EP2418610A1 - Method and system for counting stacked elements - Google Patents

Abstract

The method involves storing a raw signal corresponding to the laser telemeter measurement of the level variation on a line of points traversing a wafer (300) of a stack (30) of elements (31). The raw signal is analysed to estimate parameters relative to the structure of the stack. The stored raw signal is processed by statistical processing based on the estimated parameters. The processed signal is analysed to determine the number of stacked elements, where the parameters are average thickness of the elements of the stack, and dispersion factor of the thickness of the elements of the stack. An independent claim is also included for a device for counting stacked elements.

Description

The present invention relates to the non-contact counting of stacked elements and more particularly to the counting of postal folds.

STATE OF THE ART

There are devices for counting thin media sheets by illuminating with an LED array the edge of a plurality of sheets as described in the European patent application. EP-A2-0743616 to count packs of photographs. In such an apparatus, a network of linear CCD cameras receives light reflected from the sheets and generates a waveform signal corresponding to the intensity variations of the reflected light.

There are also other non-contact counting systems for counting envelopes and magazines as described for example in patents US 4,384,195 and US5,221,837 .

However, the increasing specifics of postal folds associated with a smaller volume of mail result in increasing heterogeneity postal folds grouped in a stack for counting. Thus, the counting methods evoked face a variety of technical challenges such as taking into account the presence of reflective envelopes, envelopes glued together, envelopes not uniformly arranged in the stack, envelopes arranged fan, envelopes of different proportions, damaged envelopes, etc. The current counting methods are mainly intended for the counting of homogeneous elements and these methods are faulted by the problems mentioned above.

The present invention provides a method and a system for counting stacked elements which particularly provides excellent reliability in the counting of stacked elements, even in cases where said elements are in the form of a heterogeneous stack.

SUMMARY OF THE INVENTION

For this purpose, according to a first aspect, the invention proposes a method of counting stacked elements forming on a slice of the stack at least one line of variation of level. The method includes measuring by telemetry a level variation on a line of points crossing the edge of the stack and recording a corresponding raw signal. The method then comprises analyzing the raw signal to estimate one or more parameters relating to the structure of the stack followed by the processing of the raw signal by one or more statistical processing based on the estimated parameters and the signal analysis. processed to determine the number of stacked items.

The present method notably implements the recording of the telemetric signal relating to the entire stack and the analysis of said signal to assess the structure of the stack prior to signal processing. This makes it possible to obtain information on the stack of elements, for example recognize inhomogeneities, so as to apply a suitable treatment. The present method can make it possible to customize raw signal processing and improve the reliability of the counting of inhomogeneous cell elements. The count made on the basis of a raw signal from a measurement signal recorded over the entire width of the slice is more reliable than that achieved by the methods of the prior art.

The method may include estimating an average thickness of the cells of the stack and a dispersion factor of the thicknesses of the cells in the stack. This makes it possible to obtain information on the composition of the elements in the stack.

Advantageously, the raw signal processing step comprises determining a parasitic threshold. The interference threshold may be used to define a spreading limit below which a probability of finding a peak in the raw signal corresponding to a cell element is low.

Advantageously, if the dispersion factor of the stack is lower than a predetermined value, then the interference threshold is proportional to the product of the dispersion factor and the average thickness. Alternatively, if the dispersion factor of the stack is greater than said predetermined value, then the interference threshold is a function of an average thickness of the elements of the stack whose thickness is less than a limiting thickness. The limiting thickness can for example be determined on the basis of the dispersion coefficient and the first average thickness and the average thickness of the thin elements of the stack can be obtained by analysis of the raw signal.

The present definition of the interference threshold makes it possible to take into account inhomogeneous cells where elements of very different proportions can coexist. Thus, when the dispersion factor is high, the interference threshold is a function of the thickness of the weakest elements in order to process the signal with sufficient accuracy. This makes it possible not to underestimate the count.

