JP3693993B2 - Paper sheet confirmation device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a paper sheet determination method applied to, for example, an automatic depositing / dispensing machine (ATM), a vending machine, and the like. More specifically, the paper sheet is stacked and transported, skewed, folded, broken, and the like. The present invention relates to a paper sheet determining apparatus that can accurately determine the number of sheets and the type even if there is a problem.
[0002]
[Prior art]
In general, a transport path for transporting paper sheets such as banknotes, cash vouchers, paper, and cards has a configuration in which a pair of transport rollers and a transport member such as a transport belt are combined to sandwich and transport paper sheets one by one. are doing.
[0003]
As a device for detecting the number of sheets conveyed on the conveyance path, for example, there is an apparatus as described in Patent Document 1 of the prior application.
[0004]
In this apparatus, as shown in FIG. 15A, detection areas (three tracks) 153a, 153b, and 153c in which the conveyance width direction is divided into three are set in the conveyance process of the paper sheet 152 guided to the conveyance path 151. are doing. Then, paper sheets are conveyed based on detection signals from the first to third thickness sensors 154a, 154b, and 154c arranged corresponding to the divided tracks and data from an image scanner (not shown). Detects and checks the status and the number of sheets being transported.
[0005]
However, as a detection method of this type, as shown in FIG. 15B, the thickness amount per scan is calculated from the sensor data amount (integrated value) of each thickness sensor 154a... Is calculated. And the number of sheets is estimated from the sensor data amount of each sensor 154a ....
[0006]
In particular, all the sensor data of the individual detection areas of the three tracks 153a, 153b, and 153c are set as the number determination condition. Therefore, it is recognized that one paper sheet 152 has been conveyed only when the results of the numbers of all tracks match.
[0007]
However, as shown in FIG. 16A, when the paper sheet 152 is greatly skewed on the conveyance path 151, the large skewed paper sheet 152 passes only through the first and second tracks 153a and 153b. In some cases, the third track 153c may not be passed.
[0008]
In this case, the third thickness sensor detects the position of the third track 153c even if the first and second thickness sensors 154a and 154b detect the positions of the first and second tracks 153a and 153b. 154c was not detected, and as a result, even though one sheet passed, the number of sheets could not be determined and there was a risk of erroneous counting.
[0009]
Further, as shown in FIG. 16B, there is a case where the number of sheets 152 can be determined by 2/3 tracks (for example, the first track 153a and the second track 153b). In this case, as indicated by an imaginary line in the figure, it is conceivable that the paper sheet 152a passes only the remaining 1/3 track (the third track 153c). For this reason, there is a possibility of erroneously recognizing the number of sheets, and as a result, there is a risk of erroneous counting.
[0010]
In addition, as shown in FIG. 17, even when the paper sheet 152 is skewed and passes through the entire area of the three tracks 153a to 153c, the edge of the paper sheet is difficult to be applied to the end track at the time of a large skew, There is a difference in the sensor data amount 155a, 155b, 155c for each end track. Therefore, one is detected from the sensor data amounts 155a and 155c at both ends, and two is detected from the sensor data amount 155b at the center. As a result, the number of sheets cannot be determined, causing erroneous detection.
[0011]
Furthermore, as another example, as already described in Patent Document 2, for example, there is known an apparatus that includes a thickness detection mechanism and an image detection mechanism in a conveyance path to determine a conveyance medium. According to this apparatus, the type and number of paper sheets can be determined even when skewing during transportation is severe or when a plurality of sheets are overlapped.
[0012]
That is, a pattern pattern and an outer shape of a paper sheet are read by an image detection mechanism such as an image reader, and the presence or absence of superimposition is checked from the output of the thickness detection mechanism. Thereafter, the figure of the paper sheet is extracted from the read outer shape, and the size and the number of the paper sheets are obtained.
[0013]
In this case, regarding the thickness detection, it is only checked whether or not it is in an overlapped state, and the outer shape is read by the image detection mechanism. In addition, regarding image detection, it is impossible to discriminate when the outline information of the transport medium cannot be obtained with image data. For example, when discriminating banknotes, if the corner portion of the outer shape is not clear, the corner portion is used as a reference, making it difficult to confirm the determination, making it impossible to determine the number or denomination.
[0014]
Specifically, when there is a banknote 181 that is completely hidden as shown in FIG. 18, the number and denomination cannot be determined. Further, as another example, a similar device as already described in Patent Document 3 is known, but this also only estimates how many bills overlap each other by thickness detection. The external shape was read by a one-dimensional optical line sensor.
