or, in a specific embo diment, for the creation o f the first and o f the second pattern.

This pattern P I comprises an arrangement of four symbo ls A, B , C, D, determined from an alphabet AT I .

The expression "two-dimensional reference pattern" should be understood here to mean an arrangement of a number o f two- dimensional symbo ls which are arranged relative to one another in a determined position.

The expression "arrangement of symbols" should be understood to mean a particular arrangement o f the two-dimensional symbo ls in the pattern.

The expression "identification pattern" should be understood to mean one possible representation out of the all the representations o f the symbo ls of the reference pattern.

The term "symbo l" should be understood here to mean a two- dimensional set of elements. An element o f a symbo l may be, for example, a dot, or a set of dots, such as a letter of the alphabet, an unevenness such as an indentation, a hole, a spur. An element may also be, for example, a color or a reflection of electromagnetic waves.

The set of these symbo ls forms what is called here an alphabet.

In this alphabet, the symbols are different from one another even if the symbo ls may comprise elements of one and the same type as described above. According to the invention, each symbol o f the alphabet comprises a number o f states, in particular more than two states, that is to say at least three states.

A state is a possible representation o f a symbo l. For example, if a symbo l is a letter, a possible representation o f a state is a lower case letter, another state or another possible representation is the same letter in upper case, yet another state or different representation is the same letter colored red, etc.

If the symbo l is a word, a representation o f a state may be a word written from left to right, or vice versa. If the symbo l is a set o f dots, a possible representation of a state may be a single dot, a pair o f dots, another representation may be a non-dot, in other words an absence of any dot, a pair of non-dots, etc.

A valid state of a symbo l is an arbitrarily chosen or authorized state out of all the possible states o f this symbo l. In other words a valid state is a single representation out of all the possible representations of the states of the symbol.

An invalid state is, a fortiori, an unauthorized state, that is to say another representation, distinct from the one used for the valid states of the same symbo l. It should be noted that an invalid state o f one symbo l, for a determined alphabet, may be identical to a valid state of another symbol of the determined alphabet.

Figure 1 shows an example o f an alphabet AT I which comprises four symbols A, B , C, D . Each symbol A, B , C , D comprises two valid states Vil , Vi2 and two invalid states Iil , Ii2. Each state o f all the symbo ls comprises two groups of elements G l to G 16. In this implementation, each group G l to G 16 comprises two elements . Each element comprises a first representation E l corresponding to a dot, that is to say a so lid circle and a second representation E2 corresponding to a non-dot, that is to say a blank circle or even an absence of any dot.

As can be seen in figure 1 , the symbo l A o f the alphabet AT I has a first valid state Vil for which the group G l , arranged vertically, is formed by a dot El, in the top part and a non-dot E2 in the bottom part.

The symbol A has a second valid state Vi2 for which the group G2, also arranged vertically, is composed of a non-dot E2 in the top part and a dot El in the bottom part. The two valid states Vil and Vi2 may, for example, represent the respective values 0 and 1 so as to produce a binary coding of an information item.

The same symbol A has two invalid states Iil and Ii2 comprising two groups G3 and G4 of two elements each arranged vertically. In the first invalid state, Iil, the group G3 is composed of two non-dots E2 one above the other whereas, in the second invalid state Ii2, the group G4 is composed of two dots El, situated one above the other but being separated by a gap.

The symbol B has a valid state Vil, for which the group G5, arranged horizontally, is composed of a dot El in the left side portion in figure 1, and a non-dot E2 in the right side portion in figure 1. The symbol B has a second valid state Vi2, for which the group G6, also arranged horizontally, is composed of a non-dot E2 arranged on the left in figure 1 and a dot El arranged on the right in figure 1. The same symbol B has two invalid states Iil and Ii2, comprising two groups of two elements G7 and G8 arranged horizontally. In the first invalid state Iil, the group G7 is composed of two non-dots E2 arranged horizontally side by side, whereas, in the second invalid state Ii2, the group G8 is composed of two dots El arranged side by side, but being separated by a central gap.

The symbol C has a valid state Vil for which a group G9, arranged diagonally in a first direction in figure 1, is formed by a dot El placed top left in figure 1 and a non-dot E2 placed bottom right. The second valid state Vi2 of the symbol C has a group G10, also arranged diagonally in the same direction as the group G9, and composed of a non-dot E2 situated top left in figure 1, and a dot El situated bottom right in figure 1. The same symbol C has two invalid states Iil and Ii2 comprising two groups of two elements Gil and G12, oriented diagonally in the same direction as the groups G9 and G10 of the two valid states Vil and Vi2 of the same symbol C. The groups Gil and G12 are respectively composed of two non-dots E2 and two dots El arranged at the ends of the diagonal and separated by a central gap.