Advantageously, the average thickness of the elements of the stack is determined according to a raw number of stacked elements estimated on the basis of the recorded raw signal. This minimizes the calculations required to determine the average thickness of the cells in the stack.

According to one embodiment, the raw signal processing includes the suppression of parasitic peaks based on the interference threshold. This makes it possible to suppress parasitic peaks that do not correspond to elements of the stack.

According to one embodiment, the step of processing the recorded raw signal comprises reconstructing interrupts based on the slope of the signal before and after the interruption when the interruption is longer than the interference threshold. This makes it possible to reconstruct the signal accurately by keeping the slope of the signal at the edges of the interruption.

According to one embodiment, the step of processing the recorded raw signal comprises reconstructing interrupts based on the value of the signal before and after the interruption when the interrupt is shorter than the interference threshold. This allows to reconstruct the signal simply when the probability is low that a peak corresponding to an element can be masked by the interruption.

According to one embodiment, the step of processing the recorded raw signal comprises sliding average filtering of the signal, the number of points of the signal on which said sliding average is calculated being determined on the basis of the interference threshold. This makes it possible to filter the signal by taking into account the structure of the stack in order to limit the peeling of peaks corresponding to an element.

The step of analyzing the processed signal may comprise, for example, the count of the number of sign changes of the derivative of the processed signal and / or the counting of the number of changes of direction of the signal processed point-to-point, and / or a control presence of glued elements in the stack based on the recorded raw signal.

Advantageously, the implementation of several of these methods is performed to increase the reliability of the count.

Alternatively, the recording step may comprise recording a plurality of raw signals corresponding to the level variation measurement on one or more lines of points traversing the entire slice of the stack. This recording may be performed using a plurality of rangefinders, a rangefinder comprising a plurality of beams and / or scanning the wafer of the stack a plurality of times. This also increases the reliability of the count.

Advantageously, each of the raw signals is recorded and processed and the convergence of the count results of the stacked elements provided by the plurality of processed signals is verified.

According to a second aspect, the invention relates to a device for counting stacked elements comprising a telemetry measurement system adapted to measure a signal corresponding to the level variation of at least one line crossing a slice of the stack formed. by said stacked elements; data storage means adapted to record said signal; calculation means adapted to implement the counting method according to the first aspect of the invention.

BRIEF DESCRIPTION OF THE FIGURES

• La figure 1 représente un schéma de fonctionnement d'un télémètre à triangulation pour le comptage d'enveloppes.
• La figure 2 représente un dispositif de comptage d'éléments empilés selon un mode de réalisation de l'invention.
• La figure 3 illustre schématiquement des étapes d'une méthode de comptage d'éléments empilés selon un mode de réalisation de l'invention.
• Les figures 4A et 4B illustrent un pic parasite dans un signal brut enregistré selon un mode de réalisation de l'invention.
• Les figures 5Aà 5D illustrent une interruption parasite dans un signal brut enregistré selon un mode de réalisation de l'invention.
• Les figures 6A et 6B illustrent la reconstruction d'une interruption dans un signal brut enregistré selon un mode de réalisation d'une étape de traitement du signal de l'invention.
• Les figures 7Aà 7C illustrent la reconstruction d'une interruption dans un signal brut enregistré selon un mode de réalisation d'une étape de traitement du signal de l'invention.
• Les figures 8A et 8B illustrent la reconstruction d'une interruption dans un signal brut enregistré selon un mode de réalisation d'une étape de traitement du signal de l'invention.
• Les figures 9A et 9B illustrent respectivement un signal brut enregistré et un signal traité selon un mode de réalisation de l'invention.
• La figure 10 illustre un schéma général d'une détection d'éléments collés dans la pile selon un mode de réalisation de l'invention.
• Les figures 11A et 11B illustrent des étapes intermédiaires de la détection d'éléments collés dans un mode de réalisation de l'invention.

