EP0509916A1 - Feduciary or security document with a printed design and security markings - Google Patents

Abstract

The invention relates to a fiduciary or security document with a printed design and security markings. According to the invention, the document (1) comprises two superposed security markings (100, 200), of which one is a periodic watermark grid and one is a coding by the cutting of the printed image (G) into parallel strips, the effect of the superposition of these two security markings thus being to affect the individual reading of said markings. The invention is used on fiduciary or security documents, for example on banknotes. <IMAGE>

Description

The invention relates to fiduciary or security documents, of the type comprising printed graphics and security signs, said documents possibly in particular being banknotes.

Documents of this type have already existed for a very long time, the security signs of which are produced by using a magnetic wire totally or alternately embedded in the paper of the document, this wire being able to be further coded: these documents are interesting because they are suitable indeed for mechanized use, the corresponding processing machines being equipped to detect the presence of the magnetic wire in the moving document, and possibly recognize the coding of this wire, in order to authenticate said document.

Such a technique actually has limits in the efficiency of document authentication, which means that complicated coding has to be provided.

There are also documents whose safety signs are based on the principle of a variation in fiber density (density or surface), with a specific coding.

A periodic dry stamping has thus been proposed along a strip parallel to an edge of the document, said stamping provided for the manufacture of paper thus making it possible to vary the density along said strip, with crushing of this strip during the printing of the document.

As a variant, a succession of stamping and counter-stamping steps was used for the form fabric, which makes it possible to obtain a succession of dark and light areas for the document, according to a particular watermark of the network type. watermarked, said network being periodic or not.

However, these techniques have limitations, because they are most of the time constraining with regard to the orientation (large edge or small edge parallel to the direction of scrolling), the shaping (double-sided) of the document, and the direction of sound. scrolling (right-left) in a processing machine.

Finally, it has been proposed to code documents with successions of bars, some of which absorb and others reflect infrared radiation, this for the authentication of said documents.

Generally, these different safety signs have been used alone or juxtaposed, with in the latter case the need to use different types of sensors for the successive detection of said signs.

Today it seems necessary to improve these authentication techniques to combat the increasingly sophisticated techniques used to try to falsify documents.

In the particular case of banknotes, the additional problem of mechanized face value discrimination must also be resolved.

Those skilled in the art, however, encounter great difficulties in seeking to combine different safety signs, on the one hand because the analysis techniques quickly become inextricable, and require the use of different sensors which are often bulky and / or hardly compatible with each other, and on the other hand insofar as we arrive at solutions which are most often restrictive with regard to the shaping and orientation of the document.

In addition, the machines used for sorting, counting and / or distribution are developed to be more and more efficient in working time per document.

This undoubtedly explains why specialists have generally restricted themselves to the use of security signs of only one type, depending on the purpose sought (in particular authentication or discrimination of the face value for banknotes).

The object of the invention is to produce a fiduciary or security document, the security signs of which provide both better assistance with authentication and easy discrimination.

The invention also aims to design a document whose security signs are arranged so as to overcome the shaping and orientation of said document, and its spatial arrangement with respect to the direction of scrolling.

The invention also aims to design a document whose security signs are as discreet as possible by examination with the naked eye, so as not to attract attention.

The object of the invention is finally to produce a document capable of being analyzed by analysis means carrying out calculation treatments which are both uncomplicated and very reliable.

It is more particularly a fiduciary or security document presenting a printed graphics and security signs, characterized in that it comprises two superimposed security signs each produced in the form of a network, of which a first security sign which is in the form of a periodic watermarked network, and a second security sign which results from a cutting of the printed graphics into parallel bands arranged and coded according to a binary coding, symmetrically with respect to an axis of symmetry of the document, the wave of said watermarked network extending in a common direction essentially not perpendicular to the direction of said cutting strips of the printed graphics, the superposition of these two security signs thus having the effect of affecting the individual reading of said signs .

Preferably, the watermarked network forming the first safety sign comprises waves with a sinusoidal surface mass profile. This in fact avoids the presence of brutal contrasts at the edges of the wave (a square or rectangular wave would be more brutal and less discreet). In particular, the variations in amplitude of the waves of the watermarked network are made around the mean plane of said document, which makes it possible to overcome the shaping of the document.

Advantageously also, the wave of the watermarked network extends substantially in a common direction making an angle of 45 ° relative to the direction of the cutting strips of the printed graphics. This achieves a possibility of reading the document not only independent of the shaping and orientation of said document, but also whatever the spatial position relative to the direction of travel.

According to another advantageous characteristic, the watermarked network defines a surface whose closed contour is inside the edges of said document, said surface being entirely crossed by strips for cutting the printed graphics.

Preferably then, the watermarked network is organized according to a square, the dimensions of this square being preferably chosen relatively large to preserve a good flatness of the document (the problem is particularly acute if using bundles or stacks comprising a large number of documents ), and to further increase the discretion of this security sign. Advantageously also, the edges of the square are bevelled: this avoids any phenomenon of insert at the edges, resulting from a high contrast, so that the discretion of this safety sign is further improved.

où T est la longueur d'onde du réseau filigrané. L'identité des largeurs des bandes de découpage et la symétrie de leur disposition permet de faciliter l'analyse du document, tout en autorisant plusieurs codages possibles : ceci est particulièrement intéressant dans le cas où le document est un billet de banque, le codage permettant alors une discrimination de la valeur faciale.According to a particular embodiment, the cutting strips of the printed graphics have the same predetermined width which is a function of the angle β between the common direction of propagation of the wave of the watermarked network and the direction of the cutting strips, when said angle is greater than a reference angle corresponding to a bandwidth reaching half the width of said document; in particular, the width e of the cutting strips is given by the formula: $$\text{e =}\frac{\text{T}}{\text{2 x sinβ}}\text{,}$$

where T is the wavelength of the watermarked network. The identity of the widths of the cutting strips and the symmetry of their arrangement makes it possible to facilitate the analysis of the document, while allowing several possible encodings: this is particularly interesting in the case where the document is a banknote, the encoding allowing then discrimination of the face value.