The symbol D has two valid states Vil and Vi2, composed of the groups G13 and G14, arranged diagonally in a direction opposite to the direction of the groups G9 to G12. The group G13 corresponding to the first valid state Vil of the symbol D is composed of a dot El placed top right and a non-dot E2 placed bottom left in figure 1. The group G14 corresponding to the second valid state Vi2 of the symbol D is composed of a non-dot E2 placed top right and a dot El placed bottom left in figure 1.

The two invalid states Iil and Ii2 of the symbol D correspond to the two groups G15 and G16 each comprising two elements oriented diagonally in the same direction as the groups G13 and G14. The group G15 is composed of two non-dots E2 whereas the group G16 is composed of two dots El placed at the ends of the diagonal and separated by a central gap.

In the creation step SI, a reference pattern PI is constructed from the symbols A, B, C, D belonging to the alphabet ATI. In this implementation, the pattern PI comprises 32 symbols, a particular arrangement of which has been determined specifically, for example according to the information that the reference pattern is required to convey. It can be seen in figure 1 that the reference pattern PI is constructed from a vertical, horizontal or diagonal arrangement of the different groups Gl to G16. The 32 symbols A, B, C and D are arranged interleaved with one another as illustrated in figure 1.

As can be seen in particular in figure 1, the interleaving of the different symbols is preferably done so that the different dots and non- dots are, in this example, aligned vertically and horizontally. As an example, it will be noted that, on the first top horizontal row X-X illustrated in figure 1 of the reference pattern PI, there are, in succession, from left to right, one element of the symbol A for which the group of elements is arranged vertically, two elements of the symbo l B arranged horizontally, one element of the symbol D arranged diagonally in the abovementioned second direction, two elements o f the symbo l B arranged horizontally and two elements o f the symbo l B also arranged horizontally. The second horizontal row of the reference pattern P I denoted Y-Y in figure 1 comprises the second element o f the symbo l A for which the first element was on the first row X-X, and, still working from left to right, one element of the symbo l C arranged diagonally in the abovementioned first direction, the second element of the symbo l D arranged diagonally for which the first element formed part of the first row X-X, in succession the elements of the two symbo ls C, two elements o f the symbo l B and one element of the symbol D .

Similarly, the first vertical line denoted Z-Z o f the reference pattern or symbol P I illustrated in figure 1 is composed, working from top to bottom, of two elements of the symbo l A arranged vertically, one element of the symbol B arranged horizontally, one element of the symbo l C arranged diagonally in the first direction, two elements o f the symbo l A arranged vertically, one element of the symbo l D arranged diagonally in the second direction and one element of the symbo l B arranged horizontally.

It will obviously be understood that a different arrangement could be adopted. It is, however, advantageous to appropriately interleave the different symbo ls so as to make the detection more difficult for a person who is unaware of the construction of the reference pattern and the alphabet used.

It will also be noted that, in the example illustrated, each of the symbo ls A, B, C and D has two valid states Vi l and Vi2, in which the different groups of two elements each time comprise a dot and a non- dot. On the other hand, the invalid states Iil and Ii2 o f the four symbo ls A, B, C and D are each time formed by a group of two elements o f identical nature, that is to say two dots or two non-dots. This particular structure further increases the robustness o f the reference pattern that can be created by the creation operation S I . There are then a number of possibilities such as PI', PI", etc. to represent the duly created pattern PI. These representations PI', PI" of the pattern PI are also hereinafter denoted identification patterns. In practice, given the fact that each symbol A, B, C, D can be represented by two distinct valid states Vil, Vi2, and that the pattern PI comprises 32 symbols, there are 232 = 4294 967296 possibilities for representing the determined pattern PI. According to another example, if the pattern comprises 64 symbols, there are then 264 = 18 446744073 709551 616 possibilities for representing the determined pattern PI. Thus, each representation of a pattern such as PI may correspond to an information item, which makes it possible to characterize an article on which one possible representation of the pattern is affixed. For example, it is possible to characterize the manufacturing source of the article, its place of distribution, and so on. All these information items make it possible to uniquely identify an article, and thus to fight against counterfeiting, if said article is, for example, distributed by an unauthorized person. As a variant, the information item is exclusively referenced in the specific arrangement of the reference pattern PI, the different representations PI', PI" etc. being created randomly so as to render the information contained in the reference pattern PI totally undetectable.