Other features and advantages of the invention will appear on reading the description which follows, illustrated by the following figures:

• The figure 1 represents an operation diagram of a triangulation range finder for counting envelopes.
• The figure 2 represents a device for counting stacked elements according to one embodiment of the invention.
• The figure 3 schematically illustrates steps of a method of counting stacked elements according to one embodiment of the invention.
• The Figures 4A and 4B illustrate a parasitic peak in a raw signal recorded according to an embodiment of the invention.
• The Figures 5A at 5D illustrate a parasitic interruption in a raw signal recorded according to a mode of embodiment of the invention.
• The Figures 6A and 6B illustrate the reconstruction of an interrupt in a raw signal recorded according to an embodiment of a signal processing step of the invention.
• The Figures 7A at 7C illustrate the reconstruction of an interrupt in a raw signal recorded according to an embodiment of a signal processing step of the invention.
• The Figures 8A and 8B illustrate the reconstruction of an interrupt in a raw signal recorded according to an embodiment of a signal processing step of the invention.
• The Figures 9A and 9B respectively illustrate a raw signal recorded and a processed signal according to an embodiment of the invention.
• The figure 10 illustrates a general diagram of a detection of elements stuck in the stack according to one embodiment of the invention.
• The Figures 11A and 11B illustrate intermediate steps of the detection of glued elements in one embodiment of the invention.

DETAILED DESCRIPTION

The present method relates to the non-contact counting of stacked elements and more particularly of substantially flat elements, such as for example a magazine, an envelope, a bank note, a credit card, etc. The elements are arranged so that they form on a slice of the stack at least one level variation line that can be detected by telemetry, for example by laser telemetry.

In the present application, an "edge" of the stacked element refers to a face of the reduced surface element that extends over almost a single dimension. For example, a rectangular envelope comprises four edges within the meaning of this description.

A slice of the stack of elements is thus formed by a grouping of edges of stacked elements.

The general principle of a triangulation rangefinder applied to the counting of stacked elements, for example envelopes, is detailed on the diagram of the figure 1 . A rangefinder 10 is generally composed of a transmission source 11 emitting a light beam 13, for example a laser emission source, and a sensor 12, for example a CCD type matrix sensor. The sensor is arranged to receive a retroreflected beam by a wafer 300 of a stack 30 of elements 31 which passes in front of the light beam 13. The elements 31 are arranged on one of their edges 311, that is to say on a face of the element 31 of reduced surface which extends over almost a single dimension. The term slice 300 of the stack 30 refers to one of the four faces of the stack 30 formed by edges 310 of the elements 31 on which the light beam 13 is reflected. For example, it may be edges opposite the edges 311 on which the elements are arranged.

The beam retroreflected by the element 31 is received by the sensor 12. The sensor 12 provides in response an electrical signal whose value is a function of the position of a point of impact of the incident light beam 13 on the element 31. Indeed, the position of the impact of the beam 13 on the element 31 is geometrically related to the impact position of the beam retroreflected on the sensor 12. The electrical signal supplied by the sensor can be viewed on a screen 23 and can be used to deduce a distance between the light source 11 and the point of impact on the element 31 of the incident beam 13 from the source bright 11.

The figure 2 illustrates a device for counting stacked elements according to one embodiment of the invention. Similar elements on the figure 1 and 2 are noted by the same references. The device comprises a rangefinder 10 comprising a light source 11 which emits a light beam 13, for example a laser, and a sensor 12, for example a CCD matrix sensor. The sensor 12 is arranged to receive a retroreflected beam by the wafer 300 of a stack 30 of elements 31 which passes in front of the beam 13. The beam retroreflected by the wafer 300 is received by the sensor 12. The sensor 12 provides in response a raw electrical signal 40 which is transmitted to a computer system 20. The computer system 20 comprises a computer data storage unit 21, a computing unit 22 and a display unit 23. The computing unit 22 processes the raw signal 40 recorded on the storage unit 21 to obtain a processed signal 50 according to a process step of the method described in the following description. The raw signal 40 and / or the processed signal 50 can be displayed on the display unit 23.