As a variant, the cutting bands of the printed graphics have the same width depending only on the coding sought and not on the angle between the common direction of propagation of the wave of the watermarked network and the direction of the cutting bands, when said angle β is less than a reference angle corresponding to a bandwidth reaching half the width of said document. This will for example be the case with a direction of propagation parallel to the direction of the cutting strips, but the analysis technique will then have to be adapted accordingly.

More preferably, the graphics of said document are printed with a pair of inks of the same shade, one of which reacts and the other not to a predetermined excitation, so as to define the cutting of said graphics into parallel strips. The excitation could be of various types (inks reacting or not with infrared, microwaves, ultraviolet, magnetic pigments, or even with a radioactive source).

In particular then, the graphics are also printed with other inks which do not react to said predetermined excitation, for example which do not absorb infrared, likewise for any graphics also provided on the back of said document.

In the case of a document comprising a graphic printed on its two faces, it is advantageous that the cutting strips of the graphic can be coded on the front, or on the back, or on the front and on the back of said document.

In particular, the graphic design of said document is printed with a couple of inks, one of which reflects infrared and the other not.

Finally, in the case where the document is a bank note, it is then interesting that it is a note whose first and second security signs are used for authentication of the note, and whose second sign security is used to mechanically discriminate the face value of said ticket.

• la figure 1 illustre un document rectangulaire conforme à l'invention, dont les premier et second signes de sécurité ont été représentés en pointillés, ces signes étant superposés ;
• la figure 2 est une vue en plan illustrant le premier signe de sécurité du document précité, qui est réalisé sous la forme d'un réseau périodique filigrané organisé ici selon un carré, tel qu'il peut se voir par transparence, avec une alternance de zones claires et opaques correspondant aux variations de la masse surfacique dans cette zone filigranée ;
• la figure 3 illustre en plan la face en relief d'une matrice permettant d'emboutir la toile formaire lors de la fabrication du document, pour obtenir un réseau périodique filigrané analogue à celui de la figure 2, les ondulations, ici sinusoïdales, de cette face en relief permettant de réaliser les variations désirées de la masse surfacique dans cette zone filigranée, les bords de cette matrice étant en outre ici biseautés pour adoucir les contrastes au niveau des bords de ladite zone ;
• les figures 4 à 8 sont des coupes, respectivement selon IV-IV, V-V, VI-VI, VII-VII et VIII-VIII de la figure 3, permettant de mieux comprendre l'organisation de la face en relief de la matrice, et en particulier de ses bords biseautés, par rapport au plan moyen de ladite face ;
• les figures 4a à 8a sont des courbes illustrant les variations de la masse surfacique de la zone filigranée obtenue avec la matrice précitée, ces courbes correspondant respectivement aux coupes des figures 4 à 8 (les courbes de variations de la masse surfacique dans le papier sont en effet des transformées directes des courbes correspondantes des variations du relief de la face de la matrice d'emboutissage) ;
• les figures 9 et 10 illustrent le document de la figure 1, avec deux codages différents des bandes parallèles du second réseau, tel que ce document apparaît par exemple lorsqu'il est examiné sous infrarouge (pour un graphisme imprimé avec un couple d'encres dont l'une réfléchit l'infrarouge et l'autre pas), avec ici huit bandes parallèles respectivement codées 1 011 1101 et 0 110 0110 ;
• la figure 11 est une vue en transparence du réseau périodique filigrané obtenu avec la matrice précédemment illustrée, avec un contour carré à bords biseautés, et avec un déphasage particulier par rapport aux axes du carré (qui sont de préférence confondus avec les deux axes de symétrie du document rectangulaire) ;
• la figure 12 est une vue à plus grande échelle, montrant une zone du document où les deux signes de sécurité sont superposés (il y a ici six bandes parallèles du second réseau, qui traversent la zone filigranée avec le premier réseau périodique), cette vue permettant de comprendre comment les deux réseaux superposés sont agencés pour une imbrication compatible avec une analyse par un organe unique au niveau duquel défile le document ;
• la figure 13 complète la vue précédente en montrant une barrette de capteurs, avec un capteur pour chaque bande parallèle du second réseau, ladite barrette étant disposée perpendiculairement à la direction de défilement du document ;
• la figure 14 illustre une variante selon laquelle la direction (DC) de propagation de l'onde du réseau filigrané n'est pas comme précédemment inclinée à 45° par rapport à la direction de défilement (DD), mais est parallèle à ladite direction de défilement, la barrette de capteurs étant dans ce cas agencée différemment, avec deux rangées décalées de capteurs comme cela est visible sur la figure ;
• les figures 15a à 15d sont des vues partielles illustrant différentes variantes d'agencement des capteurs de la barrette de la figure 13, avec respectivement des ouvertures en fentes inclinées, des ouvertures cruciformes, des capteurs multiples à deux capteurs adjacents, et des capteurs multiples à quatre capteurs disposés en carré ;
• la figure 16 est un schéma d'un dispositif d'analyse associé à la barrette de capteurs de la figure 13, montrant les moyens pouvant être utilisés pour le traitement des signaux provenant des différents capteurs, afin d'une part de vérifier le codage du second réseau et de valider le document analysé lorsque le second réseau est conforme, et d'autre part d'analyser le premier réseau et de valider le document analysé lorsque le premier réseau est conforme.

