JP2014186649A - Paper discrimination device, and paper discrimination method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a paper discrimination device and a paper discrimination method capable of improving discrimination accuracy.SOLUTION: A bill discrimination device 1A is equipped with a scanning input portion 10, and a discriminating portion 20. The scanning input portion 10 has a carrying path for carrying a bill, and an image line sensor for imaging the carried bill. The discriminating portion 20 compares an obtained image D1 outputted from the image line sensor with reference images D6-D8 to discriminate the bill. The discriminating portion 20 cuts out a discrimination object image D2 which is a part corresponding to the bill from the obtained image D1, and expands or shrinks the discrimination object image D2 so as to make the size of the discrimination object image D2 come close to the size of the reference images D6-D8, and then compares the discrimination image D2 (a collation image D5) after expansion or shrinkage with the reference images D6-D8.

Description

  The present invention relates to a paper sheet discrimination apparatus and a paper sheet discrimination method.

  Patent Document 1 describes a technique related to an apparatus for discriminating the type and authenticity of banknotes, checks and other paper sheets. In this apparatus, the outer dimension of the paper sheet is calculated from the read data pattern for the paper sheet, and the type of paper sheet (denomination for banknotes) is determined based on the outer diameter.

  Patent Document 2 describes a technique related to an apparatus for discriminating the authenticity of paper sheets. In this apparatus, pattern matching is performed not on a pixel basis but on an appropriate discrimination area basis in order to alleviate the influence of a positional deviation during the conveyance of paper sheets.

JP-A-10-198837 JP-A-10-21442

  As a paper sheet discrimination apparatus, there is an apparatus for identifying a paper sheet or determining whether it is true or false by matching an image obtained by an image line sensor with a reference image prepared in advance. In such an apparatus, the conveyance speed of paper sheets at the time of imaging may differ from apparatus to apparatus due to individual differences between apparatuses. In such a case, the size of the image of the paper sheet obtained from the image line sensor is not constant. Even if it is a genuine paper sheet, there are individual differences in external dimensions due to various reasons (for example, shrinkage due to drying after submerging). Due to these events, the paper sheet discriminating device may output an erroneous discrimination result.

  Note that in Patent Document 2 described above, an image obtained by an image sensor and a reference image are partially compared with each other for the purpose of mitigating the influence of a paper sheet misalignment. However, in such a method, the amount of information of an image used for comparison is reduced, and the discrimination accuracy is suppressed. Alternatively, it is necessary to separately acquire a high-definition image in order to accurately determine authenticity. A method of measuring the conveyance speed of paper sheets using a rotary encoder and controlling the imaging cycle of the image line sensor according to the obtained conveyance speed is also conceivable. However, in such a system, it is necessary to add parts, which contributes to an increase in manufacturing cost. In addition, the resolution of the image obtained by the image line sensor depends on the performance of the rotary encoder.

  The present invention has been made in view of such problems, and an object of the present invention is to provide a paper sheet discriminating apparatus and a paper sheet discriminating method capable of enhancing the discrimination accuracy.

  In order to solve the above-described problem, a paper sheet discriminating apparatus according to the present invention is arranged with a transport path for transporting a paper sheet to be discriminated and a direction intersecting the transport direction of the transport path as a longitudinal direction, and on the transport path. An image line sensor that captures a paper sheet that is transported, and a storage area that stores a reference image prepared in advance. By comparing the acquired image output from the image line sensor with the reference image, the paper A discriminating unit for discriminating leaves, and the discriminating unit cuts out a discrimination target image corresponding to a paper sheet from the acquired image, so that the size of the discrimination target image approaches the size of the reference image After the image is enlarged or reduced, the discrimination target image after enlargement or reduction is compared with the reference image.