A counting method according to one embodiment of the invention is now described with reference to Figures 3 to 11 . In a first step S1, a scan of the slice 300 of the stack of elements is performed. The scanning is carried out so as to obtain a telemetric measurement on a line of points passing through the entire portion 300 of the cell 30. In one embodiment, the telemeter 10 has a fixed position and the cell 30 is moved relative to the telemeter 10, for example using a In another embodiment, the stack 30 of elements is fixed and the rangefinder 10 is moved relative to the stack 30, for example by using a trolley on which the telemeter 10 is mounted.

In a second recording step S2, the raw signal 40 at the output of the rangefinder 10 is recorded on the storage unit 21 of the computer system 20. The raw signal 40 is in the form of a time record of the amplitude of the electrical signal measured at the output of the sensor 12 during the scanning of the battery. The raw signal 40 corresponds directly to a level variation measurement of the line of points passing through the entire portion 300 of the stack 30.

où p, k, N₁, N₂ sont des entiers.In the following, we consider that the raw signal is a discrete signal whose values over time can be described by the values (X ₁ ... X _n ). A derivative of the raw signal can be determined between two clouds of N ₁ consecutive points whose centroids are separated by N ₂ points by the formula: $$X_k^{\text{'}}=\frac{1}{{\mathrm{NOT}}_1+{\mathrm{NOT}}_2}\left(\frac{1}{{\mathrm{NOT}}_1}\underset{p=0}{\overset{{\mathrm{NOT}}_1-1}{\Sigma }}{X}_{k+{\mathrm{NOT}}_1+{\mathrm{NOT}}_2+p}-X_{k+p}\right),$$

where p, k, N ₁ , N ₂ are integers.

The derivative thus defined represents a slope between two means of N ₁ consecutive points of the signal separated by N ₂ points. In the following, a computed point-to-point derivative refers to a derivative calculation according to the preceding formula in which N ₁ = 1 and N ₂ = 0 are fixed.

In addition, the wafer 300 is formed by the edges 310 of the elements 31 forming a level line or ridge line. As a result, the corresponding raw signal comprises amplitude peaks. In the following, the term "peak" is generally used to denote amplitude peaks in the raw signal. The term "spreading" of a peak refers to the largest width of a peak in the raw signal. The spreading of a peak corresponds to a measured thickness of an element 31.

In a third step S3 of analyzing the raw signal, a mean spreading of the peaks, a dispersion factor of the spreads of the peaks with respect to the average spreading and an average amplitude of the peaks can be determined. These parameters relating to the structure of the stack can make it possible to adapt the treatment in the rest of the method. The average peak spreading corresponds to the average thickness of the elements 31 of the stack 30 and can be determined by the quotient of the sum of the peak spreads in the raw signal by the number of peaks. The scatter factor of peak spreads over average spread can be deduced from the standard deviation of peak spreading over mean spreading. The dispersion factor is an indicator of the inhomogeneity of the elements 31 of the stack 30.

In a first embodiment, the average spreading of the peaks can be determined on the basis of an estimate of the number of stacked elements (also called the raw number in the remainder of the description). The estimation of the raw number of stacked elements can for example be carried out by the counting of the peaks in the raw signal. Such an enumeration can for example be performed by analyzing the derivative of the raw signal 40. More precisely, the counting of the peaks in the raw signal can be based on counting the sign changes of the derivative of the computed point-to-point signal. The average peak spread in the raw signal can then be obtained by the quotient of the product of the thickness of the stack and a basic resolution of the telemetric record by the raw number of elements. The basic resolution of the telemetric recording is for example defined by the quotient of a sampling frequency of the rangefinder by a speed of movement of the edge of the stack relative to the rangefinder. The estimation step can be conducted using the computing means 22 of the computer system 20 connected to the rangefinder 10.