Other characteristics and advantages of the invention will appear more clearly in the light of the description which follows and of the appended drawings, relating to a particular embodiment, with reference to the figures in which:

• FIG. 1 illustrates a rectangular document in accordance with the invention, the first and second security signs of which have been shown in dotted lines, these signs being superimposed;
• FIG. 2 is a plan view illustrating the first security sign of the aforementioned document, which is produced in the form of a watermarked periodic network organized here according to a square, as it can be seen by transparency, with an alternation of clear and opaque zones corresponding to variations in the surface mass in this watermarked zone;
• FIG. 3 illustrates in plan the relief face of a matrix making it possible to stamp the form cloth during the manufacture of the document, in order to obtain a watermarked periodic network similar to that of FIG. 2, the undulations, here sinusoidal, of this raised face allowing the desired variations of the surface mass to be produced in this watermarked zone, the edges of this matrix being further here bevelled to soften the contrasts at the edges of said zone;
• FIGS. 4 to 8 are sections, respectively along IV-IV, VV, VI-VI, VII-VII and VIII-VIII of FIG. 3, making it possible to better understand the organization of the raised face of the matrix, and in particular of its bevelled edges, relative to the mean plane of said face;
• FIGS. 4a to 8a are curves illustrating the variations in the surface mass of the watermarked zone obtained with the above-mentioned matrix, these curves corresponding respectively to the sections of FIGS. 4 to 8 (the curves of variations in the surface mass in the paper are in effect of direct transforms of the corresponding curves of the variations in the relief of the face of the stamping matrix);
• Figures 9 and 10 illustrate the document of Figure 1, with two different codings of the parallel bands of the second network, as this document appears for example when examined under infrared (for a graphic printed with a couple of inks including one reflects infrared and the other does not), here with eight parallel bands respectively coded 1 011 1101 and 0 110 0110;
• FIG. 11 is a view in transparency of the watermarked periodic network obtained with the matrix previously illustrated, with a square contour with bevelled edges, and with a particular phase shift relative to the axes of the square (which are preferably coincident with the two axes of symmetry of the rectangular document);
• Figure 12 is a view on a larger scale, showing an area of the document where the two security signs are superimposed (here there are six parallel bands of the second network, which cross the watermarked zone with the first periodic network), this view allowing to understand how the two superimposed networks are arranged for a nesting compatible with an analysis by a single organ at the level of which the document scrolls;
• FIG. 13 completes the previous view by showing a sensor strip, with a sensor for each parallel strip of the second network, said strip being arranged perpendicular to the direction of travel of the document;
• FIG. 14 illustrates a variant according to which the direction (DC) of propagation of the wave of the watermarked network is not as previously inclined at 45 ° relative to the direction of travel (DD), but is parallel to said direction of scrolling, the sensor strip being in this case arranged differently, with two offset rows of sensors as can be seen in the figure;
• FIGS. 15a to 15d are partial views illustrating different variant arrangements of the sensors of the bar of FIG. 13, with respectively apertures in inclined slots, cruciform openings, multiple sensors with two adjacent sensors, and multiple sensors with four sensors arranged in a square;
• FIG. 16 is a diagram of an analysis device associated with the array of sensors of FIG. 13, showing the means that can be used for processing the signals coming from the various sensors, in order on the one hand to verify the coding of the second network and to validate the analyzed document when the second network is in conformity, and on the other hand to analyze the first network and to validate the analyzed document when the first network is in conformity.

FIG. 1 illustrates a document 1, here of rectangular shape, the large edge of which is noted 2 and the small edge of which is noted 3.

This document presents on one side (front or back) a printed graphic G, illustrating here a hang glider. A graphic can naturally also be provided on the other side of document 1.

According to the invention, document 1 includes two superimposed security signs 100, 200, shown here in dotted lines.

The first security sign 100 is in the form of a periodic watermarked network, delimited by a closed contour C which is inside the edges 2, 3 of document 1. This first security sign is therefore visible by transparency, and then presents a succession of bands 101, 102 which are alternately darker and lighter. The appearance of these bands 101, 102 results from variations in the basis weight in this watermarked zone.

The second security sign 200 is also produced in the form of a network, but this second sign results from a cutting of the printed graphics G into parallel strips 201, 202 which are coded.

The strips 201, 202 are first of all arranged symmetrically with respect to an axis of symmetry of the document 1, in this case the axis X′X, which is parallel to the large edge 2 of said document. There is therefore an even number of bands, arranged on either side of the axis X′X. The other axis of the document is noted Y′Y in FIG. 1.

The direction of the bands 201, 202 is denoted DD, and it will be seen that this direction coincides with the direction of travel of the document when it is a question of analyzing said document.

It is not essential that the bands 201, 202 relate to the entire document 1: in FIG. 1, there are thus distinguished two areas not concerned by the ZL coding. In the particular case of a bank note, these two ZL zones can be used for numbering.

These bands 201, 202 are also coded according to a binary coding (0 or 1), and symmetrically with respect to the axis of symmetry X′X of document 1. The coding of the bands 201, 202 is thus organized along the axis Y′Y.

It is then interesting that the graphic G of the document is printed with a pair of inks of the same shade, one of which reacts and the other not at a predetermined excitation, so as to define the division of said graphic into parallel strips.