It should be noted that PI' can be used in order to form a first pattern according to the invention. PI' may be applied on a surface of an article in order to be read with first reading conditions.

Figure 1 illustrates two possibilities of representation PI', PI" of the pattern PI, obtained by the operation S2. For each possibility PI', PI", the elements El, E2 of the valid states Vil, Vi2 of the symbols A, B, C, D of the pattern PI are represented. In the representation PI', the symbol Al, for example, will be noted, represented by its valid state Vil as illustrated in figure 1, that is to say, with a dot El in the top portion and a non-dot E2 in the bottom portion. Also, on the first horizontal row of the identification pattern PI', the symbol Bl will be noted in its valid state Vil, that is to say, with a dot El on the left of the figure and a non-dot E2 on the right of the figure. Still in the identification pattern PI', the symbol CI will also be noted, in its first valid state Vil, that is to say, with a dot El top left and a non-dot E2 bottom right, these two elements being arranged on a diagonal in the abovementioned first direction. Also in the representation PI', the symbols A2, B2 and C2 will be noted, using the second valid state Vi2 and the symbols Dl, D2 respectively using the valid states Vil and Vi2.

It should be noted that PI" can be used in order to form a second pattern according to the invention. PI' may be applied on a surface in order to be read under a second reading condition.

In the representation PI", it can be seen that the symbol Al indicated previously in the representation PI' is identical whereas the symbol Bl is replaced by B2 represented with its second valid state Vi2, namely with a non-dot E2 on the left and a dot El on the right. In other words, this symbol B in this position of the pattern PI is used here with its representation corresponding to a valid state other than in the representation PI'. Similarly, the symbol C mentioned previously, and denoted CI in figure 1 for the representation PI', is used, for the representation PI", with its second valid state Vi2, namely with a non-dot E2 top left and a dot El bottom right and denoted C2.

Observation of the two representations PI' and PI" illustrated in figure 1 shows that a certain number of symbols A, B, C and D are thus used either in a first valid state Vil, or in a second valid state Vi2, such that the two representations PI' and PI" are totally different, although the same reference pattern PI has been used.

Figure 2 schematically represents another implementation of a creation SI of a reference pattern P2.

Figure 2 shows an alphabet AT2 which comprises three symbols A', B', C Each symbol A', B', C comprises two valid states Vil, Vi2 and fourteen invalid states II to 114. As in the preceding example, the two valid states Vil and Vi2 can represent the respective values 0 and 1 with a view to a binary coding of an information item in an identification pattern. Each state of all the symbols comprises a group of elements such as Gl and G2 for the two valid states Vil and Vi2 or Gil, GI2, GI3 ... GI13, GI14 for the invalid states II to 114. Each group of elements Gl to GI14 comprises four elements each having a first representation El corresponding to a dot, that is to say a solid circle and a second representation E2 corresponding to a non-dot, that is to say, a blank circle or even an absence of dot. In the example illustrated in figure 2, the two groups Gl and G2 corresponding to the two valid states Vil, Vi2 each time comprise two representations El, that is to say two solid dots, and two representations E2, that is to say two non-dots. The advantage of such a choice is that an identification pattern P2', P2", obtained as will be described below, is uniform, that is to say that each group Gl, G2 corresponding to the two valid states comprises the same density of dots El. If these elements or dots El are produced by printing an ink, the result thereof is that the surface density of the ink on the pattern remains constant.

As can be seen in figure 2, each group of two elements for the different symbols A', B' and C is approximately in the form of a square at the four corners of which are arranged the dots El and the non-dots E2. Thus, for the symbol A' in the valid state Vil, two dots El are aligned vertically on the left portion of this square relative to figure 1 and two non-dots E2 are aligned vertically on the right portion of the square. For the valid state Vi2 of the same symbol A', two dots El are aligned vertically on the right side of the square and two non-dots E2 are aligned on the left portion of the square.

For the symbol B', two dots El are aligned horizontally in the top portion of the square for the valid state Vil, and, in the bottom portion of the square, for the valid state Vi2. For the symbol C, two dots El are arranged diagonally opposite at two corners of the square on a diagonal oriented in a first direction for the valid state Vil, and, on the other diagonal oriented in the other direction for the valid state Vi2. The non-dots E2 are each time arranged at the corners which are left free by the dots El.

The different invalid states may comprise, for each of their groups Gil to GI14: no dot El, a single dot El, two dots El or even three or four dots El, such as, for example, for the group GI14. Here again, the non-dots E2 are arranged at the positions left free by the dots El .