In a second embodiment, the average spreading of the peaks can be deduced from a fundamental frequency resulting from a spectral analysis of the raw signal, for example by decomposing the raw signal in Fourier series and looking for the average frequency corresponding to the frequency at which the cumulative energy crosses a threshold of 50% of the total energy of the spectrum. This average frequency corresponds to an average period of folds. Given that the speed of travel of the wafer of the stack relative to the rangefinder is constant, the average thickness of the elements 31 is deduced therefrom.

The scatter factor of peak spreads over average spread can be determined by the quotient of the standard deviation of peak spread and average spread.

In a fourth step S4 of treatment of signal, the raw signal 40 is processed by one or more statistical processing based on the parameters estimated from the raw signal. This improves the quality of the recorded signal. The processing step is performed on the basis of the recorded raw signal 40 corresponding to the entire slice of the stack taking into account the estimated structure of the stack. This makes it possible to process the raw signal by taking into account the inhomogeneity of the stack and to identify abnormalities in the raw signal. The applicant has shown that by treating the entire measured signal over a whole line, a much more reliable count can be obtained even with a stack of heterogeneous elements.

The processing of the raw signal 40 can be performed by considering a parasitic threshold in the raw signal. The parasitic threshold represents a threshold below which the spreading of a peak can be considered abnormal for a given cell. The interference threshold may result from an observation of the stack of elements by the user and a visual appreciation of the presence of fine elements in the stack. Setting the interference threshold may make it possible to adapt the processing of the raw signal according to the stack of elements to improve the quality of the signal without unduly suppressing a peak corresponding to an element of the stack.

Advantageously, the interference threshold can be determined on the basis of the recorded raw signal. For example, if the dispersion factor is less than a predetermined value, for example between 0.4 and 0.5, the interference threshold may be proportional to the product of the average spread of the peaks in the raw signal. by the dispersion factor of peak spreading. This makes it possible to take into account the disparity of the stacked elements in the signal processing.

For example, with reference to the first embodiment for determining the average spread, assuming the sampling frequency f of the range finder equal to 2000 points per second, the speed v of displacement of the stacked elements relative to the range finder equal to 0.1 m / s, the width D of the stack equal to 0.5 m, the estimated rough number B of stacked elements equal to 623 and the dispersion factor σ equal to 50%, we obtain a parasitic threshold S _p : $$\mathit{sp}=\frac{D\times f}{B\times v}\times \sigma \cong 8\mathit{points}$$

Alternatively, when the dispersion factor is greater than the predetermined value, it is possible to filter the low frequencies in the raw signal and determine the interference threshold based on a mean peak spreading value in the filtered raw signal. In other words, the interference threshold is then a function of a second average spread calculated on the thinnest peaks whose thickness is less than a limiting thickness. The limiting thickness can be determined on the basis of the first average thickness and the dispersion factor. This allows, in case of strong inhomogeneity of the elements of the stack, not to neglect fine peaks corresponding to small proportion elements in the stack. The limiting thickness can be determined as an average thickness of a portion of the thinnest folds. For example, we can classify the elements in order of increasing thickness and take the average thickness of the elements included in the 20% the finer.

The signal processing step S4 comprises one or more statistical treatments applied to the signal such as spurious peak elimination, interrupt elimination and fine sliding averaging.

A first treatment may include the removal of spurious peaks in the raw signal. The Figures 4A and 4B illustrate the presence of a parasitic peak 41 in the raw signal 40 respectively before the raw signal and in the raw signal. Such parasitic peaks may be caused by a glare of the sensor 12 caused, for example, by abnormal reflection of the light beam. Reflective material envelopes, commonly used for promotional campaigns, can cause such glare. Paper dust deposited on the edge of envelopes can also cause such dazzling. A detection of the spurious peaks in the raw signal can be achieved by comparing a spurious peak spreading value with a threshold, for example the parasitic threshold and / or by comparing the amplitude value of the spurious peak with the average amplitude. of the raw signal. Indeed, the parasitic peaks generally have a spread much lower than the parasitic threshold and a very strong amplitude, greater than an average amplitude of the peaks. Typically, a paper dust is about 0.1 millimeter thick. Taking again the hypotheses of the previous example, a parasitic peak due to the presence of a dust generally has a thickness of 2 points.

