OPTICAL SENSOR FOR EXAMINING VALUABLE DOCUMENTS

Abstract

An optical sensor for examining value documents, such that at a point in time before the check of the value documents, a self-test of the optical sensor is carried out, during which the light sources thereof are switched on, and, with the aid of monitor elements, the respective light intensity of the light source assigned to the respective monitor element is detected which impinges on the respective monitor element at the time of the self-test. During the check of a value document following the self-test, the light sources illuminate the value document, and measured values are recorded. The recorded measured values are then corrected with the aid of the light intensities detected by the monitor elements at the time of the self-test to take into account a change in the light intensity emitted by the light sources that occurs in the course of the service life of the light sources.

Claims

1.-15. (canceled)

16. A method for checking a value document introduced into a capture range of an optical sensor, wherein the optical sensor has: a photodetector line, which has a number K of several detector elements arranged next to one another, which are configured to detect in each case the intensity of the light emanating from a detection region of the value document, a line of light sources, which has a number N of several light sources arranged next to one another, which are configured to illuminate the value document, a monitor detector line, which has several photosensitive monitor elements arranged next to one another, wherein each monitor element is assigned to one of the light sources and is configured to detect the intensity of the light emitted by this light source and impinging on the respective monitor element, wherein in the method the following steps are carried out: carrying out a self-test of the optical sensor at a point in time before the value document is checked, wherein the N light sources are switched on during the self-test and the respective light intensity MS.sub.j of the light source assigned to the respective monitor element is detected with the aid of the monitor elements, said light intensity impinging on the respective monitor element at the time of the self-test, in order to verify the light intensities MS.sub.j detected by the monitor elements at the time of the self-test, wherein j=1 . . . N, introducing a value document into the capture range of the optical sensor, simultaneously switching on the N light sources in order to illuminate the value document introduced into the capture range with the light of the light sources, and recording measured values of the respective value document by means of the detector elements, wherein the recorded measured values correspond to the light intensity emanating from the respective value document as a result of the illumination and wherein the K detector elements of the photodetector line each record at least one measured value D.sub.i of the value document, wherein i=1 . . . K, correcting the measured values D.sub.i based on the light intensities MS.sub.j detected by the monitor elements at the time of the self-test, wherein the at least one measured value D.sub.i of the respective detector element is corrected during the correction in each case with a correction factor FK.sub.i that is computed individually for the respective detector element based on the light intensities MS.sub.j detected by several of the monitor elements at the time of the self-test, and checking the value document based on the measured values corrected with the aid of the respective correction factor FK.sub.i.

17. The method according to claim 16, wherein the respective light intensities MS.sub.j detected by the monitor elements at the time of the self-test, each limited by a proportional factor A.sub.ij, are computationally included in the respective correction factor FK.sub.i, which indicates which proportion of the light intensity emitted by the respective light source impinges on the respective detector element due to the optical beam path from the light sources to the detector elements.

18. The method according to claim 16, wherein the proportional factors A.sub.ij are determined empirically before the value document check, by a reference area being placed in the capture range of the detector elements instead of the value document, the light sources being switched on one after the other in order to illuminate the reference area and the light intensity emanating from the reference area and impinging on the detector elements being detected in each case.

19. The method according to claim 16, wherein the proportional factors A.sub.ij are computed before the value document check by means of a numerical simulation based on a model of the optical beam path from the N light sources to the K detector elements.

20. The method according to claim 19, wherein in the numerical simulation of the optical beam path from the N light sources to the K detector elements, a transfer matrix is computed, the matrix elements of which correspond to the proportional factors A.sub.ij, and the transfer matrix, in particular the matrix elements of the transfer matrix A corresponding to the proportional factors A.sub.ij, is/are employed to correct the measured values D.sub.i.

21. The method according to claim 19, wherein when computing the respective correction factor FK.sub.i of the respective detector element, for several or all of the N light sources, the proportional factors A.sub.ij are multiplied with the light intensities MSj detected by the respective monitor elements at the time of the self-test.

22. The method according to claim 16, wherein to compute the respective correction factor FK.sub.i, for several or all of the N light sources, the respective proportional factors A.sub.ij are multiplied with the respective light intensities MS.sub.j detected by the monitor elements at the time of the self-test and the results of these multiplications are summed up to compute the respective correction factor FK.sub.i.