Although it is possible to use different types of excitation (magnetic pigments, microwaves, UV radiation, radioactive source), it is interesting to choose infrared radiation. The wavelength of the infrared will then be chosen so as to obtain the best efficiency of the couple constituted by the two safety signs 100, 200, so that the response curves concerned during the analysis of the document coincide at least in part.

We will preferably choose a wavelength slightly less than a micrometer, and in particular between 0.8 and a micrometer (this is therefore the lower infrared region, which is very far from thermal infrared sometimes used for document analysis, where the wavelengths are at least three micrometers).

When the graphic design of the document is printed with a pair of inks, one of which reflects infrared and the other not, the examination of said document under infrared corresponds to an image of the type of those illustrated in FIGS. 9 and 10.

In FIG. 9, there is thus successively a band 202 coded 1 (absorbs infrared, so lets see the part concerned of the graphics as well as the zone concerned by the first watermarked network 100), a band 201 coded 0 (reflects infrared , therefore hides the graphics, thus leaving only the area concerned with the first watermarked network 100) visible, then two bands 202 coded 1. The symmetry of the coding with respect to the axis X′X then implies the successive presence of two bands 202 , a band 201, and finally a band 202.

The binary coding illustrated in FIG. 9 is therefore 10111101.

FIG. 10 illustrates another coding with the same number of parallel bands: the coding is then 01100110 (the symmetry of the coding with respect to the axis X′X is naturally always respected).

In FIGS. 9 and 10, eight parallel bands are provided, so that there are in fact 2 ⁴ , ie 16 different codings.

More generally, with 2n bands coded 0 or 1, there will be 2 ⁿ different codings.

The coding by cutting of the printed graphics can concern the front, the back, or both. In the latter case, the document will be read more easily if the same coding is used on the front and back, the corresponding bands being thus directly superimposed; this possibility may prove to be interesting insofar as it makes it possible to better resist aging.

In practice, we will choose a number of bands at least equal to the number of documents to be discriminated (this will be the case for banknotes, for example, when we use the second sign for the mechanized discrimination of the face value of the ticket analyzed), the number of strips remaining moreover limited by the technological possibilities of the analysis means working on very thin strips.

It will also be possible to print the graphics (on the front and / or back) with other inks that do not react to the excitation corresponding to the coding in parallel bands (for example to infrared radiation).

This possibility could be used for banknotes, offset printing, in particular intaglio printing easily allowing a juxtaposition of colors, thanks to the cut rollers (there are no problems of "register" with colors, because we use the same printing plate).

FIG. 2 makes it possible to better distinguish the watermarked area corresponding to the first security sign 100, as it appears seen through transparency.

The watermarked network 100 is therefore periodic (regular alternation of opaque and clear zones), and the period is noted T. In addition, as will be explained in detail below, this watermarked network comprises waves which are preferably with profile of sinusoidal surface mass.

FIG. 2 also shows that the wave of the watermarked network 100 extends in a common direction DC which is essentially not perpendicular to the direction DD of the cutting strips of the second network 200.

In this case, the aforementioned directions DC and DD form an angle β between them which is here 45 °, which allows a reading of the document in two perpendicular directions (parallel to the large edge, which is generally the case for the processing machines, especially for banknotes, or parallel to the small edge).

As a variant, it is possible to choose other values for the angle β between the two directions mentioned above, but to the detriment of the corresponding advantage. FIG. 14 illustrates a particular case where the directions DC and DD are essentially parallel, this case inducing a particular arrangement of the detection sensors, as will be described later with reference to this figure.

It should also be noted in FIG. 2 the presence of a particular phase shift for the waves of the first network 100 relative to the center of the square which is here at the intersection of the axes X′X and Y′Y of the document. The choice of such a phase shift, for example bringing as is the case here the edge of a strip at the center 0 of the square, will depend on the analysis mode used and the corresponding processing means. We will see that this allows a sensor located at any distance from the X′X or Y′Y axes to always receive the same signal (to within π or 2 π).

The arrangement illustrated in FIG. 1 remains in any case the most interesting, because the arrangement of the two superimposed networks, namely the watermarked periodic network 100 and the coded network 200 in parallel strips for cutting the printed graphics, allows a reading of the globally indifferent document (independent of the facing, orientation and direction of passage of the document).

FIG. 3 illustrates the raised face of a matrix 110 making it possible to stamp the form cloth during the production of the document, to obtain a watermarked periodic network similar to that of FIG. 2. This raised face has undulations, here sinusoids, which propagate in a common direction DC inclined at 45 °.

The raised face of the matrix 110 thus presents a succession of hollows 111 and bumps 112 (better visible in the cross section of FIG. 4), which make it possible to produce the alternately light 102 and opaque 101 areas for the watermarked network 100 of the document.

The associated curve IV of FIG. 4a, showing the variations in the surface mass in the watermarked area of the document (in the direction DC), is then in direct correspondence with the curve of the variations in the relief of the matrix 110 illustrated in FIG. 4 .

It is interesting to note in FIG. 4a that the amplitude variations of the sine waves of the watermarked network are made around the mean plane noted PM of the document (which makes it possible to have independence of reading with regard to the shaping of the document).

The period T will preferably be chosen to be large in relation to the dimensions of the document, for example of the order of 10 mm for a banknote, so that the security sign 100 is as discreet as possible. The same goes for the side of the square, which will for example be of the order of 60 mm.

The sections of FIGS. 5 to 8 also make it possible to better distinguish the particular beveling of the edges 113 of the matrix 110. This beveling is in fact organized either downwards (chamfered edges 113 ′), or upwards (chamfered edges 113˝ ) relative to the mean plane of the raised face of the matrix 110.