A second treatment includes eliminating interruptions in the raw signal. The Figures 5A to 5D respectively illustrate interrupts 42 of the raw signal 40 positioned on a peak, in a trough, ascent of a peak and downward of a peak of the signal. Such interruptions may be caused by an absence of reflection due for example to a deterioration of the envelopes.

In a first embodiment of the elimination of interruptions illustrated by the Figures 6A and 6B the elimination of the interrupt 42 is carried out by reconstruction of the raw signal 40 on the basis of the derivative of the signal calculated between two clouds of N ₁ consecutive points separated by N ₂ points before and after the interruption. Typically, N1 can be chosen between 2 and 5 so as to limit the influence of disturbances in the calculation. N2 is usually chosen close to the parasitic threshold. This treatment makes it possible to quickly provide an approximation of the slope of the signal near the edges. This treatment can advantageously be implemented when the size of the interrupt 42 is greater than the parasitic threshold.

In a second embodiment of the elimination of interruptions illustrated in Figures 7A-7C the elimination of the interrupt 42 is performed by reconstructing the raw signal 40 on the basis of the derivative of the calculated point-to-point signal before and after the interruption. This processing makes it possible to precisely determine the slope of the signal at the edges of the interruption 42. This processing can advantageously be implemented when the size of the interrupt 42 is greater than the parasitic threshold.

In a third embodiment illustrated in Figures 8A and 8B , the elimination of the interruption 42 is performed by reconstructing the raw signal 40 based on the value of the signal before and after the interruption. This processing makes it possible to reconstitute the signal simply by a linear approximation of the signal. This treatment is advantageously implemented when the size of the interrupt 42 is less than the parasitic threshold.

A third treatment may include fine filtering of the gross signal by moving average. A mathematical formula for performing such filtering can be of the form: $$\widetilde{X_{\mathrm{not}}}=\frac{1}{\mathrm{NOT}}{\displaystyle \underset{p=0}{\overset{\mathrm{NOT}}{\Sigma }}}{X}_{\mathrm{not}-p}$$

The choice of the number N of points of the signal on which said sliding average is calculated may be performed on the basis of a threshold, for example the interference threshold determined during the analysis step S3 of the raw signal. In one embodiment, the number N can be chosen equal to half the parasitic threshold. This treatment makes it possible to ensure a low-pass filter of the raw signal and to free itself from micro-waves.

The Figures 9A and 9B respectively illustrate a raw signal 40 and a processed signal 50 in an embodiment in which the signal processing step S4 successively comprises the removal of spurious peaks, the reconstruction of the interrupts and the fine sliding average filtering.

In a fifth counting step S5, the processed signal 50 is analyzed to precisely determine the number of elements 31 of the stack 30.

In a first embodiment of the step S5 of counting, the determination of the number of elements 31 in the processed signal 50 is performed by serial analysis of the derivative of the processed signal calculated between two clouds of N ₁ consecutive points separated by N ₂ points. In this calculation, it is possible for example to choose N ₁ between 1 and 5 as a function of the compactness of the stack and N ₂ equal to the parasitic threshold. The more the elements of the pile are tight, the more N ₁ is generally chosen small. The analysis of the derivative includes counting the number of sign changes of the derivative.

In a second embodiment of the step S5 counting illustrated in FIG. figure 11 in addition to determining the number of elements on the basis of the serial analysis of the derivative of the processed signal, a detection of glued elements in the stack is performed. This makes it possible to improve the counting of the elements 31. In fact, elements of the stack may, for example, be damaged by a fall of an object on the edge 300 of the stack 30. The edges 310 of the damaged elements may then form a flat line almost continuous. The raw signal corresponding to such damaged peaks has an almost constant amplitude.