23. The method according to claim 22, wherein to correct the measured values D.sub.i of the value document, the respective measured value D.sub.i is multiplied with the correction factor FK.sub.i, wherein the correction factors FK.sub.i are computed according to the following formula: $F{K}_i=\frac{{.Math.}_{j=1}^N{A}_{\mathrm{ij}}.Math.{\mathrm{MA}}_j}{{.Math.}_{j=1}^N{A}_{\mathrm{ij}}.Math.{\mathrm{MS}}_j}$

24. The method according to claim 16, wherein to compute the respective correction factor FK.sub.i of the respective detector element, for several or all of the N light sources in each case the ratio is formed between the light intensity MS.sub.j detected by the monitor elements at the time of the self-test and the light intensities MA.sub.j detected by the monitor elements at an earlier point in time, the respective ratio MS.sub.j/MA.sub.j is multiplied with the respective proportional factor A.sub.ij and the results of these multiplications are summed up in order to compute the respective correction factor FK.sub.i.

25. The method according to claim 24, wherein to correct the measured values D.sub.i of the value document, the respective measured value D.sub.i is multiplied with the respective correction factor FK.sub.i, wherein the correction factors FK.sub.i are computed according to the following formula: $F{K}_i=\frac{{.Math.}_{j=1}^N{A}_{\mathrm{ij}}}{{.Math.}_{j=1}^N{A}_{ij}.Math.\frac{{\mathrm{MS}}_j}{M{A}_j}}$

26. The method according to claim 16, wherein a calibration procedure is carried out before the self-test, in which a calibration medium is introduced into the capture range of the optical sensor and the N light sources are switched on in order to illuminate the calibration medium, wherein the monitor elements of the optical sensor each detect a light intensity MA.sub.j, which corresponds to the light intensity emitted by the respective light source during the calibration procedure, wherein j=1 . . . N, and/or the detector elements of the optical sensor each detect a light intensity DA.sub.i emanating from the calibration medium, wherein i=1 . . . K.

27. The method according to claim 26, wherein to compute the respective correction factor FK.sub.i of the respective detector element also the light intensities MA.sub.j of the light sources are considered that the monitor elements have detected within the scope of the calibration procedure, wherein the respective light intensity MA.sub.j is put in each case in mutual relation with the light intensity MS.sub.j detected by the respective monitor element at the time of the self-test, wherein j=1 . . . N.

28. The method according to claim 26, wherein the light intensities DA.sub.i detected from the calibration medium by the detector elements within the scope of the calibration procedure are employed to correct the measured values D.sub.i detected from the value document, wherein the measured values D.sub.i of the respective detector element recorded from the value document are multiplied with a further correction factor F.sub.i, which corresponds to the reciprocal of the light intensity DA.sub.i detected from the calibration medium by means of the respective detector element during the calibration procedure of the optical sensor.

29. The method according to claim 16, wherein the temperature dependency of the monitor elements is also taken into account when correcting the measured values D.sub.i, wherein, in particular with the aid of a temperature sensor installed in the optical sensor, a temperature of the monitor elements is measured at the time of the self-test and a temperature-dependent factor is determined based on the temperature measured in each case, with which the light intensities MS.sub.j detected by the monitor elements at the time of the self-test are corrected.

30. An optical sensor for checking a value document, which is introduced into a capture range of the optical sensor for its checking, wherein the optical sensor has: a photodetector line, which has a number K of several detector elements arranged next to one another, which are configured to detect in each case the intensity of the light emanating from a detection region of the value document, a line of light sources, which has a number N of several light sources arranged next to one another, which are configured to illuminate the value document, a monitor detector line, which has several photosensitive monitor elements arranged next to one another, wherein each monitor element is assigned to one of the light sources and is configured to detect the intensity MS.sub.j of the light emitted by this light source and impinging on the respective monitor element, a control device which is adapted to switch on the N light sources simultaneously for the optical check of the value document in order to illuminate the value document with the light from the light sources, and to control the detector elements of the photodetector line in such a manner that these record measured values of the value document that correspond to the light intensity emanating from the value document as a result of the illumination, wherein the K detector elements each record a measured value D.sub.i of the value document, wherein i=1 . . . K, and carry out a self-test of the optical sensor, in which the N light sources are switched on and, with the aid of the monitor elements, the respective light intensity MS.sub.j of the light source assigned to the respective monitor element is detected, which impinges on the respective monitor element at the time of the self-test, in order to verify the light intensities MS.sub.j detected by the monitor elements at the time of the self-test, wherein j=1 . . . N, an evaluation device which is adapted to correct the measured values D.sub.i of the value document based on the light intensities MS.sub.j detected by the monitor elements at the time of the self-test, wherein, when correcting, the at least one measured value D.sub.i of the respective detector element is corrected in each case with a correction factor FK.sub.i, which is computed individually for the respective detector element based on the light intensities MS.sub.j detected by several of the monitor elements at the time of the self-test, and check the value document based on the measured values corrected with the aid of the respective correction factor FK.sub.i.