This alphabet AT2 being defined thus, it is possible to create a reference pattern P2 during a creation step denoted SI in figure 2, by having a certain number of symbols A', B' and C according to a determined arrangement. In the example illustrated in figure 2, the reference pattern P2 is in the form of a square comprising 16 symbols A', B' and C according to a specific arrangement illustrated in figure 2.

During a step denoted S2, each symbol of the reference pattern

P2 is assigned one of the two valid states Vil or Vi2. This assignment is made randomly or in a specifically determined manner according to an information item that the reference pattern is required to represent. Figure 2 shows, as an example, two possibilities of representation denoted P2' and P2". It will obviously be understood that a large number of other possibilities can be envisaged with the two valid states Vil and Vi2. The two representations P2' and P2" differ by the use, for certain symbols of the reference pattern P2, of different valid states. Thus, the symbol A' located in the top left corner of the representation P2 is represented in the form of the valid state Vil corresponding to the group Gl of the alphabet AT2. The same symbol A' in the top left corner of the representation P2" is represented in the form of the valid state Vi2 corresponding to the group G2.

It should be noted that P2' and P2" may be used, in a specific embodiment, to form respectively first and second patterns according to the invention. Each pattern may then be applied so that they can be read under distinct reading conditions.

Figure 3 schematically represents an implementation of a method for marking an article SUP using two representations PI' and PI" of a reference pattern PI as described and illustrated in figure 1 in order to form a marking according to the invention. In this example, PI' is a first pattern and PI" is a second pattern.

This method comprises the step SI of creation of a reference pattern PI comprising an arrangement of 32 symbols A, B, C, D, as described in relation to figure 1 , a step S2 of assignment of valid states to the created pattern P I and a step S3 o f marking the representations P I ' and P I " obtained of the valid states of the symbo ls A, B , C and D on the article SUP .

The term "marking" should be understood to mean a printing, for example o f ink, an etching, that is to say a production o f ho llows and bumps on a medium, for example a metal medium, or any type o f known representation o f a pattern on an article. This marking may be visible, when it is, for examp le, produced in the visible scale, or invisible when it is produced in the form o f signs, for example magnetic, which can only be detected by appropriate measuring instruments .

There are a number of means that make it possible to render the identification pattern undetectable to the naked eye. For example, it is possible to use an ink whose chemical nature renders it invisible to the naked eye but simply visible under specific lighting, for example using ultravio let or infrared lighting.

Advantageously, in order to strengthen the security o f the marking of the representations of the reference pattern, it is possible to use inks whose chemical nature renders it invisible to the naked eye. The composition o f this ink may comprise one or more pigments and/or co lorants which absorb in the visible or invisible range o f the electromagnetic spectrum and/or may also comprise one or more pigments and/or co lorants which are luminescent. Non-limiting examples of appropriate pigments and/or colorants which absorb in the visible or invisible ranges o f the electromagnetic spectrum include the derivatives o f phthalo cyanine. Nonlimiting examp les o f appropriate luminescent pigments and/or co lorants include the derivatives o f lanthanides. The presence o f pigment(s) and/or a colorant(s) makes it possible to enhance and strengthen the security o f the marking against counterfeiting .

It is also possible to mix the identification pattern with a noise, for examp le by incorporating the identification pattern in an environment of the surface of the article SUP, for example preexisting irregularities. These irregularities must be comparable to the size of the pattern.

Moreover, it is possible to use a very small printing scale, from 150 to 500 dpi units (dpi standing for dots per inch, that is to say a number of pixels per inch, one inch being equivalent to 2.54 centimeters) which reflects the precision of the printing. With such a printing scale, the identification pattern cannot be distinguished with the naked eye from the background of the printing medium. Furthermore, the reproduction of the identification pattern would then entail using an even finer scale in order to avoid any reproduction error, which is almost impossible in practice. As illustrated in figure 3, in the step SI of creation of the pattern PI, a particular arrangement of symbols from the symbols A, B, C, D of the alphabet ATI is determined as explained above. Then, in the assignment step S2, a valid state Vil or Vi2 is assigned to each symbol of the pattern PI, so as to obtain different possible representations PI', PI", etc. out of all the possible representations.

For each pattern PI' and PI", a selection of marking technique, resolution, and eventually of ink may be conducted in order to obtain marked patterns that can be read undr different reading conditions.

Then, in the marking step S3, the representation or representations such as PI', PI", etc. of the pattern PI is/are marked on the article SUP by marking the representations El, E2 of the valid states Vil, Vi2 which have been assigned to each symbol of the pattern PI.