In a first step S11 of the detection of the glued elements, the spreading of the peaks in the processed signal 50 can be determined. For example, the spreading of the peaks can be deduced from an analysis of the sign changes of the derivative of the processed signal 50. Indeed, a peak is framed by two minima and the analysis of the derivative of the signal makes it possible to deduce the spreading peaks in the signal. A device for implementing the detection of glued elements may include a Schmitt flip-flop configured to switch when the derivative of the raw signal input to the flip-flop changes from sign. Thus, the flip-flop outputs a square-wave signal and the size of the slots corresponds to the spreading of the peaks of the processed signal.

In a step S12 of the detection of the glued elements, the spreading of the peaks is compared with a threshold to determine the peaks likely to contain glued elements. The threshold may be equal to the interference threshold or the average spreading value in the raw signal. For a given peak, if the spread of the peak is below the threshold, the control ends.

If the spread of the peak is greater than the threshold, a portion of raw signal corresponding to a processed signal portion comprising the determined peak is analyzed. This allows a finer analysis because the processed signal has undergone changes that may have removed small amplitude variations. Said portion of the raw signal may advantageously be finely filtered by sliding average filtering. This eliminates high frequency disturbances.

In a step S13 of the detection of the glued elements, the point-to-point derivative of the portion of the raw signal corresponding to the determined peak is calculated. In a step S14, the calculated derivative is analyzed. If a change of sign of the derivative of the raw signal which was not detected on the derivative of the processed signal is detected in the raw signal, then it can be considered that there are two elements glued to each other. other and can be incremented a counter in a step S16. The additional peak detected could be masked in the signal processed by statistical processing. The control ends after the step of incrementing. The figure 11A illustrates a raw signal 40 having a peak additional 402 of low amplitude.

If an additional peak is not detected, a determination of a second derivative of the portion of the raw signal corresponding to the processed signal portion comprising the determined peak is performed in a step S16. The second derivative can be obtained point-to-point from the derivative calculated in step S13.

If a peak of the second derivative is detected in said portion of the raw signal at a distance less than the mean value of spreading a peak of the signal then it can be considered that there are two elements glued one to the other and a counter can be incremented in a step S16. Otherwise, the control ends. The Figure 11B illustrates a derivative 43 and a second derivative 44 of a raw signal 40 (not shown). The second derivative 44 comprises a peak 441 corresponding to an inflection point 431 of the derivative. In one embodiment, a peak of the second derivative can be considered to represent glued folds if the peak is detected at a distance D less than about half of the spread E _a of the abnormal peak. Thus, the glued-ply capture window is equal to about 50% of the spreading of the abnormal spike determined and centered on the peak. This corresponds to a factor of four in dispersion, that is to say twice as big or twice as fine as an average element.

In a third embodiment of the counting step S5, determining the number of elements 31 in the processed signal 50 comprises the serial analysis of consecutive quadruplets of points of the processed signal. For a given quadruplet (X _p , X _{p + 1} , X _{p + 2} , X _{p + 3} ) the analysis can include the determination of a slope between first two points (X _p , X _{p + 1} ) of the quadruplet and a slope between the last two points (X _{p + 2} , X _{p + 3} ) of the quadruplet. A peak of the processed signal can be determined by a change of direction of the determined slopes. Complementarily, flat peaks can be identified by the identification of high and low half-peaks distant from each other. The high half-peaks and the low half-peaks can be respectively determined on a given quadruplet of points in the signal processed by the passage of a horizontal slope (respectively negative) for the first two points (X _p , X _{p + 1} ) a quadruplet at a positive slope (respectively horizontal) for the last two points (X _{p + 2} , X _{p + 3} ) of the quadruplet. This makes it possible to detect almost flat peaks of significant spreading.