  In addition, the paper sheet discrimination method according to the present invention images a paper sheet, which is a discrimination target transported on the transport path, using an image line sensor arranged with the direction intersecting the transport direction of the transport path as the longitudinal direction. An imaging step, and a discrimination step for discriminating paper sheets by comparing the acquired image output from the image line sensor with a reference image prepared in advance. After cutting out the discrimination target image which is a corresponding part and enlarging or reducing the discrimination target image so that the size of the discrimination target image approaches the size of the reference image, the discrimination target image after enlargement or reduction and the reference image It is characterized by comparing.

  In these paper sheet discrimination apparatuses and paper sheet discrimination methods, the size of the discrimination target image corresponding to the paper sheet cut out from the acquired image is enlarged or reduced, so that the size of the discrimination target image approaches the size of the reference image. It is done. Thus, for example, even when the length of the discrimination target image is not constant due to variations in the conveyance speed, or even when the authentic paper sheets are expanded or contracted for some reason, their influence is excluded from the discrimination target image. Further, the determination accuracy can be increased by accurately comparing the determination target image and the reference image without reducing the amount of information. In addition, these paper sheet discriminating apparatuses and paper sheet discriminating methods do not require additional parts such as a rotary encoder, so that it is possible to reduce the manufacturing cost of the apparatus and to obtain a high resolution image acquired by the image line sensor. Can be maintained.

  Further, in the paper sheet discrimination device, the discrimination unit discriminates paper sheets by comparing a plurality of reference images stored in the storage area with an enlarged or reduced discrimination target image, and each of the plurality of reference images. May be characterized in that they are images with the same size for each of a plurality of paper sheets having different dimensions. Similarly, the paper sheet discrimination method discriminates paper sheets by comparing a plurality of reference images with an enlarged or reduced discrimination target image in the discrimination step, and each of the plurality of reference images has a size relative to each other. The images may be characterized in that the sizes are the same for each of a plurality of different paper sheets. Conventionally, when the size of a paper sheet as a discrimination target varies depending on the type, the type is discriminated based on the size of the paper sheet as a discrimination target. In such a method, for example, after detecting the outline of a paper sheet in the acquired image, its length and area are obtained, and discrimination is performed based on the length and area. However, depending on the paper sheet, it may be difficult to obtain the length and area accurately, and there is a problem that the discrimination accuracy is lowered. On the other hand, according to the above-described paper sheet discriminating apparatus and paper sheet discriminating method, the size of the reference image relating to a plurality of paper sheets having different dimensions is aligned, so that the pattern can be obtained regardless of the difference in dimensions for each type. Etc. can be determined based on the above. Thereby, even if it is difficult to accurately obtain the length and area, the type of the paper sheet can be determined with high accuracy. Further, it is possible to omit the configuration necessary for the discrimination processing based on the length and area of the paper sheet. Alternatively, by using the discrimination processing based on the length and area of the paper sheet and the paper sheet discrimination device or the paper sheet discrimination method, the discrimination accuracy of a plurality of paper sheets having different dimensions can be significantly improved. Can do. Further, according to the paper sheet discrimination apparatus and the paper sheet discrimination method described above, the procedure of the discrimination process for each paper sheet can be made common, so that the configuration (for example, program configuration) of the discrimination unit can be simplified. As an example, even when a new type is added to the paper sheet to be identified, it is only necessary to add a reference image of the paper sheet of that type, eliminating the need for large-scale program changes. be able to.

  The paper sheet discrimination device may be characterized in that the discrimination unit cuts out the discrimination target image from the acquired image, adjusts the inclination of the discrimination target image, and then enlarges or reduces the discrimination target image. Similarly, the paper sheet discrimination method may be characterized in that, in the discrimination step, the discrimination target image is cut out from the acquired image, the inclination of the discrimination target image is adjusted, and then the discrimination target image is enlarged or reduced. When the width of the transport path is wider than the width of the paper sheet, the paper sheet may be transported in a state of being inclined with respect to a direction orthogonal to the longitudinal direction of the image line sensor. In such a case, if the discrimination target image is tilted with respect to the reference image, even if the paper sheet is authentic, it does not coincide with the reference image, and the discrimination accuracy decreases. On the other hand, by adjusting the inclination of the discrimination target image as described above, the discrimination target image and the reference image can be compared more accurately, and the discrimination accuracy can be further improved.