It should be noted that PI' and PI" can be marked simultaneously or separately, using the same or separate equipement, depending on the reading conditions to be obtained.

It should also be noted that the patterns PI' and PI" may overlap, either partially or totally.

The identification patterns PI' and PI" makes it possible to code an information item concerning the article SUP and/or its method of use. In the example illustrated in figure 3, two representations PI', PI" of the pattern PI have been marked on the article SUP. In other words, for each representation PI', PI" of the pattern PI, the dot/non- dot representations of each symbol A, B, C, D of the pattern PI have been marked on the article SUP.

Moreover, the creation of an alphabet in which the invalid states of the symbols comprise the same representations as those used for the valid states, but in a different arrangement, makes the recognition of the identification pattern more difficult. If the example of the alphabet AT2 is taken, in practice, the invalid state 110 for the symbol B' corresponds to the valid state Vil of A'. Any one not knowing the reference pattern P2 could therefore make an incorrect reading by believing the symbol A' had been detected. The knowledge of the reference pattern P2 on the other hand makes it possible to correctly locate the symbol A' and eliminate the ambiguity.

If the user does not know the reference pattern beforehand, it is therefore almost impossible for he or she to be able to recognize it from the states represented on the article SUP. In particular, there is one chance in 4 294 967296 of determining the pattern PI, when the latter comprises thirty two symbols as illustrated in figures 1 and 3. In the case where the pattern PI comprises sixty four symbols, there is then one chance in 264 of determining the pattern PI.

The reference pattern PI or P2 can also be used to encrypt a digital information item. The valid states Vil, Vi2 of the symbols of the reference pattern are then used to code this digital information item. This digital information item may be a numerical value of binary type.

For example, it is possible to code the binary value 1100 (or 12 in decimal) with the first four symbols SP1, SP2, SP3, SP4 illustrated in figure 3 of the determined pattern PI. For this, it is possible to establish the following code, for each symbol A, B, C, D of the determined pattern PI: the first valid state Vil corresponds to the value 1 and the second valid state Vi2 corresponds to the value 0. In this case, the first symbol SP1 of the determined pattern PI corresponds to the value 1, the second symbol SP2 of the determined pattern PI corresponds to the value 1, the third symbol SP3 of the determined pattern PI corresponds to the value 0, the fourth symbol SP4 of the determined pattern PI corresponds to the value 0. Thus, the first four symbols represent the decimal value 12. It is also possible to choose other codes for which each first valid state of the symbols of the pattern PI corresponds to 1 or 0, depending on the code chosen. This digital information item may be a date of birth, a product reference, for example a reference to its place of manufacture, its place of distribution, its content, a datum characteristic of the article on which the pattern is represented, etc.

The method of marking from the reference pattern P2 illustrated in figure 2 proceeds in the same way with, in the marking step S3, the use of representation P2" of the reference pattern P2 on an article SUP.

Figure 4 represents an implementation of a method of identifying an identification marking comprising patterns PI' and PI", affixed or marked on the article SUP.

An alphabet and a reference pattern, in other words a particular arrangement of a certain number of symbols of the stored alphabet, are stored beforehand.

In the example illustrated, the identification method comprises a step Dl in which an image is captured of an area of the article SUP, a step D2 of identifying symbols, and a step D3 of identifying the identification pattern PI', PI". Furthermore, this method may comprise a step D4 of decrypting a digital information item.

In the step Dl, at least an image ID1 is captured of an area of the article SUP with, for example, a decoding camera, the image being stored in a memory.

The wording "camera" should be understood as a single camera or a plurality of cameras each adapted for specific reading conditions.

Then, in the step D2, the symbols contained in this stored image ID1 are identified from the stored alphabet which was used when marking the representations PI' and PI" on the article SUP. The symbo l identification step D2 is applied to a captured image which comprises one or more identification patterns .

The arrangement o f the symbo ls o f the reference pattern, and therefore the position o f each symbo l in the reference pattern, are known beforehand.

The stored reference pattern is therefore superposed on an area of the captured image. Then, a symbo l of the reference pattern is selected, for example a first symbol positioned top left in the reference pattern, and the representation o f a symbol o f the alphabet contained in the area o f the captured image lo cated in the same position as that of the selected symbo l is determined. If the determined representation corresponds to at least one representation o f a valid state of the selected first symbo l o f the reference pattern, it is considered that the symbo l selected in the reference pattern has been identified in the area of the captured image. Otherwise, it is considered that the selected symbo l has not been identified.