Description

[0026] The invention will hereinafter be described by way of example with reference to the accompanying drawings. The figures are described as follows:

[0027] FIG. 1 structure of an optical sensor for checking the value documents,

[0028] FIG. 2 schematic sketch of the distribution of the illumination light emitted by the light sources in the value document plane,

[0029] FIG. 3 gray level representation of the proportional factors A.sub.ij determined with the transfer matrix,

[0030] FIG. 4 light intensities detected by the monitor elements at the time of the calibration procedure (MA.sub.j) and at the time of the self-test (MS.sub.j),

[0031] FIG. 5 correction factors Fk.sub.i computed for the i=1 . . . K detector elements,

[0032] FIG. 6 measured values D.sub.i of a value document recorded by the i=1 . . . K detector elements and corrected measured values D.sub.i*=FK.sub.i.Math.D.sub.i.

[0033] FIG. 1 shows the structure of an optical sensor 100 adapted to check value documents, which can be installed e.g. in a value document processing apparatus. To illuminate a value document 10 introduced into the capture range of the optical sensor 100, a line of light sources is employed, which has several light sources 1 arranged next to one another along the y-direction on a light source mount 11, for example N=12 light sources. The light sources 1 are e.g. UV LEDs, which are suitable for luminescence excitation of the value document. The light emitted by the light sources 1 is directed via a mirror 2 onto photo-sensitive monitor elements 3 (e.g. photodiodes), by means of which the light intensity emitted by the light sources 1 can be verified, which usually decreases over the service life of the light sources. There is exactly one monitor element 3 for each of the N light sources 1 (e.g. N=12 monitor elements). It is achieved by apertures 15 that one monitor element 3 each captures only the light of exactly one of the light sources 1: the monitor element 3.sub.1 captures the light from light source 1.sub.1, monitor element 3.sub.2 captures the light from light source 1.sub.2, etc. The monitor elements 3 are attached to a monitor mount 1.sub.3.

[0034] As an alternative to reflection on a mirror 2, the monitor elements 3 could also be arranged in the radiation region of the respective light source 1 in such a manner that they directly capture part of the light emitted by the light sources. Or a partly transmissive mirror could be arranged between the light sources and the value document, which allows most of the illumination light to pass through and reflects part of the illumination light onto the monitor elements. Alternatively, the monitor elements could also detect the illumination light from the light sources scattered back on a reference area.

[0035] The light emitted by the light sources 1 is directed partly directly, partly via an elliptical mirror 6 onto the value document 10 to be checked and is transmitted through the measuring window 8 of the optical sensor 100 and through a spectral filter 7, which blocks the visually visible portion of the illumination light. In addition, a further spectral filter of this type could also be provided in the beam path in front of the monitor elements 3 in order to also block the visually visible portion of the illumination light here. Alternatively, the spectral filter 7 could also be arranged immediately behind the light sources in order to spectrally filter both the light impinging on the value document and the light impinging on the monitor elements 3.

[0036] The luminescent light emanating from the value document 10 is imaged by a line of Selfoc lenses 5 supplied with a UV-blocking filter onto several detector elements 4 of a line of photodetectors, which are arranged next to one another along the y-direction on a detector element mount 14 and which has e.g. K=112 detector elements. Only a single photodetector line can be employed on which the luminescence light is imaged. With the aid of additional lenses, the luminescence light can also be imaged onto several photodetector lines which are offset from one another along the x-direction and which capture different spectral ranges of the luminescence light and have corresponding spectral filters. Alternatively, several photodetector lines of a two-dimensional image sensor can also be employed, onto which the light emanating from the value document can be imaged.