[0015]
Therefore, when there is a banknote 181 that is completely hidden as shown in FIG. 18 where the outer corner portion of the transport medium is not clear, the number and denomination cannot be determined in the same manner. Furthermore, there is also a disclosure about folding detection that the folded banknote is present when the entire banknote image is overlapped with the entire transport form image and the overlapping state of the banknotes is analyzed and does not coincide with the overall image. There is no further disclosure.
[0016]
[Patent Document 1]
JP 2001-266105 A
[Patent Document 2]
JP-A-5-46842
[Patent Document 3]
Japanese Patent Laid-Open No. 7-141547.
[0017]
[Problems to be solved by the invention]
Therefore, the present invention accurately determines the number of sheets and the type even if paper sheets in various states due to troublesome conveyance states such as overlap conveyance and skew conveyance and the paper sheets themselves are broken, torn, or missing are conveyed. An object of the present invention is to provide an apparatus for determining paper sheets.
[0018]
[Means for Solving the Problems]
The present invention provides a thickness sensor in which a plurality of thickness detection elements for detecting the thickness of a paper sheet passing through a conveyance path are arranged over the entire width of the conveyance path, and detection detected for each thickness detection element. Contour extraction means for extracting contour data representing the outer shape of the paper sheet passing through the conveyance path from the waveform, storage means for storing basic contour data of the paper sheet serving as a reference, and extracted by the contour extraction means A paper sheet determination apparatus comprising: comparison means for comparing contour data with contour basic data stored in the storage means; and determination means for determining at least the number of paper sheets from the comparison result of the comparison means. It is characterized by being.
[0019]
Here, the paper sheet is a general term for banknotes, cash vouchers, paper, cards, and the like.
[0020]
As a result, it has been transported by comparing the contour data representing the outer shape of the paper sheet obtained from the detected waveform actually detected over the entire length in the transport width direction with the basic contour data stored in the storage means. It is possible to determine the number of sheets that pass through whether the sheet is conveyed one by one or stacked.
[0021]
In particular, the thickness sensor can subdivide and detect the thickness data of the entire width by a plurality of thickness detection elements arranged over the entire width of the paper sheet transport path in the transport width direction. For this reason, the whole area of the conveyance width direction can be detected with high accuracy. And the outline shape of the passing paper sheet can be accurately grasped by the contour data extracted from these detected waveforms.
[0022]
Further, the thickness distribution state and the conveyance state of the paper sheet can be clearly understood from both the thickness data and the contour data, and the number of passing sheets can be determined with high accuracy. For this reason, even if it is a case where the thickness for two sheets is detected, it can be discriminate | determined whether the part is a thing by a folding or an overlap conveyance.
[0023]
In addition, even if the corner portion of the paper sheet extracted by the contour extracting means is not clear, the three-dimensional data of the paper sheet taking thickness data into account can be obtained, so that the overlapping state is clarified. The number of sheets can be accurately recognized.
[0024]
Therefore, even if a paper sheet is transported in a skewed manner, or a folded or missing paper sheet is transported, it can be reliably determined whether the sheet is transported or stacked. For this reason, there is no risk of miscounting the number of sheets, and count management with high reliability can be performed.
[0025]
In yet another invention, a thickness sensor in which a plurality of thickness detection elements for detecting the thickness of a paper sheet passing through the conveyance path are arranged over the entire width of the conveyance path, and detection for each thickness detection element. Extracted from the detected waveform, contour extracting means for extracting contour data representing the outer shape of the paper sheet passing through the conveying path, storage means for storing basic contour data of the paper sheet serving as a reference, and extraction from the contour extracting means A paper sheet comprising: a comparison unit that compares the contour data that is stored with the basic contour data stored in the storage unit; and a determination unit that determines at least the type of the paper sheet from the comparison result of the comparison unit It is a definite device.
[0026]
According to this, even if the conveyed paper sheets are in a troublesome conveying state such as overlap conveyance or skew conveyance, the type of the paper sheets can be accurately determined from the acquired data.
[0027]
In another invention, a thickness sensor in which a plurality of thickness detection elements for detecting the thickness of the paper sheet passing through the conveyance path are arranged over the entire width of the conveyance path, and detection for each thickness detection element. Contour extraction means for extracting contour data representing the outer shape of the paper sheet passing through the transport path from the detected waveform, and an image of the outer shape data of the paper sheet in the transport process of the transport path. Image data acquisition means acquired by the above, tolerance value calculation means for calculating an allowance value from the outline data and the outline data, storage means for storing basic outline data of a paper sheet as a reference, and the outline data A paper sheet comprising: a comparison unit that compares the contour basic data based on an allowable value calculated by the allowable value calculation unit; and a determination unit that determines at least the number of sheets from the comparison result of the comparison unit. This is a classifying device The features.