Then, another symbo l o f the reference pattern is selected, preferably a symbo l immediately adjacent to the preceding one, and the comparison o f the symbo l representation o f the area o f the image situated in the position o f the second selected symbo l with the representations o f the valid states thereof is reiterated. If the symbo l representation o f the image corresponds to at least one representation of a valid state of the second symbo l, the second symbo l is considered to be identified. Then, the preceding step is reiterated for each symbo l of the reference pattern by comparing a representation o f the image situated in the same position as the selected symbo l o f the reference pattern with the representations o f the valid states thereof until all the symbo ls o f the reference pattern are identified or not.

This step D2 may be carried on for both patterns according to the two reading conditions .

If at least one symbo l is identified, the fo llowing step D3 o f identifying the identification pattern is performed; otherwise, it is possible, for example, to move the camera, or the article, and reiterate the method from a new captured image, or it is possible to move within the captured image by a step equal to the height or to the width of a symbo l o f the stored alphabet and the method is reiterated in another area of the captured image. The physical movement, in order to capture a new image, can be done by moving the camera, for example, step by step, the width o f a step substantially corresponding to the width or a height of a representation o f an element of a symbo l, or by moving the article while keeping the camera fixed. The movement in the captured image can be obtained by so ftware means according to a step determined in terms of number o f pixels according to the resolution of the image.

In a practical embodiment, it is possible to provide a fairly large quantity o f identification patterns arranged on the article, side- by-side and one above the other, for example 20 □ 20 identification patterns . The viewing camera may then be fixed and encompass a large number o f identification patterns to perform the abovementioned identification.

When the step D3 is performed, the number o f identified symbo ls is calculated. If all the symbo ls are identified, it is considered that the identification pattern has been identified. The expression "identified symbo l" should be understood here to mean a symbo l detected in the image ID 1 which corresponds to a symbol of the stored reference pattern, and which is also arranged on the article according to the arrangement of the symbo ls o f the reference pattern.

If, however, the symbo ls detected in the image ID 1 are not all identical to those o f the stored reference pattern, the number o f identified symbo ls is then calculated, and if this number is greater than a probability thresho ld, it is considered that the identification pattern has been identified.

This step D3 may be carried on for both patterns according to the two reading conditions .

For example, according to a preferred embo diment, the two- dimensional reference pattern has a dimension 16 □ 16, that is to say that it comprises 256 states, or 128 pairs of states . In this case, it is considered that a recognition o f 30 pairs of states, correctly arranged relative to the reference pattern, is sufficient to consider that the identification pattern has been recognized. In other words, if an alphabet is considered which comprises symbo ls for which the valid states are represented by a pair o f states, and that a reference pattern of dimension 16 □ 16 comprises 128 symbo ls, it is sufficient to identify 30 symbo ls out of the 128 symbo ls o f the identification pattern to consider that the identification pattern has been identified. In this case, the probability thresho ld is equal to approximately 24% .

According to another example, if an alphabet is considered which comprises symbo ls for which the valid states are represented by four states, and that a reference pattern o f dimension 16 □ 1 6 comprises 64 symbo ls, it is sufficient to identify 10 symbols out of the 64 symbo ls o f the identification pattern to consider that the identification pattern has been identified. In this case, the probability thresho ld is equal to approximately 16% .

A partial alteration o f a certain number of symbo ls o f an identification pattern can therefore be tolerated. This represents an affirmation o f the robustness o f the identification o f the identification pattern.

If the percentage o f symbo ls identified is sufficient, it is then considered that the identification pattern contained in the image corresponds to a representation of the reference pattern which has been stored. This probability thresho ld makes it possible to tolerate the errors in reading the representations contained in the image, or the errors of representation o f the identification pattern when marking the latter on the article.

It will be noted that, in the marking o f the valid states o f the reference pattern on the article, invalid states seem to appear because of the alphabet used in which the symbo ls comprise at least one invalid state. This makes it all the more difficult for an unauthorized user to determine the valid symbo ls, and thereby renders the identification pattern almost undetectable.

A margin o f movement heightwise and widthwise is tolerated for each dot of each symbo l. Slight differences o f diameter in the dots or spots which define the representations of the symbo ls do not affect the detection. Nor do slight differences of position affect the detection since the reading o f the identification pattern is done by comparing a position of one symbol relative to another.

In the case where it is considered that the identification pattern has not been recognized, that is to say if the identification conditions defined above are not valid, the fact that the article is not identified is signaled (arrow denoted "no" in figure 4) . If, however, the article is identified (arrow denoted "yes" in figure 4), the information contained in the identification pattern can be taken into account, for example to trace the manufacture of the article or to use the article in a certain way (authorized washing temperature, steam pressing recommended, etc.) .