[0037] The value document 10 can be introduced statically into the capture range of the optical sensor 100 for its check. However, the value document is preferably transported past the optical sensor along the x-direction in order to successively scan the various sections of the value document with the optical sensor. The transport past is achieved e.g. by means of appropriate means of transport, e.g. a conveyor belt and/or transport rollers, which are employed in a value document processing apparatus for transporting the value documents.

[0038] A control device 30 of the optical sensor 100 is connected to the light sources 1, the monitor elements 3 and the detector elements 4 and has appropriate hardware and software which ensures that the light sources 1 are switched on simultaneously and the detector elements 4 are caused simultaneously and/or or after the illumination to capture luminescence measured values D.sub.i of the value document 10. The measured values D.sub.i (i=1 . . . K) of the value document captured by the detector elements 4 are transmitted to an evaluation device 20 connected to the detector elements.

[0039] The control device 30 is adapted to carry out a self-test of the optical sensor. The self-test can be carried out at a point in time immediately before the value documents are checked, e.g. when starting the value document processing apparatus, or in the gap between two value documents to be checked transported past the sensor. During the self-test, the control device causes the N light sources 1 to be switched on simultaneously and the respective light intensity MS.sub.j (j=1 . . . N) of the light source 1 assigned to the respective monitor element to be detected. For example, during the self-test, the light intensities MS.sub.j (j=1 . . . N) detected by the monitor elements at the time of the self-test are N) are verified for whether they exceed a specific threshold value, which is required to illuminate the value document, e.g. for the optical excitation of a measurable luminescence. The control device 30 is also connected to the evaluation device 20 in order to transmit the light intensities MS.sub.j (j=1 . . . N) detected by the monitor elements at the time of the self-test to the evaluation device 20, and possibly also the light intensities MA.sub.j (j=1 . . . N) of a calibration medium detected by the monitor elements during the calibration procedure.

[0040] The evaluation device 20 corrects the measured values D.sub.i of the respective value document with the aid of correction factors FK.sub.i, in which the light intensities MS.sub.j (j=1 . . . N) detected by the monitor elements 3 at the time of the self-test are incorporated, in order to compute corrected measured values

D.sub.i*=FK.sub.i.Math.D.sub.i(i=1 . . . K)  (1)

The evaluation device 20 subsequently checks the value document by means of the corrected measured values D.sub.i*, e.g. its authenticity, condition or the type of value document. To check the value document, the corrected measured values D.sub.i* are e.g. compared with at least one reference value that is expected for the respective value document or the respective section of the value document. The corrected measured values can also be summarized over a section (region of interest, ROI) of the value document, e.g. be averaged before the comparison with the respective reference value is carried out.

[0041] When manufacturing the optical sensor, N light sources that are as identical as possible are preferably selected, which hardly differ in their emitted light intensity, and the photodetector line is preferably selected in such a manner that the sensitivities of the individual detector elements hardly differ. In this case, a correction of the measured values D.sub.i alone with the aid of the factors FK.sub.i according to formula (1) is preferred.

[0042] Otherwise, if it is unavoidable in the manufacture of the optical sensor that the individual light sources and/or the individual detector elements are different, before the delivery of the optical sensor, the sensor manufacturer preferably carries out a calibration procedure for the optical sensor, in which the possibly different emission intensities of the light sources and the possibly different sensitivities of the detector elements are checked and quantitatively determined. During the calibration procedure, a calibration medium with homogeneous optical properties, e.g. a homogeneous white surface, is introduced into the capture range of the optical sensor 100. Then the N light sources 1 are switched on simultaneously in order to illuminate the calibration medium and the reflected light intensity DA.sub.i (i=1 . . . K) emanating from the calibration medium is detected by means of the detector elements 4 of the optical sensor.

[0043] If it turns out during the calibration procedure that the individual light sources 1 and/or the individual detector elements 4 are actually significantly different, the measured values D.sub.i of the value documents are subjected to a further correction by means of a further correction factor which is likewise multiplied with the measured values D.sub.i:

D.sub.i*=F.sub.i.Math.Fk.sub.i.Math.D.sub.i  (2)

The further correction factor F.sub.1 results from the measured values of the detector elements DA.sub.i captured from the calibration medium during the calibration procedure according to the formula

F.sub.i=c/DA.sub.i  (3)

wherein a fixed numerical value is assumed for c.