[0028]
Here, the allowable value refers to the basic outline data and basic outline data of the reference paper sheet when the outline data and the basic outline data of the reference paper sheet are compared to determine the number of paper sheets. If there is contour data between the data to which the allowable value is added, it is a value used as the value of the allowable range that can be determined to match the basic contour data.
[0029]
According to this, since the number of sheets is determined in comparison with the outline basic data based on the allowable value calculated from the outline data acquired by the image data acquisition means, the number of sheets can be determined more accurately. .
[0030]
In yet another invention, a thickness sensor in which a plurality of thickness detection elements for detecting the thickness of a paper sheet passing through the conveyance path are arranged over the entire width of the conveyance path, and detection for each thickness detection element. Contour extracting means for extracting contour data representing the outer shape of the paper sheet passing through the transport path from the detected waveform, storage means for storing basic contour data of the paper sheet as a reference, and the contour data A fold detection means for detecting a fold of the paper sheet, and when the fold detection means detects that there is a fold, the contour basic data A paper sheet determining apparatus comprising: a comparing unit that compares the number of sheets; and a determining unit that determines at least the number of sheets from the comparison result of the comparing unit.
[0031]
Here, expanding a fold means creating a data by folding back a portion detected as having a fold with a line-symmetric figure.
[0032]
According to this, even in a folded paper sheet, the original state without folding is found from the developed data. For this reason, it is possible to reliably determine the number of sheets.
[0033]
DETAILED DESCRIPTION OF THE INVENTION
An embodiment of the present invention will be described below with reference to the drawings.
FIG. 1 shows a banknote processing apparatus 11 built in an ATM. This banknote processing apparatus 11 is provided with a banknote transport group 13 for transporting banknotes on the upper side of the apparatus main body 12, and is allowed to store and feed banknotes on the lower side. The bill storage group 14 is provided, and the upper and lower sides of the apparatus body 12 are separated into a transport system and a storage system.
[0034]
The banknote transport group 13 on the upper side has a deposit / withdrawal port 15 at the top of the apparatus main body 12, and the bills inserted into the deposit / withdrawal port 15 are guided to the identification unit 16 disposed at the upper middle position, After identifying the fake, denomination, number of sheets, and front and back, it is led to the temporary holding unit 17 and temporarily held. When it is identified that the captured banknote is in the state of being conveyed on the back side, it is led to the front / back reversing unit 18 and the front and back sides are aligned, and then guided to the temporary holding unit 17. After being led to 19 and temporarily suspended, it is returned to the original deposit / withdrawal port 15. And the optimal storage timing in operation is selected from the temporary storage unit 17 and stored in the banknote storage group 14 on the lower side.
[0035]
On the other hand, the lower bill storage group 14 includes first to third stackers S1 to S3 that are fixedly installed, and an operation cartridge C1 and a collection cartridge C2 that are allowed to be attached and detached. Among these, the first to third stackers S1 to S3 have a bill collecting function and a payout function, and store three denominations of 10,000 yen, 1,000 yen, and 5,000 yen in denominations, and any stacker at the time of withdrawal. Each sheet is fed out from S1 to S3 and guided to the identification unit 16 to check the denomination and then conveyed to the deposit / withdrawal port 15.
[0036]
Further, the operation cartridge C1 distributes and stores banknotes in the stackers S1 to S3 at the start of business or banknote replenishment, and collects banknotes from the stackers S1 to S3 at the end of business or when the bank is full. The collection cartridge C2 collects reject banknotes such as poorly identified banknotes and forgotten banknotes generated during deposit / withdrawal transactions and replenishment processing.
[0037]
By the way, even if a banknote is conveyed in the above-mentioned identification part 16 in the above-mentioned conveyance state, it has the banknote determination apparatus 21 mentioned later which determines what is the number of passages at that time and what denomination is. ing.
[0038]
FIG. 2 shows the above-described banknote determination apparatus 21. The banknote determination apparatus 21 has, for example, a one-dimensional CCD image sensor 23 and a thickness on the identification conveyance path 22 of the identification section 16 as shown in FIG. A sensor 24 is arranged so that the bill A that has been conveyed is read by two types of detection means.
[0039]
As shown in FIG. 2 (B), the above-described identification transport path 22 drives transport rollers R1 and R2 that are vertically opposed and are installed in the transport width direction. A is nipped and conveyed one sheet at a time. At this time, the transport rollers R1 and R2 are driven to rotate at a constant speed by receiving a rotational force from a transport motor (not shown).
[0040]
The CCD image sensor 23 is disposed over the entire width in the transport width direction along the above-described transport rollers R1 and R2, and acquires the outline data of banknotes passing through the image by an image.