According to another imp lementation, the identification pattern may represent a digital information item and the identification metho d comprises a step D4 in which the value of the digital information item is identified on the basis o f the valid states o f the symbo ls o f the recognized identification pattern and the code used in encrypting said digital information item.

The identification method may also comprise conventional error correction steps in order to determine the expected valid states of the symbo ls which have not been able to be identified in the step D2 of determining the states of the identified symbols.

Thus, using error corrections, it is possible to consider that the identified symbo ls correspond to those of the determined reference pattern with a better reliability.

Also proposed is a system of two-dimensional representation o f a reference pattern configured to implement the method described in figure 3.

In one embo diment, the representation system comprises at least one memory configured to store an alphabet such as AT I or AT2 comprising a number o f symbo ls, creation means for creating a reference pattern P I or P2 from this alphabet and representation means configured to represent at least one possibility o f representation of the created reference pattern P I . These representation means are also configured to assign, randomly, to each symbo l o f said created reference pattern, a valid state out of the valid states o f said symbo l, and to store the representations o f the reference pattern P I deriving from the assignment in a memory.

Furthermore, the system comprises marking means for marking the representations of the reference pattern P I on a medium SUP . These marking means are, for example, inkj et or laser printing means, engraving means based on techniques such as photo-engraving, photogravure, laser etching, engraving using a router, or even intaglio etching on metal (engraving using a graver or an acid) according to the surface used. Furthermore, the marking means may be printing means based on flexography, screen printing, stamp printing (marking metho d based on the transfer o f ink using a flexible stamp), lithography, or xerography. For example, the marking means are able to mark the representations o f the reference pattern using an ink visible to the naked eye or invisible to the naked eye.

These marking means may even be thermal printing means or embossing means which can be used, for example, on polymeric or metal surfaces .

Each marking technique is used according to the type of article on which the identification pattern or patterns is/are marked.

Also proposed is a system for identifying the identification pattern obtained according to the two-dimensional representation method described in figure 3 , configured to imp lement the identification method described in figure 4.

In one embodiment, the identification system comprises a detector or a plurality o f detectors, for example a digital cameras, able to detect an area of the article comprising representations of symbo ls of both patterns. The identification system also comprises at least one memory configured to store an alphabet and a reference pattern defined from said alphabet, determination means for determining the symbo ls that have a valid state out of the symbols of the detected area, and comparison means for comparing all the symbo ls that have a valid state and their arrangement in the detected area with the defined reference pattern stored in memory.

Furthermore, the robustness o f the identification o f the identification pattern can be enhanced by having a number o f representations (identical or not) of the reference pattern on the article. The need to have to place the article according to an exact position when the detector is fixed is thus avoided.

The means mentioned above may be produced as so ftware in a microprocessor and/or by logic circuits .

The identification pattern thus defined allows for a robust identification o f articles. Such an identification pattern is also illegible to the unauthorized user and does not detract from the article on which it is inscribed. This identification pattern makes it possible to safely identify an article for better traceability o f the article. In a commercial context where there are numerous counterfeitings, the proposed identification pattern makes it possible to convey information so as to fight against any fraudulent use of articles .

Furthermore, the use o f an additional pattern associated with different reading conditions makes more complex the reading or the reproduction of the pattern by a counterfeiter.

On figure 5 , a marking according to an aspect o f the invention has been represented. The marking is marked on a surface 1 00.

The marking comprises a first pattern 10 1 , for example a first pattern as described by reference to figures 1 to 4. This first pattern comprises elements schematically represented as ellipsis 103.

The marking also comprises a second pattern 102 comprising elements schematically represented by squares 1 04.

In the examp le o f figure 5 , the first pattern has a first reso lution and the second pattern has a second resolution. The reso lution o f the first pattern is the smallest dimension 105 o f an element 103. The resolution o f the second pattern is the smallest dimension 106 of an element 104. The first resolution 105 is chosen in order to be 5 times smaller than the second resolution 106, so that the first pattern 101 may be readable even if it is superposed on the second pattern 1 02.

On figure 6, a marking according to another aspect o f the invention has been represented. The marking is marked on a surface 100.

The marking comprises a first pattern 1 1 1 and a second pattern 1 12, respectively comprising elements 1 13 and 1 14.