[0044] During the calibration procedure, the light intensities MA.sub.j (j=1 . . . N) emitted by the N light sources 1 and impinging on the N monitor elements 3 are also detected with the aid of the monitor elements 3. When determining the correction factor FK.sub.i, these are employed as benchmark values for the light intensities MS.sub.j (j=1 . . . N) emitted by the N light sources 1 during the self-test and incident on the N monitor elements 3.

[0045] If necessary, the calibration procedure can also be repeated from time to time after delivery of the optical sensor (e.g. monthly or every six months) in order to capture the change in the light intensity of the light sources over the course of the service life. However, since the calibration procedure is costly (use of service personnel), instead of repeating the calibration procedure frequently, it is advantageous to utilize a self-test of the optical sensor as an alternative to capture the change in the light intensity of the light sources over the course of the service life. During the self-tests of the optical sensor, the light intensities MS.sub.j (j=1 . . . N) of the light sources 1 incident on the monitor elements 3 are detected regularly (e.g. before each start of the value document processing apparatus). Since the self-test is carried out closely in time to the value-document check, the light intensity of the respective light source noted during the respective self-test corresponds fairly exactly to the light intensity of the respective light source 1 present during the value-document check.

[0046] In the following, the determination of the correction factors FK.sub.i for the K detector elements (i=1 . . . K) with the aid of the light intensities MS.sub.j (j=1 . . . N) detected by the monitor elements 3 at the time of the self-test is described. The correction factors can be determined within the scope of the self-test, e.g. immediately before the check of the value documents.

[0047] To determine the correction factors FK.sub.i for the K detector elements (i=1 . . . K) it is determined individually for each detector element 4 what proportion of the light intensity emitted by the respective j-th light source (j=1 . . . N)—considering the optical beam path between the light sources and the detector elements—impinges on the respective i-th detector element (i=1 . . . K). Each of these proportions is represented by a proportional factor A.sub.ij.

[0048] Into the proportional factor A.sub.ij, j=1 . . . N, there are incorporated e.g. the position of the respective j-th light source along the line of light sources and the position of the respective i-th detector element along the photodetector line and/or their relative position or their distance along the line direction (y-direction), as well as the radiation angle of the light sources (approximately the same for all light sources), the angle of reception of the detector elements (approximately the same for all detector elements) and the distance of the light sources 1 and of the detector elements 4 from the value document plane of a value document introduced into the capture range of the optical sensor.

[0049] FIG. 2 shows an example of the distribution of the illumination light emitted from the light sources 1.sub.1, 1.sub.2, . . . , 1.sub.N. The capture ranges E.sub.1, E.sub.2, . . . , E.sub.K of the individual detector elements (i=1 . . . K) are entered schematically in the value document plane, which result from the optical imaging by means of the line of Selfoc lenses 5. It can be seen that the light contribution radiated by a light source arranged on the left edge of the line of light sources (e.g. the light source 1.sub.1) onto the capture ranges (e.g. E.sub.1, E.sub.2, E.sub.3) of the detector elements 4 located at the left edge of the photodetector line is much larger than its light contribution to the capture range (e.g. E.sub.K) of a detector element 4 located on the right edge of the photodetector line. The proportional factors A.sub.11, A.sub.21, A.sub.31 of the light source 1.sub.1 on the detector elements located on the left edge of the photodetector line are correspondingly large compared to its proportional factor A.sub.K1 on the detector element located on the right edge of the photodetector line.

[0050] The proportional factors A.sub.ij can be computed quantitatively by means of a numerical simulation based on a model of the optical beam path, in which the optical beam path from the N light sources of the optical sensor to the K detector elements of the optical sensor is modeled.

[0051] For the optical simulation, in the case of the optical sensor 100 it is assumed, for example, that the light sources have Lambertian radiation, the distance h of the light sources 1 from the value document plane is 30 mm, the light sources 1 are at a distance of 10.5 mm from one another and the detector elements 4 are at a distance of 1 mm from one another (in the y-direction). In the numerical simulation, e.g. a transfer matrix A is computed, the matrix elements of which have the proportional factors A.sub.ij, (i=1 . . . K, j=1 . . . N).