This results in "bevelled" edges for the watermarked area, as is apparent from curves V to VIII giving the corresponding variations in the surface mass, and this on either side of the mean plane PM of the document. A watermarked square is thus produced, the edges of which are "lace", which avoids sudden contrast transitions around the watermarked zone, and further accentuates the discretion of the security sign.

FIG. 11 illustrates (in transparency) the watermarked periodic network 100 obtained with a form cloth previously stamped with the above-mentioned matrix 110: the bevelled edges 103 of the square will be noted in particular. The opaque 101 and light 102 regions correspond to what has been previously described with reference to FIG. 2.

FIG. 12 shows on a larger scale the zone of document 1 where the two security signs 100 and 200 are superimposed.

The zones in bands 101 and 102 of the watermarked periodic network 100, alternately opaque and clear, have the same width which is equal to the half-period T / 2 of the sine wave of said network.

The inclination of these bands 101 and 102 is identified by the angle β between the directions DC and DD (the angle β here is 45 °).

FIG. 12 also makes it possible to distinguish the parallel coded strips 201, 202 from the second security sign 200 corresponding to the cutting of the printed graphics.

The coded bands have the same width e which is determined, in most cases, as a function of the watermark network, that is to say more precisely of the period T and of the angle β.

FIG. 12 shows a right triangle ABC corresponding to a particularly advantageous arrangement for reading the document, a triangle whose hypotenuse AB corresponds to the width e of each of the bands 201 or 202, and one side of which corresponds to the half-period T / 2: so we have the relation $$\text{e =}\frac{\text{<figure-callout id="2" label="T" filenames="imgf0001.png,imgf0003.png" state="{{state}}">T</figure-callout>}}{\text{2 sin β}}\text{.}$$

ce qui correspond par exemple à une largeur de bandes de 10 mm (avec six bandes), pour une période de 14,14 mm.In the particular case illustrated here, we have β = 45 °, therefore $$\text{e =}\frac{\text{<figure-callout id="2" label="T" filenames="imgf0001.png,imgf0003.png" state="{{state}}">T</figure-callout>}}{\sqrt{\text{2}}}\text{,}$$

which corresponds for example to a strip width of 10 mm (with six strips), for a period of 14.14 mm.

However, the above relationship can only be used within certain limits, that is, as long as the angle β is greater than a reference angle β _o corresponding to a bandwidth e _o reaching half the width ( l) of the document: this borderline case would indeed correspond to the presence of two bands, symmetrical with respect to the axis X′X.

For example, with a banknote whose width would be of the order of 80 mm, there will be a reference angle β _o of the order of 10 °.

When the angle β becomes less than this reference angle β _o , the width e of the strips 201, 202 for cutting the printed graphics is essentially chosen according to the coding sought.

The particular case of a zero angle is illustrated in FIG. 14: the bands 101, 102 of the first network 100 are then orthogonal to the bands 201, 202 of the second network 200, and one can then choose a width e advantageously equal to half period T / 2 (the representation would then correspond to a perfect grid of the square in six orthogonal bands).

In practice, the number of cutting strips will first be chosen as a function of the number of documents to be coded and of the manufacturing techniques making it possible to produce these coded strips, and also of symmetry constraints. This choice will also be guided by the precision of the reading machine used for document analysis. We will then determine the possible angles β, it being understood that an angle of 45 ° offers the maximum of advantages, as has been explained above.

The document thus comprising two superimposed security signs 100, 200 of the aforementioned type is very interesting insofar as the superposition of these two signs has the effect of affecting the individual reading of said signs.

We thus succeed in considerably arguing the efficiency of authentication.

When the document is a banknote, the first security sign 100 and the second security sign 200 are used to authenticate the ticket, and the second security sign 200 is used for mechanized discrimination of the face value of said note.

This will emerge more clearly from the analysis method and the associated device, which will now be described with reference to FIGS. 13 to 16.

FIG. 13 indeed illustrates the zone of document 1 where the two security signs 100 and 200 are superimposed (as for FIG. 12), with in addition a read bar 301 equipped with detection means.

The detection means are here in the form of sensors 300, with at least one sensor per coded band 201 or 202 of the second network 200 (here one per band). These means are organized in a general direction D which is perpendicular to the direction DD which is that of the scrolling of the document in the reading machine (the direction DD is also that of the coded strips 201, 202), and with a prohibition d equal to the width e of said coded bands 201 or 202.

It is also advantageous for the detection means 300 to be located on the median axis (a) of the associated bands 201 or 202 of the second network 200: this thus avoids any risk of alteration of the analysis in the event of a document shift. compared to the sensors of the reading module (there would be a loss of signal by increase in noise).

It may be a single reading bar, the sensors of which comprise transmitting and receiving means, and under which the document to be analyzed scrolls. It may alternatively be two superimposed reading bars, one of which comprises transmitting means and the other of receiving means, and between which the document scrolls. analyze. Figure 13 then schematically shows either this single bar, or one of the two superimposed bars (the other being below it).

FIG. 13 also makes it possible to understand that, when a sensor 300 associated with a coded strip 201 or 202 reads a minimum surface density (sensor at the center of an inclined strip 102, on the axis of said strip), the sensor 300 associated with the symmetrical strip 201 or 202 (conjugate strip) reads a maximum surface mass (sensor at the center of an inclined strip 101, on the axis of said strip): this results from the fact that the arrangement of the watermarked network with a sine wave profile is such that there is phase opposition of the waves on either side of the X′X axis of the document, at the same distance from said axis.

More generally, there is at all times an inter-relation between the response of a coded strip 201 or 202 and the response of the symmetric coded strip (conjugate strip), when the document scrolls under the reading bar 301.