[0041]
The thickness sensor 24 arranges a plurality of thickness detection elements 25 in a single row in a dense manner without gaps across the entire width in the conveyance width direction of the identification conveyance path 22 so as to increase the resolution. Similarly, a sensor-corresponding roller R3 is provided on the lower surface side over the entire width in the conveyance width direction. When the bill A is guided from the horizontal direction between the upper and lower opposing surfaces of the upper surface side thickness detection element 25 and the lower surface side sensor-corresponding roller R3, the thickness of the bill A is increased. Detect operation.
[0042]
3 shows the internal structure of one thickness detection element 25 constituting the thickness sensor 24, FIG. 3 (A) is a front sectional view of the thickness detection element 25, and FIG. 3 (B) is the thickness. A sectional view of the side surface of the detection element 25 is shown.
[0043]
Each thickness detecting element 25 has a T-shaped movable metal piece 33 attached to the lower portion of the socket body 31 by being biased downward by a coil spring 32 and biased from the lower surface of the socket body 31, and magnetically on the upper portion. An output terminal 35 having a built-in coil 34 protrudes from the upper surface of the socket body 31, and the output terminal 35 is connected to the control unit.
[0044]
The above-described thickness detection element 25 has a proximity sensor detection configuration. The detection operation is such that the movable metal piece 33 moves up and down due to the vertical expansion and contraction action of the coil spring 32, and the bill thickness is adjusted when the bill A passes. The movable metal piece 33 is pushed upward by the amount. The amount of mechanical displacement when displaced by being pushed up is converted into a voltage change by the magnetic coil 34 disposed on the upper portion of the socket body 31 (by a change in the distance between the movable metal piece 33 and the magnetic coil 34). It is output from the output terminal 35 as an electrical thickness detection signal.
[0045]
FIG. 4 is an enlarged view of the thickness sensor 24 in a linear array in which a plurality of thickness detection elements 25 are closely attached at regular intervals to an elongated socket body 31 arranged in the conveyance width direction. FIG. 4A is a plan view showing the arrangement state of the output terminals 35, and FIG. 4B is a side view showing the arrangement state of the movable metal pieces 33.
[0046]
When the above-described thickness sensor 24 is used, the thickness data of the entire width can be subdivided and detected by the plurality of thickness detection elements 25 arranged over the entire width of the identification transport path 22 in the transport width direction. For this reason, the whole area of the conveyance width direction can be detected with high accuracy, and the entire thickness distribution state of the bill A passing through the position of the thickness sensor 24 can be accurately captured. Thereby, the outline of the three-dimensional banknote described later can be extracted.
[0047]
FIG. 5 shows a block diagram of the control circuit of the banknote determination device 21. The control circuit includes an outline extraction unit 51 connected to the thickness sensor 24, an allowable value calculation unit 52 connected to the CCD image sensor 23, and a volume calculation. The unit 53, the storage unit 54, the contour comparison unit 55, and the determination unit 56 are configured.
[0048]
In the above-described contour extraction unit 51, thickness data subdivided in the transport width direction of the bill A detected for each thickness detection element 25 constituting the thickness sensor 24 is obtained, and from this subdivided thickness data Contour data representing the outer shape of the banknote is extracted.
[0049]
FIG. 6 shows an example of a detection waveform detected for each thickness detection element 25, and the detection waveform is output as time variation of the output voltage corresponding to the thickness of the banknote from each thickness detection element 25 along with the conveyance of the banknote. Is done.
[0050]
For example, as shown in FIG. 6 (A), when two bills A1, A2 of the same size are stacked and transported in an obliquely shifted state (continuous delivery), each thickness detection element of the thickness sensor 24 As for the output waveform detected from 25, the 1 sheet detection waveform W1 and the 2 sheet detection waveform W2 are obtained in the conveyance width direction. It can be accurately extracted from these waveform data that the overlap conveyance has occurred and that the overlap state is shifted obliquely.
[0051]
Similarly, as shown in FIG. 6B, when two bills A3 and A4 of different sizes are overlapped and conveyed as if they were one sheet, each thickness detection element 25 of the thickness sensor 24 As for the detected output waveform, one sheet detection waveform W3 and two sheet detection waveform W4 are obtained in the conveyance width direction. Also in this case, it can be accurately extracted from the waveform data that the overlap conveyance has occurred and that two sheets of different sizes are in an overlapping state.
[0052]
FIG. 7 shows an example of the outline extraction of banknotes, and the outline extraction processing of banknotes when the two banknotes A1 and A2 of the same size described in FIG. .