  Further, the paper sheet determination apparatus may be characterized in that when the determination unit enlarges or reduces the determination target image, interpolation processing is performed on pixels included in the determination target image. Similarly, the paper sheet discrimination method may be characterized in that, in the discrimination step, when the discrimination target image is enlarged or reduced, interpolation processing is performed on pixels included in the discrimination target image. In the above apparatus and method, there is a case where the discrimination target image is disturbed (or the definition is reduced) when the discrimination target image is enlarged or reduced. Even in such a case, by performing interpolation processing on the pixels included in the discrimination target image, it is possible to maintain the fineness of the discrimination target image and accurately compare the discrimination target image and the reference image.

  When interpolation processing is performed on pixels included in the discrimination target image, the interpolation processing is preferably processing by bilinear interpolation or bicubic interpolation. According to the bilinear interpolation, the discrimination accuracy is inferior to that of the bicubic interpolation, but it is possible to facilitate the arithmetic processing and shorten the processing time. Moreover, according to bicubic interpolation, the discrimination accuracy can be increased.

  In addition, the paper sheet discrimination device is configured to enlarge or reduce the discrimination target image and compare the discrimination target image with the reference image when the size of the discrimination target image included in the acquired image is outside a predetermined range. It is good also as not performing. Similarly, in the determination step, when the size of the determination target image included in the acquired image is outside a predetermined range, the paper sheet determination method enlarges or reduces the determination target image and determines whether the determination target image is the reference image. The comparison may not be performed. As a result, it is possible to quickly discriminate paper sheets that are clearly not to be discriminated, and to save useless processing.

  According to the paper sheet discriminating apparatus and the paper sheet discriminating method according to the present invention, the discrimination accuracy can be improved.

It is a block diagram which shows the structure of a banknote discrimination | determination apparatus as one Embodiment of the paper sheet discrimination | determination apparatus which concerns on this invention. It is a top view which shows the structural example of a scanning input part. It is a figure which shows notionally a mode that a discrimination | determination target image is cut out from an acquired image. It is a figure which shows notionally a mode that a collation image is produced | generated from the discrimination | determination target image cut out from the acquired image by the discrimination | determination target image cutting part. It is a figure which shows notionally a mode that several reference images are created from various banknotes. It is a flowchart which shows a banknote discrimination | determination method.

  Hereinafter, embodiments of a paper sheet discriminating apparatus and a paper sheet discriminating method according to the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted. In the following description, the image size in a certain direction means the number of pixels from one end of the image to the other end in the direction.

  FIG. 1 is a block diagram showing a configuration of a bill discriminating apparatus 1A as an embodiment of a paper sheet discriminating apparatus according to the present invention. The bill discriminating apparatus 1A of the present embodiment is a device for discriminating the type and authenticity of a bill as a paper sheet. As shown in FIG. 1, the bill discriminating apparatus 1 </ b> A includes a scanning input unit 10, a discriminating unit 20, and a control unit 30 that controls these operations.

  The scan input unit 10 is a component for capturing an image of a bill that is a discrimination target and generating an acquired image. Here, FIG. 2 is a plan view illustrating a configuration example of the scan input unit 10. As shown in FIG. 2, the scan input unit 10 includes a conveyance path 11, an image line sensor 12, two passage detection sensors 13, and a control circuit 14.

  The conveyance path 11 has a conveyance belt and a drive mechanism that moves the conveyance belt in a predetermined conveyance direction (arrow A in the figure), and conveys the bills B placed on the conveyance belt. The image line sensor 12 is arranged with the direction intersecting the transport direction A of the transport path 11 as the longitudinal direction, and images the banknote B transported on the transport path 11.