• . on dispose des moyens de détection 300 à raison d'au moins un par bande 201, 202 du second réseau 200, ces moyens étant organisés selon une direction générale D perpendiculaire à la direction de défilement DD, avec une interdistance d égale à la largeur e des bandes parallèles 201, 202 dudit second réseau ;
• . on vérifie le codage du second réseau 200 en additionnant la réponse de chaque bande codée 0 ou 1 et de sa symétrique qui lui est conjuguée, afin d'éliminer l'influence du premier réseau 100, et en comparant les résultats obtenus aux valeurs théoriques de codage ;
• . on analyse le premier réseau 100 par soustraction des réponses de chaque bande exempte de codage codée 0 et de sa symétrique.

This thus leads to formulate the characteristic of the document analysis process, according to which:

• . there are detection means 300 at the rate of at least one per band 201, 202 of the second network 200, these means being organized in a general direction D perpendicular to the direction of travel DD, with a prohibition d equal to the width e parallel bands 201, 202 of said second network;
• . the coding of the second network 200 is verified by adding the response of each coded band 0 or 1 and of its symmetric which is conjugate to it, in order to eliminate the influence of the first network 100, and by comparing the results obtained with the theoretical values of coding;
• . the first network 100 is analyzed by subtracting the responses from each band free from coded coding 0 and from its symmetrical.

According to this method, by adding the response of each coded band and that of its conjugate band, it is possible both to eliminate the signal from the first watermarked network, and to improve the response to the coding of the bands, for example the response to the infrared: preferably for this, use a decoding by synchronous integration for each pair of coded bands (a pair consisting of a coded band and its symmetric or conjugate), then a mutual comparison of the results obtained with the theoretical coding values.

By subtracting the responses from the pairs of codeless bands (coded 0), in particular the levels of absorption of infrared radiation, it is possible to analyze the first watermarked network all the more easily when the signal-to-noise ratio of this network is clearly improved. (there is indeed a signal whose amplitude is double thanks to the phase opposition of the watermarked network between the coded bands conjugate of the same channel).

In the case of FIG. 14, for which the directions DC and DD are substantially parallel (the two superimposed networks then forming a grid of the watermarked area), it is necessary to modify the reading bar 301.

Instead of a single row of sensors 300 arranged perpendicular to the direction of travel DD, the read bar 301 then comprises two parallel rows of sensors 300 ′, 300˝, arranged perpendicular to the direction of travel DD, with one row per half strip: these two rows of sensors (each comprising three sensors 300 ′ or 300 ici here) are then offset between them by a predetermined distance d ₁ which is preferably substantially equal to the half wavelength T / 2 of the first network, so as to find the opposition of the previous phase between homologous sensors.

The 300′or 300˝ sensors of the same row are also located on the median axis (a) of the associated coded bands 201, 202 of the second network, and are equidistant from one another by a distance d substantially equal to the width. e of said coded bands.

• . on dispose des moyens de détection 300′, 300˝ à raison d'au moins un par bande 201, 202 du second réseau 200, ces moyens étant organisés selon une direction générale D perpendiculaire à la direction de défilement DD et situés sur l'axe médian a des bandes associées 201, 202, avec, d'un côté dudit axe X′X du document 1 des premiers moyens de détection 300′ alignés entre eux, et, de l'autre côté dudit axe X′X, des seconds moyens de détection 300˝ également alignés entre eux mais décalés des premiers moyens de détection 300′ d'une distance d₁ sensiblement égale à la demi-longueur d'onde T/2 du premier réseau 100 ;
• . on vérifie le codage du second réseau 200 en additionnant la réponse de chaque bande codée 0 ou 1 et de sa symétrique qui lui est conjuguée, afin d'éliminer l'influence du premier réseau 100, et en comparant les résultats obtenus aux valeurs théoriques de codage ;
• . on analyse le premier réseau 100 par soustraction des réponses de chaque bande exempte de codage codée 0 et de sa symétrique.

This thus leads to formulate the characteristic of such a variant of the analysis method, according to which:

• . there are detection means 300 ′, 300˝ at the rate of at least one per band 201, 202 of the second network 200, these means being organized in a general direction D perpendicular to the direction of travel DD and located on the axis median has associated bands 201, 202, with, on one side of said axis X′X of document 1, first detection means 300 ′ aligned with each other, and, on the other side of said axis X′X, of second means detection 300˝ also aligned with each other but offset from the first detection means 300 ′ by a distance d ₁ substantially equal to the half-wavelength T / 2 of the first network 100;
• . the coding of the second network 200 is verified by adding the response of each coded band 0 or 1 and of its symmetric which is conjugate to it, in order to eliminate the influence of the first network 100, and by comparing the results obtained with the theoretical values of coding;
• . the first network 100 is analyzed by subtracting the responses from each band free from coded coding 0 and from its symmetrical.

Thus, here again we find the same analysis process with addition of the responses of the conjugate coded bands, and subtraction of the absorption levels of the couples of bands for the channel or channels not used for coding (coded bands 0).

Such an analysis process is therefore very interesting, because it allows a double analysis of the two superimposed safety signs with a single array of sensors, and this notwithstanding the fact that the superposition of these two signs has the effect of affecting the reading. each of them.

This analysis process will be detailed below, with reference to FIG. 16 which schematically illustrates a complete device for analyzing the signals from the various sensors, in order on the one hand to verify the coding of the second network and to validate the document analyzed. when the network read is compliant, and on the other hand to analyze the first network and to validate the document analyzed when the network read is also compliant.