[0053]
When waveform data is obtained from each thickness detection element 25 when it is overlapped and conveyed, the waveform data is combined from the waveform data as shown in FIG. The three-dimensional figure 71 is generated by complementing the data with a straight line. The vertices (corner portions) of the three-dimensional figure 71 are connected by straight lines, and as shown in FIG.
[0054]
FIG. 8 shows an example of the outline extraction of banknotes in other cases. The banknotes in the case where two banknotes A3 and A4 having different sizes described in FIG. The contour extraction process will be described.
[0055]
When the waveform data when the superimposed transport is performed is obtained, the waveform data from each thickness detecting element 25 is combined from the waveform data as shown in FIG. The three-dimensional figure 81 is generated by complementing the data with a straight line. The contours of the bills that are overlapped and conveyed by connecting the vertices of the solid figure 81 with straight lines are extracted.
[0056]
The above-described waveform data complementing process supplements the data between the detection elements by linear interpolation when the resolution of the thickness sensor itself (which is proportional to the arrangement interval of the plurality of thickness detection elements 25) is about several millimeters. Find a three-dimensional outline with high accuracy. In this case, the closer the thickness detection element 25 is, the higher the resolution is. If finely set, the necessity for complementation is eliminated.
[0057]
On the other hand, in the allowable value calculation unit 52, when there is contour data between the contour basic data and the data obtained by adding the allowable value to the contour basic data when the contour data is compared with the basic contour data in the storage unit 54, An allowable value is calculated as an allowable range value that can be determined to match the contour basic data.
[0058]
Here, the allowable value is calculated from the difference between the outline data of the banknote image acquired from the CCD image sensor 23 and the above-described contour data.
[0059]
In this way, even if the resolution of the thickness sensor itself is low, the allowable value is calculated from the difference from the outline data of the image with a high resolution, and is compared with the contour basic data based on the allowable value. Therefore, the number of coins and the denomination can be determined with higher accuracy.
[0060]
FIG. 9 shows an example of image data acquisition from the CCD image sensor 23. When two bills A1 and A2 of the same size are stacked and conveyed obliquely, an image of the bill on which the CCD image sensor 23 overlaps is shown. Data 91 is acquired, and binarization processing of the image data 91 is performed to determine the outer shape of the banknote. In this determination, the outline of the banknote can be determined by finding the square corner of the banknote from the image data 91.
[0061]
The allowable value is calculated by calculating the difference between the image data (outer shape data) and the contour data described above by the allowable value calculation unit 52. For example, since the image data of the CCD image sensor has a better resolution than the thickness sensor, both the image data and the contour data are collated.
[0062]
When the difference (thickness sensor data-image data) between the outer shape data from the image data of the CCD image sensor and the contour data from the thickness sensor is calculated, if there is a difference of, for example, 2 mm in the bill longitudinal direction, the contour data And the basic contour data in the storage unit 54 are compared by adding a tolerance of 2 mm to the bill longitudinal direction of the basic contour data.
[0063]
Specifically, when the bill longitudinal direction is extracted as 121 mm from the contour data and the bill longitudinal direction is extracted as 119 mm from the image data, the allowable value is calculated to be 121-119 = 2 mm. For this reason, at the time of comparison with the contour basic data (120 mm), an allowable value of 2 mm is added, and if it is within the range of 120 mm to 122 mm, it is determined that the denomination matches the contour basic data.
[0064]
Further, the volume calculation unit 53 compares the volume obtained by calculating the three-dimensional data of the conveyed banknote obtained from the thickness data and the contour data with the volume basic data stored in the storage unit 54. It is possible to accurately recognize the number of bills and the denomination.
[0065]
The storage unit 54 stores basic contour data and basic volume data serving as a reference for vertical and horizontal lengths, volumes, and the like, which are determined for each denominated banknote.
[0066]
For example, if there are two denominations to be transported, if the thickness of the A denomination to be transported horizontally is 0.1 mm and the length and width are 70 mm x 120 mm, the volume is 840 mm. ³ It becomes. Also, if the thickness of B type is 0.1mm and the length and width are 60mm x 110mm, the volume is 660mm. ³ It becomes. Both of these are stored for determination criteria.
[0067]
The contour comparison unit 55 compares the actually detected contour data with the basic contour data stored in the storage unit 54 to perform bill discrimination processing.
[0068]
FIG. 10 shows an example of banknote discrimination processing. As shown in FIGS. 10A and 10B, the contour data of the banknote is extracted from the detected waveform acquired from the thickness sensor 24. The extracted contour data is compared with the basic contour data stored in the storage unit 54.