  The two passage detection sensors 13 are respectively arranged on both sides in the transport direction A with respect to the image line sensor 12, and control signals 14 indicate signals Sa and Sb indicating that the bill B is passing over the passage detection sensor 13. Output to. The control circuit 14 receives the signal Sa from the passage detection sensor 13 located in front of the image line sensor 12, and when it detects that the bill B has started to pass over the passage detection sensor 13, the image is started to start imaging. The line sensor 12 is controlled. The control circuit 14 receives the signal Sb from the passage detection sensor 13 located behind the image line sensor 12 and terminates the imaging when it detects that the bill B has passed over the passage detection sensor 13. The image line sensor 12 is controlled. The image line sensor 12 generates the acquired image D1 by scanning the surface of the moving bill B with a constant period. The acquired image D1 output from the image line sensor 12 is sent to the determination unit 20 described later via the control circuit 14. In addition, a signal indicating that the bill B has passed is sent to the control unit 30.

  The lateral width of the transport path 11 in the direction orthogonal to the transport direction A is slightly wider (for example, about +5 mm) than the width of the banknote B in the same direction. The size in the transport direction of the acquired image D1 generated by the image line sensor 12 is determined by the number of scans from the start of imaging to the end of imaging. Further, the size of the acquired image D1 in the direction orthogonal to the transport direction is determined by the lateral width of the transport path 11.

  Refer to FIG. 1 again. The determination unit 20 includes a processing unit 21, a storage unit 22, and a determination unit 23. The processing unit 21 and the determination unit 23 are preferably realized by an arithmetic processing device such as a CPU executing a predetermined program. The storage unit 22 is preferably realized by, for example, a volatile memory, except for a reference image storage area 22c described later. This memory may be integrated into a chip with the above arithmetic processing unit, or may be provided separately from the above arithmetic processing unit.

  The processing unit 21 includes a scanning processing unit 21a, a discrimination target image cutout unit 21b, an inclination calculation unit 21c, a discrimination target image rotation unit 21d, and a discrimination target image expansion / contraction unit 21e. The storage unit 22 includes an acquired image storage area 22a, a discrimination target image storage area 22b, a reference image storage area 22c, and a collation image storage area 22d.

  The scanning processing unit 21a inputs the acquired image D1 from the control circuit 14 (see FIG. 2) of the scanning input unit 10 based on an instruction from the control unit 30 that has detected that the banknote B has passed, and this acquired image. D1 is stored in the acquired image storage area 22a. The discrimination target image cutout unit 21b reads the acquisition image D1 from the acquisition image storage area 22a, and cuts out the discrimination target image that is a portion corresponding to a banknote from the acquisition image D1 in units of pixels. Here, FIG. 3 is a diagram conceptually showing a state in which the discrimination target image D2 is cut out from the acquired image D1. The acquired image D1 includes a discrimination target image D2 and a background image D3 around the discrimination target image D2. The discrimination target image cutout unit 21b determines a boundary line between the discrimination target image D2 and the background image D3, and cuts out the discrimination target image D2. The discrimination target image cutout unit 21b stores the cut out discrimination target image D2 in the discrimination target image storage area 22b.

  The inclination calculation unit 21c reads the determination target image D2 from the determination target image storage area 22b, and calculates the inclination of the determination target image D2 with respect to the transport direction A (see FIG. 2). The inclination calculation unit 21c provides inclination data D4 indicating the inclination of the discrimination target image D2 to the discrimination target image rotation unit 21d. Here, the reason for calculating the inclination of the discrimination target image D2 is that the width of the conveyance path 11 is wider than the width of the bill B as described above, and therefore, the direction perpendicular to the longitudinal direction of the image line sensor 12 is used. This is because the bill B may be conveyed in a tilted state.