[0052] The proportional factors A.sub.ij, which were computed by the numerical simulation of the optical beam path of the optical sensor 100, are shown in FIG. 3 in the form of gray scale values. The proportional factors A.sub.ij are represented for the N=12 (j=1 . . . 12) light sources from left to right and for the K=112 detector elements (i=1 . . . 112) from top to bottom. As to be expected from the above description of FIG. 2, the largest proportional factors A.sub.ij result in the “diagonal” of FIG. 3. As an alternative to the numerical simulation of the optical beam path, the proportional factors A.sub.ij can also be determined by measuring the individual light intensity emitted by the light sources and impinging on the respective detector element.

[0053] The illumination intensity B.sub.i decisive for the respective i-th detector element (i=1 . . . K), is generally obtained by summing up the light intensities L.sub.j (j=1 . . . N) emitted by the light sources, each weighted by the proportional factor A.sub.ij, to

B.sub.i=Σ.sub.j=1.sup.NA.sub.ij.Math.L.sub.j  (4)

and in matrix notation

[00001]$\begin{array}{cc}\left(\begin{array}{c}{B}_1\\ B_2\\ .Math.\\ B_K\end{array}\right)=\left(\begin{array}{ccc}{A}_{1,1}& .Math.& A_{1,N}\\ .Math.& \ddots & .Math.\\ A_{K,1}& .Math.& A_{K,N}\end{array}\right).Math.\left(\begin{array}{c}{L}_1\\ L_2\\ .Math.\\ L_N\end{array}\right).& \left(5\right)\end{array}$

[0054] Assuming that the monitor elements each only detect a specific, fixed proportion 1/ß of the light intensity emitted by the light sources, the light intensity L.sub.j (j=1 . . . N) emitted by the light sources can be computed from the light intensities MS.sub.j detected by the monitor elements in the self-test by L.sub.j=β.Math.MS.sub.j. At the time of the calibration procedure, L.sub.j=β.Math.MA.sub.j is assumed accordingly.

[0055] For the time of the self-test, there results the illumination intensity BS.sub.i relevant for the respective i-th detector element (i=1 . . . K) by summing up the light intensities MS.sub.j (j=1 . . . N) detected by the monitor elements during the self-test, each weighted by the proportional factor A.sub.ij

BS.sub.i=Σ.sub.j=1.sup.NA.sub.ij.Math.β.Math.MS.sub.j  (6)

and for the time of the calibration procedure there results accordingly

BA.sub.i=Σ.sub.j=1.sup.NA.sub.ij.Math.β.Math.MA.sub.j  (7)

[0056] In a first embodiment example, the correction factors FK.sub.i of the individual detector elements (i=1 . . . K) are computed from the two illumination intensities BS.sub.i and BA.sub.i according to the following formula:

[00002]$\begin{array}{cc}\begin{array}{cc}{\mathrm{FK}}_i=\frac{{\mathrm{BA}}_i}{{\mathrm{BS}}_i}=\frac{{.Math.}_{j=1}^N{A}_{\mathrm{ij}}.Math.{\mathrm{MA}}_j}{{.Math.}_{j=1}^N{A}_{\mathrm{ij}}.Math.{\mathrm{MS}}_j}& \mathrm{wherein}i=1.Math.K,j=1.Math.N.\end{array}& \left(8\right)\end{array}$

Accordingly, the respective correction factor FK.sub.i results from the quotient of the illumination intensity BA.sub.i=Σ.sub.j=1.sup.NA.sub.ij.Math.MA.sup.j present during the calibration procedure and the illumination intensity BS.sub.i=Σ.sub.j=1.sup.NA.sub.ij.Math.MA.sup.j present during the self-test. The computation of the correction factors FK.sub.i according to the formula (8) is an approximation in which it is assumed for the sake of simplicity that all monitor elements have the same sensitivity and the optical image on each of the monitor elements is the same.

[0057] In a second embodiment example, the correction factors FK.sub.i of the individual detector elements (i=1 . . . K) are computed according to the following formula:

[00003]$\begin{array}{cc}\begin{array}{cc}{\mathrm{FK}}_i=\frac{{.Math.}_{j=1}^N{A}_{\mathrm{ij}}}{{.Math.}_{j=1}^N{A}_{\mathrm{ij}}.Math.\frac{{\mathrm{MS}}_j}{{\mathrm{MA}}_j}},& \mathrm{wherein}i=1.Math.K,j=1.Math.N.\end{array}& \left(9\right)\end{array}$

The computation of the correction factors FK.sub.i according to the formula (9) is also an approximation, wherein it is assumed that the light sources 1 emit approximately the same light intensity at the time of the calibration procedure.