[0069]
As a result, as shown in FIG. 10C, two A denominations stored in the storage unit 54 and different-sized squares (50 mm × 100 mm) not stored in the storage unit 54 are obtained. Therefore, it is possible to determine that a square having a different size is invalid data.
[0070]
FIG. 11 shows another example of banknote discrimination processing. As shown in FIGS. 11 (A) and 11 (B), banknote contour data is extracted from the detected waveform acquired from the thickness sensor 24. The extracted contour data is compared with the basic contour data stored in the storage unit 54.
[0071]
As a result, as shown in FIG. 11 (C), squares of the same size of the A denomination and the B denomination stored in the storage unit 54 are obtained. Thereby, it can be determined that two different denominations are being conveyed in an overlapping manner.
[0072]
The confirmation unit 56 collates the volume data of the banknote calculated by the volume calculation unit 53 with the comparison result compared by the contour comparison unit 55. At this time, if it is within the error range, the number and denomination of the banknote are determined.
[0073]
For example, as described above with reference to FIG. 10, when determining the number of coins and denominations when the banknotes of the same size are generated, the data obtained from the two A denominations and the volume calculated by the volume calculation unit 53 are used. Is 840mm ³ It is determined whether it is within the range of × 2 ± volume tolerance. At this time, if the value is within the error range (within the allowable range), it can be determined that the two denominations of the A denomination are consecutive.
[0074]
Similarly, when determining the number and denomination when the banknotes of different sizes described above in FIG. 11 are generated, the data obtained for the A denomination and the B denomination and the volume calculated by the volume calculation unit 53 are 840 mm. ³ + 660mm ³ It is judged whether it is within the range of ± volume tolerance. At this time, if it is a value within an error range, it can be determined that two banknotes of the A denomination and the B denomination are continuously issued.
[0075]
Further, when determining, it is possible to predict a certain number of sheets even with either volume or contour data, but the determination accuracy is further increased by using two data. Furthermore, even when the skew becomes large, contour data representing the outer shape of the banknote can be extracted from the thickness data of the thickness sensors arranged over the entire width of the transport path, so the number of coins and the denomination of the coins can be determined accurately. Is possible.
[0076]
FIG. 12 shows an example of detection when a folded banknote is conveyed. As shown in FIG. 12A, when a banknote in a folded state is conveyed, the image of the banknote is detected by the CCD image sensor 23. Read.
[0077]
In the read CCD image sensor 23, as shown in FIG. 12B, non-rectangular image data 121 having a corner shape is obtained.
[0078]
Further, in the thickness sensor 24, as shown in FIG. 12C, contour data 122 that is not a quadrangle having a corner shape is obtained.
[0079]
Then, when the contour extraction unit 51 extracts the contour data 122 having a corner break, as shown in FIG. 12D, the corner break portion a of the contour data is developed along the contour. Collation can be confirmed by comparing the developed data and the basic contour data by the contour comparison unit 55.
[0080]
In the contour comparison unit 55, the contour data 122 that is not a quadrangle cannot be checked against the basic contour data in the storage unit 54, and since the thickness of the triangular portion at the upper right corner in the figure is doubled, The occurrence can be detected. In this case, expansion data is created by performing expansion as described later. As a result, the determination unit 56 can accurately determine the number and denomination even if the folded banknote is conveyed.
[0081]
FIG. 13 shows the above-described algorithm for unfolding a folded banknote. The contour data of the folded banknote is displayed in graphic form, and each vertex number at that time and the number of branches connected to it are obtained.
[0082]
In this case, since there are six vertices (corner portions) of the folded banknote, the numbers 1 to 6 are attached to the vertices, and the matrix display (vertex number, number of branches),
(1,2)
(2, 2)
(3, 3): Can be folded back with 3 branches
(4,2)
(5,3): Can be folded back with 3 branches
(6,2)
It becomes.
[0083]
As a result, since folding can be performed at the position of the branch number 3 (in FIG. 13, folding can be performed at the positions of the vertex numbers 3 and 5), each triangle of the upper right triangle has a thickness of 2 from the above thickness data. In the case of development from the fact that it is a side and contour data, the folding position of the corner-folded portion is determined by applying a condition for the contour to be a quadrangle.
[0084]
When there are no two vertices corresponding to the number of branches 3 described above, or when there are four or more vertices, it is considered that there is an overlap in banknotes.
[0085]
The processing operation in the case of determining the number of banknotes and the denomination using the banknote determination device 21 configured as described above will be described with reference to the flowchart of FIG.
When a bill passes through the identification transport path 22, the thickness sensor 24 detects the thickness of the bill that passes through it, and the image of the bill is acquired by the CCD image sensor 23 (step n1).