  The discrimination target image rotation unit 21d reads the discrimination target image D2 from the discrimination target image storage area 22b, and adjusts the inclination of the discrimination target image D2 based on the tilt data D4 provided from the tilt calculation unit 21c. For example, the discrimination target image rotation unit 21d adjusts the inclination of the discrimination target image D2 so that the long side direction of the discrimination target image D2 matches the transport direction A. The discrimination target image rotation unit 21d sends the discrimination target image D2 after the tilt adjustment to the discrimination target image expansion / contraction unit 21e.

  The discrimination target image expansion / contraction part 21e enlarges or reduces the discrimination target image D2 so that the size of the discrimination target image D2 in the transport direction and the direction orthogonal to the transport direction approaches (preferably matches) the size of the reference image. Reduce. In addition, the discrimination target image expansion / contraction unit 21e may include a pixel interpolation unit 21f that performs an interpolation process on pixels included in the discrimination target image D2. The pixel interpolation process is a process for correcting a portion where the data between adjacent pixels becomes discontinuous due to enlargement or reduction of the discrimination target image D2 so that these data are pseudo-continuous. is there. Examples of such interpolation processing include processing by bilinear interpolation or bicubic interpolation. The discrimination target image expansion / contraction unit 21e stores the discrimination target image D2 that has been enlarged or reduced and interpolated in the collation image storage area 22d as the collation image D5.

  Here, FIG. 4 is a diagram conceptually illustrating a state in which the collation image D5 is generated from the discrimination target image D2 cut out from the acquired image D1 by the discrimination target image cutout unit 21b. FIG. 4A shows a discrimination target image D2 acquired by the image line sensor 12 when the transport speed in the transport path 11 (see FIG. 2) is faster than a specified speed, for example. In the discrimination target image D2, the size L1 in the transport direction (longitudinal direction) is shorter than the standard size La. Moreover, FIG.4 (b) has shown the discrimination | determination target image D2 acquired by the image line sensor 12 about the banknote B dried, for example after being submerged. In the discrimination target image D2, the size L1 in the longitudinal direction is shorter than the standard size La, and the size L2 in the short direction is shorter than the standard size Lb. FIG. 4C shows a discrimination target image D2 that is inclined by an angle θ with respect to the transport direction.

  The discrimination target image rotating unit 21d and the discrimination target image expanding / contracting unit 21e are based on the discrimination target image D2 as shown in FIGS. 4A to 4C, and the standard image shown in FIG. A collation image D5 having an inclination and standard sizes La and Lb is generated.

  The reference image storage area 22c is a storage area for storing a reference image for a genuine banknote prepared in advance. Unlike the other areas of the storage unit 22, the reference image storage area 22c is preferably realized by, for example, a nonvolatile memory. . This non-volatile memory may be integrated into a chip with the arithmetic processing device constituting the processing unit 21 or may be provided separately from the arithmetic processing device. Here, FIG. 5 is a diagram conceptually showing how the reference images D6 to D8 are created from various banknotes. There are 10,000 yen bills shown in FIG. 5 (a), 5,000 yen bills shown in FIG. 5 (b), and 1000 yen bills shown in FIG. 5 (c). Images D6 to D8 are stored in the reference image storage area 22c. Moreover, each of the reference images D6 to D8 is an image in which the sizes are aligned with respect to each of a plurality of banknotes having different dimensions. In other words, the size L3 in the longitudinal direction of the 10,000 yen bill, the 5,000 yen bill, and the thousand yen bill is different from each other, but the reference images D6 to D8 stored in the reference image storage area 22c are the sizes in the longitudinal direction. Is an image adjusted to be equal to the standard size La. Further, the sizes in the short direction of the reference images D6 to D8 are equal to the standard size Lb. As an example, the size (number of pixels) of the standard size La is 160 pixels, and the size (number of pixels) of the standard size Lb is 80 pixels.