There are of course multiple ways of making the reading strip, as will emerge from the variants described below by way of example.

The reading strip 301 can have substantially circular equidistant openings 302 associated with each sensor 300, as illustrated in FIG. 13.

As a variant, openings in the form of slots 303 can be provided (FIG. 15a): each slot is then inclined so as to be substantially perpendicular to the direction of propagation of the wave of the first network (each slot is thus inclined according to the same angle β with respect to the direction of travel DD).

According to another variant, there are provided cruciform openings 304 (FIG. 15b), the two branches of which are respectively parallel and perpendicular to the direction of propagation of the wave of the first grating. This makes it possible to further increase the integration surface for the first network and to have a higher average value for the measured signal, since in this case an integrated sampling process is used.

According to yet another variant illustrated in FIGS. 15c and 15d, at least one of the sensors is multiple (here the six sensors are multiple). In FIG. 15c, each multiple sensor 300 is constituted by two identical identical sensors 300 ₁ arranged on either side of the median axis (a) of each coded strip 201 or 202. The response of the sensor 300 is then the sum of the responses from the two sensors 300 ₁ . In FIG. 15d, each multiple sensor 300 consists of four identical sensors 300 ₂ arranged in a square, the square being centered on the median axis (a) of each coded strip 201 or 202, and the edges of the square being parallel and perpendicular to the DD scroll direction.

It goes without saying that the variants of FIGS. 15a to 15d can be adapted to the case of the bar with two offset rows illustrated in FIG. 14, with then two offset rows of inclined or cruciform slots, or two offset rows of multiple sensors.

In general, the sensors 300 or 300 ′, 300 ˝ of the read bar 301 will preferably be organized to present the same gain and the same original setting, so as to ensure the balancing of the different channels.

The single or multiple sensors may be photo-diodes, or photo-transistors, or even photoresist cells, each of these sensors preferably being associated with optical filters to be perfectly calibrated at the desired wavelength.

We will now describe a complete device for analyzing the signals coming from the different sensors of the read bar, with reference to FIG. 16.

We find the reading bar 301, here with six sensors 300 for a document with six coded bands parallel to the direction of travel, including three sensors producing a respective signal noted SA, SB, SC, and three other sensors producing a respective signal SA ′, SB ′, SC ′ corresponding to the conjugate coded bands.

The analysis device comprises means 400 for processing the signals coming from the sensors 300.

These processing means comprise two units, each of which is associated with a network 100 or 200 of the document.

The first unit allows the verification of the coding of the second network of the scrolling document at the level of the sensor strip, and the validation of the document analyzed when this network is compliant.

This first unit firstly comprises summing means 401 associated with each pair of coded bands. The signals obtained thus correspond to signals SA + SA ′, SB + SB ′ and SC + SC ′ (these additions each time include the sum of a signal and of this same signal phase shifted by π), with preferably an amplification prior by means of intermediate amplifiers 413. These signals are sent to associated integrating means 402 allowing integration over the entire length of the document analyzed.

Signals IA, IB, IC are thus obtained associated with each pair of coded bands. These signals are sent to comparator means 403 to compare the results obtained with the theoretical coding values of the second network of the document to be analyzed.

Preferably, switching means 408, 409 are provided upstream and downstream of the integrating means, these switching means (shown here by switches) being controlled respectively by the passage of the front edge and the rear edge of the document. in front of a fixed member, such as a photo-diode (at least one of the sensors of the reading bar can alternatively provide itself an additional function of detection of the passage of the ticket, which avoids having to plan a separate photo-diode): the control of the means 408, 409 is shown here schematically by a central control unit 415.

Thanks to this fixed detection member (integrated or separate photo-diode), it is then ensured to carry out an integration over the entire length of the document. This is particularly interesting in the case of banknotes of the same width and different lengths.

The comparator means 403 first make it possible to verify that each value IA, IB, IC is indeed within a predetermined range, the limits of which are defined as a function of the inks, the opacity of the paper, and other parameters relating to the document concerned. .

The comparator means 403 are equipped with a contrast alarm 410 intervening when a difference between the results is outside a predetermined range. In this case, all the differences I _i -I _j are compared with the limits of the range, and the alarm 410 intervenes if there is no ink reacting to the known excitation (infrared radiation for example), or if the ink does not react properly to this excitation.

sont comparées aux limites de la fourchette. Cette variante est intéressante à la fois pour la symétrie des résultats si les réponses s'inversent, pour la sensibilité élevée pour une échelle donnée dans les faibles écarts de contraste, et pour le fait que l'on a une réponse maxi pour du noir et mini pour du blanc.As a variant, the contrast alarm 410 intervenes when a relationship between results is outside of a predetermined range. The comparator means 403 then comprise logarithmic report amplifiers and a window comparator (positive or negative). In this case, all Log values $$\text{((}\frac{{\text{I}}_{\text{i}}}{{\text{I}}_{\text{j}}}\text{)}$$

are compared to the limits of the range. This variant is interesting both for the symmetry of the results if the responses are reversed, for the high sensitivity for a given scale in small differences in contrast, and for the fact that there is a maximum response for black and mini for white.

Preferably, the first unit finally comprises decoding means 411 downstream of the comparator means 403, in order to identify the document, and in particular when the document is a banknote, in order to discriminate the face value of the note. These decoding means 411 have in memory the inequalities I _i <I _j for each document, which makes it possible to easily identify the analyzed document.