[0086]
Contour data representing the outer shape of the banknote is extracted by the contour extraction unit 51 from the detection waveform detected for each thickness detection element 25 of the thickness sensor 24 (step n2).
[0087]
In addition, the allowable value calculation unit 52 calculates the allowable value from the difference between the contour data obtained from the image of the banknote acquired by the CCD image sensor 23 and the contour data extracted by the contour extraction unit 51 (step n3). .
[0088]
The contour basic data incorporating the calculated allowable value and the contour data are compared by the contour comparison unit 55 (step n4).
[0089]
At this time, the contour comparison unit 55 determines whether or not the banknote is folded from the comparison result (step n5),
If it is determined that there is no banknote breakage, the denomination and the number of sheets are estimated (step n6).
[0090]
Further, the volume calculation unit 53 obtains the volume from the thickness data and the contour data (step n7),
From the comparison result obtained by comparing the obtained volume with the basic volume data stored in the storage unit 54, the determination unit 56 determines whether the volume is within an allowable range (step n8).
[0091]
At this time, if it is within the allowable range, it is determined that the bill is determinable, and its denomination and number are determined (step n9).
[0092]
However, if it is determined that the value is outside the allowable range in step n8, the rejection process is performed because it cannot be determined.
[0093]
Even when it is determined in step n5 that there is a fold, if it is determined that the folded portion cannot be expanded, a rejection process is performed.
[0094]
By the way, even if it is determined that there is a fold in the above step n5, the data can be expanded by folding the folded portion with a line-symmetric figure (step n10).
The development data which developed the folding of the banknote is created, converted into a regular banknote without folding, and processed. Thereafter, the processes after step n6 described above are performed, and the denomination and number of banknotes are determined (step n11).
[0095]
As described above, since the thickness sensor 24 subdivides the entire width in the conveyance width direction of the identification conveyance path 22 with the plurality of thickness detection elements 25 and detects the thickness data of the entire width, the entire area in the conveyance width direction is accurately detected. Can be detected.
[0096]
And the three-dimensional outline of a banknote can be correctly grasped | ascertained by extracting the outline data showing the external shape of a banknote from this thickness data. For this reason, the number of banknotes and the denomination can be accurately determined even without image data.
[0097]
Furthermore, even when the banknotes are completely hidden as shown in FIG. 18, the number of banknotes and the denomination can be accurately determined from the three-dimensional data of the banknotes.
[0098]
As described above, if a banknote determination device is incorporated into an ATM identification unit that handles banknotes, this banknote determination device is arranged in the full width by a plurality of thickness detection elements arranged over the entire width in the conveyance width direction of the identification conveyance path. Since the thickness data is subdivided and detected, the entire area in the conveyance width direction can be detected with high accuracy, and a bill can be accurately captured regardless of skew conveyance. For this reason, there is no risk of misjudging that it has not passed, and count management with high reliability can be performed.
[0099]
Further, even without image data, the number of banknotes and denominations can be accurately determined, and even when there is a banknote that is completely hidden as shown in FIG. Since the three-dimensional data of the banknote based on the data is obtained, the number of banknotes and the denomination can be determined accurately.
[0100]
In correspondence between the configuration of the present invention and the configuration of the above-described embodiment,
The paper sheet determination device of this invention corresponds to the banknote determination device 21 of the embodiment,
Similarly,
The conveyance path corresponds to the identification conveyance path 22,
Paper sheets correspond to banknotes A, A1, A2, A3, A4,
The contour extracting means corresponds to the contour extracting unit 51,
The storage means corresponds to the storage unit 54,
The comparison unit and the break detection unit correspond to the contour comparison unit 55,
The confirmation means corresponds to the confirmation unit 56,
The image data acquisition means corresponds to the CCD image sensor 23,
The allowable value calculation means corresponds to the allowable value calculation unit 52, but the present invention can be applied based on the technical idea shown in the claims and is not limited to the configuration of the above-described embodiment. Absent.
[0101]
【The invention's effect】
According to the present invention, the three-dimensional outline of the paper sheet is obtained from the detection waveform of the thickness sensor, thereby determining the number of paper sheets conveyed and the type of the paper sheet with high accuracy and high reliability. be able to.
[Brief description of the drawings]
FIG. 1 is an internal configuration diagram of a banknote handling apparatus built in an ATM.
FIG. 2 is a schematic configuration diagram showing a banknote confirmation device.
FIG. 3 is a longitudinal sectional view showing an internal structure of a thickness detection element.
FIG. 4 is an external view showing a thickness sensor.
FIG. 5 is a control circuit block diagram of the banknote determination device.
FIG. 6 is an explanatory diagram showing an example of contour data extraction.
FIG. 7 is an explanatory view showing a three-dimensional contour extraction example of a banknote.