In this example, different inks are used so that the two patterns may be read under distinct reading conditions. A black or colored ink is used for printing the first pattern and a second ink is chosen in the group formed by black or colored IR transparent ink is used for printing the second pattern. Typical black or colored ink printed by inkj et or laser techniques are already known by the one skilled in the art and so ld under the market by know providers such as Hewlett Packard, Olivetti i-j et or Canon. Typical black or colored IR transparent ink are also known by the one skilled in the art and described in EP 1789505B 1 or WO2007133533 A l .

In this example, the resolution o f each pattern may be the same or different.

Claims

1 . Marking comprising on a surface :

- a first pattern for coding a first numerical information item on said surface,

- a second pattern for coding a second numerical information item on said surface,

wherein the first and second pattern are adapted to be read under corresponding distinct first and second reading conditions .

2. The marking as claimed in claim 1 , wherein the first pattern is associated with a first reso lution so that it can be read under the first reading condition and the second pattern is associated with a second resolution so that it can be read under the second reading conditions, the first resolution being distinct from the second reso lution.

3. The marking as claimed in claim 2, wherein the first reso lution is at least five times finer than the second resolution or at least five times coarser than the second resolution.

4. The marking as claimed in any one o f the preceding claims, wherein the first pattern is marked using a first ink so that it can be read under the first reading condition and the second pattern is marked using a second ink so that it can be read under the second condition, the first ink being distinct from the second ink.

5. The marking as claimed in claim 4, wherein the first ink is chosen in the group formed by black or colored ink and the second ink is chosen in the group formed by black or colored IR transparent ink.

6. The marking as claimed in any one of claims 1 to 5 , wherein the first pattern comprises a specific arrangement o f a plurality o f symbo ls belonging to an alphabet, each symbo l o f the alphabet comprising at least three states, including at least two valid states and at least one invalid state, each representation o f a symbo l in said arrangement corresponding to a valid state of said symbol.

7. The marking as claimed in any one of claims 1 to 6 , wherein the second pattern comprises a specific arrangement of a plurality o f symbo ls belonging to an alphabet, each symbo l o f the alphabet comprising at least three states, including at least two valid states and at least one invalid state, each representation o f a symbo l in said arrangement corresponding to a valid state of said symbol.

8. The marking as claimed in any one of claims 1 to 6 , wherein the second pattern is a one-dimensional barcode or a two-dimensional barcode.

9. The marking as claimed in claim 8 , wherein the first pattern is embedded in at least one black portion of the barcode.

10. The marking as claimed in claim 8 , wherein the first pattern is embedded in at least one white portion of the barcode.

1 1 . The marking as claimed in any one o f the preceeding claims, wherein the first pattern and second pattern overlap fully or partially.

12. A method for applying a marking on a surface, comprising the fo llowing steps :

- defining first reading conditions,

- defining a first pattern to be applied for coding a first numerical information item,

- defining second reading conditions distinct from the first reading conditions,

- defining a second pattern to be applied for coding a second numerical information item,

- providing a surface,

- applying said first pattern on said surface using a first application process so that the first pattern can be read under the first reading conditions,

- applying said second pattern on said surface using a second application process so that the second pattern can be read under the second reading conditions .

13. The method according to claim 12, wherein the first application process applies the first pattern with a first resolution and the second application process applies the second pattern with a second resolution, the first reso lution being distinct from the second reso lution.

14. The method according to claim 1 3 , wherein the first reso lution is at least five times finer than the second resolution or at least five times coarser than the second resolution.

15. The method as claimed in any one of claims 12 to 14, wherein the first application process uses a first ink and the second application process uses a second ink, the first ink being distinct from the second ink.

16. The method as claimed in claim 1 5 ,

wherein the first ink is chosen in the group formed by black or co lored ink and the second ink is chosen in the group formed by black or colored IR transparent ink.

17. A process for marking obj ects by inkj et printing, wherein proj ected ink forms the patterns according to claim 12.

1 8. A system for applying a marking according to any one o f claims 1 to 1 1 on a surface, comprising :

- processing means configured for defining a first and a second pattern,

- application means for applicating on a surface said first pattern and said second pattern adapted to be read under corresponding distinct first and second reading conditions .

19. The system according to claim 1 8 , wherein the application means comprises first application means adapted for the first reading conditions and second application means adapted for the first reading conditions, the first and second application means being distinct.

20. Method for reading a marking according to any one o f claims 1 to 1 1 on a surface, comprising :

- reading said marking using a first reading process adapted for first reading conditions for reading the first pattern,

- reading said marking using a second reading process adapted for second reading conditions for reading the second pattern.

21 . System for reading a marking according to any one 1 to 1 1 on a surface, comprising :

- first reading means adapted for first reading conditions,

- second reading means adapted for second reading conditions