[0058] The correction factors FK.sub.i can either be computed by the control device within the scope of the self-test and forwarded thereby to the evaluation device 20. However, the correction factors FK.sub.i can also first be computed by the evaluation device, which receives the respective light intensities MS.sub.j (j=1 . . . N) detected during the self-test from the control device. After the correction factors FK.sub.i have been determined according to formula (8) or (9) from the illumination intensities MS.sub.j and MA.sub.j (j=1 . . . N) detected during the calibration procedure and the self-test while employing the proportional factors A.sub.ij shown in FIG. 3, the evaluation device 20 can carry out a correction of the measured values D.sub.i (i=1 . . . K) of a value document according to formula (1).

[0059] As an example, the illumination intensities MS.sub.j and MA.sub.j (j=1 . . . N) for all N=12 light sources 1 detected by the N=12 monitor elements 3 during the calibration procedure and during a self-test of the optical sensor are represented in FIG. 4. During the self-test, a greatly reduced light intensity of the second light source (j=2) was noted, which is reduced by about half compared to the light intensity measured earlier during the calibration procedure. From these values for MS.sub.j and MA.sub.j (j=1 . . . N) there result (while employing the proportional factors A.sub.ij shown in FIG. 3) from formula (8) or (9) the correction factors FK.sub.i represented in FIG. 5 for all K=112 detector elements 4.

[0060] With the correction factors FK.sub.i from FIG. 5, the measured values D.sub.i of the value document captured by the individual detector elements 4 (cf. FIG. 6) are corrected according to formula (1), e.g. multiplied, in order to compute the changed light intensity over the course of the service life, in particular the light intensity of the second light source, which has been strongly reduced since the calibration procedure, from the measured values detected. FIG. 6 shows the measured values D.sub.i*=FK.sub.i.Math.D.sub.i corrected in the example under consideration. For the detector elements i=1 to around i=45, this results in a clear upward correction of the measured values D.sub.i.

[0061] In addition to the aging of the light sources 1 over the course of the service life, the temperature dependency of the monitor elements 3 can also be corrected if necessary. For this purpose, the temperature of the monitor elements at the time of the self-test T.sub.S is measured with the aid of a temperature sensor 16 installed near the monitor elements 3 (e.g. on the monitor mount 13) and taken into account when correcting the measured values D.sub.i. If the temperature of the monitor elements is always approximately the same at the time of a calibration procedure carried out by the sensor manufacturer, it is not necessary to take the temperature measured at the time of the calibration procedure into account. Otherwise, if a calibration procedure takes place with a greatly changed temperature of the monitor elements, the temperature of the monitor elements T.sub.A measured during the calibration procedure is preferably also deposited in the optical sensor and taken into account when correcting the measured values D.sub.i.

[0062] For example, a table can be stored in the evaluation device 20 of the optical sensor, in which a correction factor t as a function of the temperature T is assigned in each case to different temperatures T of the monitor elements. To correct the temperature dependency of the monitor elements 3, the light intensities MS.sub.j detected by the monitor elements at the time of the self-test are then replaced in the equations (8) or (9) by MS.sub.j′=t(T.sub.S) MS.sub.j, wherein t(T.sub.S) results from the temperature T.sub.S measured during the self-test according to the above-mentioned table. It is assumed that the correction factors t(T.sub.S) are approximately the same for all monitor elements 3. Analogously, the light intensities MA.sub.j detected by the monitor elements at the time of the calibration procedure can be replaced by MA′.sub.j=t(T.sub.A).Math.MA.sub.j, wherein t(T.sub.A) results from the temperature T.sub.A measured during the calibration procedure according to the above-mentioned table. Alternatively, the measured values D.sub.i of the value document captured by the individual detector elements 4 could simply be multiplied with an additional factor t(T.sub.A)/t(T.sub.S).

[0063] Moreover, the temperature dependency of the detector elements can optionally be measured during the self-test with corresponding additional temperature sensors and—if the detector elements have a correspondingly great temperature dependency—any temperature changes since the calibration procedure can be taken into account when correcting the measured values D.sub.i.