The second unit firstly comprises differentiating means 404 associated with each pair of coded bands. The signals obtained thus correspond to signals | SA-SA ′ |, | SB-SB ′ | and | SC-SC ′ |, again preferably with prior amplification by intermediate amplifiers 413. Each difference corresponds, due to the phase opposition for the watermarked network, to twice the starting signal free of disturbances due to the dirtiness of the document and the look of the paper.

There are also provided selector means 405, downstream of each of the differentiating means 404, in order to keep only the responses relating to the bands free of coding (coded 0). These means are shown diagrammatically here by switches controlled by the central unit 415, the switch associated with the bands SC and SC ′ (coded 0) being here closed.

It is interesting to send the signals obtained to additional summing means 412 (the signals being in phase, the signal is indeed obtained n times, with here n = 1, 2 or 3).

There are then filtering means 406 allowing filtering of the signals at the fundamental frequency of the first network, which makes it possible to isolate the useful signal. This signal is finally sent to means 407 of recognition and validation, in order to analyze the first network of the document, and to validate the document when the watermarked network is compliant, or failing to involve an associated alarm 414. These means 407 could include a window comparator on the amplitude and / or a threshold detection of the harmonic distortion, or a detection of the number of periods.

It goes without saying that the amplifier means 413, summers 401 and integrators 402 of the first unit can be grouped together in a single functional unit, and the amplifier means 413 and differentiators 404 of the second unit.

The method and the analysis device which have just been described in detail considerably increase assistance with authentication.

If the document is falsified, this can result from a non-observance of the coding in parallel bands (second network), but then the decoding circuit will not validate the document and moreover the reading of the watermarked network on the considered channel will not be not possible because of the infrared sensitive ink. This can also result from a falsification of the watermarked periodic network (first network), but then, if the amplitude is too high, detection is easy; if the phase is not respected, the signal from the channel difference is then very attenuated, and if the profile is not sinusoidal, the measurement of the harmonic distortion allows detection.

The invention is not limited to the embodiments which have just been described, but on the contrary encompasses all the variants incorporating, with equivalent means, the essential characteristics set out above.

Claims (15)

1 / Fiduciary or security document presenting a printed graphics and security signs, characterized in that it comprises two superimposed security signs (100, 200) each produced in the form of a network, including a first sign of security (100) which is in the form of a periodic watermark network, and a second security sign (200) which results from a cutting of the printed graphics (G) into parallel strips (201, 202) arranged and coded according to binary coding (0 or 1), symmetrically with respect to an axis of symmetry (X′X) of the document (1), the wave of said watermarked network extending in a common direction (DC) essentially not perpendicular to the direction (DD) of said cutting strips of the printed graphics, the superposition of these two security signs (100, 200) thus having the effect of affecting the individual reading of said signs. 2 / Document according to claim 1, characterized in that the watermarked network forming the first safety sign (100) comprises waves with sinusoidal surface mass profile. 3 / Document according to claim 2, characterized in that the amplitude variations of the waves of the watermark network (100) are around the mean plane of said document (1). 4 / Document according to one of claims 1 to 3, characterized in that the wave of the watermarked network (100) extends substantially in a common direction (DC) making an angle of 45 ° relative to the direction ( DD) strips (201, 202) for cutting the printed graphics. 5 / Document according to one of claims 1 to 4, characterized in that the watermarked network (100) defines a surface whose closed outline (C) is inside the edges (2, 3) of said document (1), said the surface being entirely crossed by strips (201, 202) for cutting the printed graphics. 6 / Document according to claim 5, characterized in that the watermarked network (100) is organized according to a square. 7 / Document according to claim 6, characterized in that the edges (103) of the square are bevelled. 8 / Document according to one of claims 1 to 7, characterized in that the strips (201, 202) for cutting the printed graphics have the same predetermined width (e) which is a function of the angle (β) between the common direction (DC) of propagation of the watermarked network wave (100) and the direction (DD) of the cutting strips (201, 202), when said angle (β) is greater than a reference angle (β _o ) corresponding to a bandwidth (e _o ) reaching half the width (l) of said document (1). 9 / Document according to claim 8, characterized in that the width (e) of the cutting strips (201, 202) is given by the formula $$\text{e =}\frac{\text{T}}{\text{2 x sinβ}}\text{,}$$

where T is the wavelength of the watermarked network (100). 10 / Document according to one of claims 1 to 7, characterized in that the strips (201, 202) for cutting the printed graphics have the same width (e) depending only on the desired coding and not on the angle ( β) between the common direction (DC) of propagation of the wave of the watermarked network (100) and the direction (DD) of the cutting strips (201, 202), when said angle (β) is less than a reference angle (β _o ) corresponding to a bandwidth (e _o ) reaching half the width (l) of said document (1). 11 / Document according to one of claims 1 to 10, characterized in that the graphics (G) of said document (1) is printed with a pair of inks of the same shade, one of which reacts and the other does not at a predetermined excitation, so as to define the cutting of said graphics into parallel strips (201, 202). 12 / Document according to claim 11, characterized in that the graphic (G) is also printed with other inks which do not react to said predetermined excitation, for example which do not absorb infrared, similarly for a possible graphics also provided on the back of said document. 13 / Document according to claim 1 and claim 11 or 12, comprising a printed graphic (G) on both sides, characterized in that the strips (201, 202) for cutting said graphic are coded on the front, or on the back , or on the front and back of said document (1). 14 / Document according to one of claims 11 to 13, characterized in that the graphics (G) of said document (1) is printed with a pair of inks, one of which reflects infrared and the other does not. 15 / Document according to one of claims 1 to 14, characterized in that said document is a banknote, the first and second security signs (100, 200) used for authentication of the note, and whose the second security sign (200) is used for mechanized discrimination of the face value of said ticket.

Images