In one embodiment of the step S5 of counting, the determination of the number of elements in the processed signal can be carried out successively according to the three embodiments presented above so as to obtain several counting results of the number of elements of the pile.

In one embodiment, the counting method can be performed several times successively to obtain several counting results of the number of elements of the stack. It is also possible to use a multibeam rangefinder or a plurality of rangefinders to obtain a plurality of counting results.

In embodiments in which a plurality of counting results are obtained, the counting step S5 may be followed by a convergence determination step S6 in which at least one disturbing measure can be eliminated. advantageously, the counting device operator may be notified if the counting results have a standard deviation greater than a predetermined tolerated value.

Although described through a certain number of exemplary embodiments, the counting method and the device according to the invention comprise various variants, modifications and improvements which will be obvious to those skilled in the art, it being understood that these various alternatives, modifications and improvements are within the scope of the invention as defined by the following claims.

Claims (15)

A method of counting stacked elements forming on a slice of the stack at least one level variation line formed by the edges of the stacked elements comprising the steps of: recording a raw signal corresponding to the measurement by telemetry of the variation of level on a line of points crossing the edge of the stack; - analysis of the raw signal to estimate one or more parameters relating to the structure of the stack; processing of the raw signal recorded by one or more statistical processes on the basis of the estimated parameters; analyzing the signal processed to determine the number of stacked elements. Counting method according to claim 1, wherein the estimated parameters comprise: an average thickness of the elements of the stack, and a dispersion factor of the thicknesses of the elements of the stack. The counting method of claim 2, wherein the raw signal processing includes determining a parasitic threshold proportional to the product of the dispersion factor and the average thickness. The counting method of claim 2, wherein the raw signal processing includes determining an estimated interference threshold based on an average thickness of the stack elements having a thickness less than a limiting thickness. The counting method according to at least one of claims 2 to 4, wherein the average thickness is determined as a function of a raw number of stacked elements estimated on the basis of the recorded raw signal. The counting method according to one of claims 3 to 5, wherein the raw signal processing includes the suppression of spurious peaks based on the interference threshold. The counting method according to at least one of claims 3 to 6, wherein the step of processing the recorded raw signal comprises reconstructing interrupts based on the slope of the signal before and after the interruption when the interruption is longer than the parasitic threshold. The counting method according to at least one of claims 3 to 7, wherein the step of processing the recorded raw signal comprises reconstructing interrupts based on the value of the signal before and after the interruption when the interruption is shorter than the interference threshold. Counting method according to one of claims 3 to 8, wherein the step of processing the recorded raw signal comprises the sliding average filtering of the signal, the number of points of the signal on which said sliding average is calculated being determined on the base of the parasitic threshold. Counting method according to at least one of the preceding claims, wherein the step of analyzing the processed signal comprises counting the number of sign changes of the derivative of the processed signal and / or counting the number of change of direction. of the point-to-point signal. The counting method according to at least one of the preceding claims, wherein the analyzing step comprises checking for the presence of pasted elements in the stack on the basis of the recorded raw signal. A counting method according to at least one of the preceding claims, wherein the step of recording comprises recording a plurality of raw signals corresponding to the level variation measure on one or more lines of points traversing the entire slice of the pile. Counting method according to claim 12, wherein the recording of the plurality of signals The roughs are performed using a plurality of rangefinders, a range finder comprising a plurality of beams and / or sweeping the wafer of the stack a plurality of times. The counting method according to at least one of claims 12 and 13, further comprising a step of verifying the convergence of the count results of the stacked elements provided by the plurality of raw signals. Device for counting stacked elements comprising: a telemetry measurement system adapted to measure a signal corresponding to the level variation of at least one line crossing a slice of the stack formed by said stacked elements; data storage means adapted to record said signal; calculation means adapted to implement the counting method according to at least one of claims 1 to 14.

Images