FIG. 8 is an explanatory view showing a three-dimensional contour extraction example of another banknote.
FIG. 9 is an explanatory diagram showing an example of obtaining image data from a CCD image sensor.
FIG. 10 is an explanatory diagram showing an example of bill discrimination processing.
FIG. 11 is a descriptive diagram illustrating another example of banknote discrimination processing.
FIG. 12 is an explanatory diagram showing an example of a discrimination process for folded banknotes.
FIG. 13 is an explanatory diagram showing a folded banknote unfolding algorithm.
FIG. 14 is a flowchart showing the processing operation of the banknote confirmation device.
FIG. 15 is a conveyance explanatory view showing a conventional paper sheet conveyance detection structure.
FIG. 16 is a conveyance explanatory view showing an example of conveyance detection during a large skew of a conventional paper sheet.
FIG. 17 is a conveyance explanatory view showing an example of erroneous counting of a conventional paper sheet.
FIG. 18 is a transport explanatory view showing a transport state in which a conventional banknote is completely hidden.
[Explanation of symbols]
16: Identification part
21 ... Bill confirmation device
23 ... CCD image sensor
24 ... Thickness sensor
51. Outline extraction unit
52 ... Allowable value calculation unit
54. Storage unit
55. Outline comparison unit
56 ... Determining part
A, A1, A2, A3, A4 ... bills

Claims (7)

A thickness sensor in which a plurality of thickness detection elements for detecting the thickness of the paper sheet passing through the transport path are arranged over the entire width of the transport path;
Contour extracting means for extracting contour data representing the outer shape of the paper sheet passing through the conveyance path from the detected waveform detected for each thickness detecting element;
Storage means for storing basic outline data of paper sheets as a reference;
Comparing means for comparing the contour data extracted by the contour extracting means with the basic contour data stored in the storage means;
A confirmation means for confirming at least the number of sheets from the comparison result of the comparison means;
A paper sheet confirmation device. A thickness sensor in which a plurality of thickness detection elements for detecting the thickness of the paper sheet passing through the transport path are arranged over the entire width of the transport path;
Contour extracting means for extracting contour data representing the outer shape of the paper sheet passing through the conveyance path from the detected waveform detected for each thickness detecting element;
Storage means for storing basic outline data of paper sheets as a reference;
Comparing means for comparing the contour data extracted by the contour extracting means with the basic contour data stored in the storage means;
Confirming means for confirming at least the type of the paper sheet from the comparison result of the comparing means;
A paper sheet confirmation device. A thickness sensor in which a plurality of thickness detection elements for detecting the thickness of the paper sheet passing through the transport path are arranged over the entire width of the transport path;
Contour extracting means for extracting contour data representing the outer shape of the paper sheet passing through the conveyance path from the detected waveform detected for each thickness detecting element;
An image data acquisition unit that is opposed to the conveyance path and acquires the outline data of the paper sheet by an image in the conveyance process of the conveyance path;
An allowable value calculating means for calculating an allowable value from the outline data and the outline data;
Storage means for storing basic outline data of paper sheets as a reference;
Comparing means for comparing the contour data and the basic contour data based on the tolerance calculated by the tolerance calculation means;
A confirmation means for confirming at least the number of sheets from the comparison result of the comparison means;
A paper sheet confirmation device. A thickness sensor in which a plurality of thickness detection elements for detecting the thickness of the paper sheet passing through the transport path are arranged over the entire width of the transport path;
Contour extracting means for extracting contour data representing the outer shape of the paper sheet passing through the conveyance path from the detected waveform detected for each thickness detecting element;
Storage means for storing basic outline data of paper sheets as a reference;
A fold detection means for detecting a fold of a paper sheet conveyed from the contour data;
When the fold detection means detects that there is a fold, a comparison means for creating development data obtained by developing the detected fold from the outline data and comparing the outline basic data;
A confirmation means for confirming at least the number of sheets from the comparison result of the comparison means;
A paper sheet confirmation device. The determining means compares the paper size data in the contour data by the comparing means with the paper size data in the contour basic data stored in the storage means, and the comparison result is consistent. 5. The paper sheet determining apparatus according to claim 1, wherein the determined number is determined as the number of paper sheets. The determining means compares the size data of the paper sheet in the contour data by the comparing means with the size data for each type of paper sheet in the basic contour data stored in the storage means, and compares 3. The paper sheet determining apparatus according to claim 2, wherein a type that matches the results is determined as a paper sheet type. 5. The paper sheet determination apparatus according to claim 4, wherein the fold detection unit detects that there is a fold when there is non-rectangular data in the contour data.

Images