  The determination unit 23 reads the collation image D5 from the collation image storage area 22d, and reads the reference images D6 to D8 from the reference image storage area 22c. Then, the determination unit 23 compares the collation image D5 with the reference images D6 to D8 to determine whether the banknote B to be determined is a 10,000 yen bill, a 5,000 yen bill, or a thousand yen bill. It discriminates, and discriminates its true / false.

  In the configuration described above, the determination unit 20 adjusts and determines the inclination of the determination target image D2 by the determination target image rotation unit 21d when the size of the determination target image D2 included in the acquired image D1 is outside the predetermined range. It is not necessary to enlarge or reduce the discrimination target image D2 by the target image expansion / contraction unit 21e and to compare the collation image D5 and the reference images D6 to D8 by the determination unit 23.

  Operation | movement of 1 A of banknote discrimination | determination apparatuses of this embodiment provided with the above structure is demonstrated with the banknote discrimination method by this embodiment. FIG. 6 is a flowchart showing the bill discriminating method according to the present embodiment.

  First, the banknote B which is a discrimination target is placed on the transport path 11 shown in FIG. 2, and the banknote B is imaged using the image line sensor 12 while moving the transport path 11 in the transport direction A (imaging). Step S10). At this time, the image line sensor 12 generates the acquired image D1 by scanning the surface of the moving bill B with a certain period.

  Next, the determination unit 20 determines the type and authenticity of the bill B by comparing the acquired image D1 output from the image line sensor 12 with the reference images D6 to D8 prepared in advance (determination step). S20). This determination step S20 includes the following steps S21 to S25.

  In step S21, the discrimination target image cutout unit 21b cuts out the discrimination target image D2 that is a portion corresponding to the banknote B from the acquired image D1. Next, in step S22, the determination unit 20 determines whether or not the size of the determination target image D2 is within a predetermined range. When the size is within the predetermined range (step S22; Yes), the determination unit 20 performs the following steps S23 to S25. If the size is out of the predetermined range (step S22; No), the determination unit 20 does not perform the subsequent steps S23 to S25 and ends the process.

  Subsequently, in step S23, the discrimination target image rotation unit 21d adjusts the tilt of the discrimination target image D2 based on the tilt angle of the discrimination target image D2 calculated by the tilt calculation unit 21c (FIG. 4C and FIG. 4 (d)). Subsequently, in step S24, the discrimination target image expansion / contraction unit 21e enlarges or reduces the size of the discrimination target image D2, thereby bringing it close to the size of the reference images D6 to D8 (FIGS. 4A and 4B). And FIG. 4 (d)). At this time, the pixel interpolation unit 21f may perform interpolation processing on the pixels included in the determination target image D2. Subsequently, in step S24, the determination unit 23 determines the type and authenticity of the bill B by comparing the collation image D5 generated in the determination target image expansion / contraction unit 21e with the reference images D6 to D8.

  The effects obtained by the banknote discriminating apparatus 1A and the banknote discrimination method according to the present embodiment described above will be described. As described above, when the conveyance speed of the bill B in the conveyance path 11 is different for each apparatus, the length of the image of the bill B obtained from the image line sensor 12 in the conveyance direction A is not constant. For example, when the transport speed is faster than the standard, the size L1 in the transport direction of the determination target image D2 is shorter than the standard size La (for example, 160 pixels), and when the transport speed is slower than the standard, the size L1. Becomes longer than the standard size La. Moreover, even if it is a genuine banknote B, an individual difference may arise in an external dimension for various reasons. For example, when the banknote B is accidentally submerged and then dried, the banknote B may shrink and be smaller than the original outer dimensions. In such a case, the size L1 in the transport direction of the discrimination target image D2 is shorter than the standard size La, and the size L2 in the direction orthogonal to the transport direction is shorter than the standard size Lb (for example, 80 pixels). Therefore, if the discrimination process is performed under these circumstances, an erroneous discrimination result may be output even though all of the bills are genuine.

  With respect to such a problem, in the present embodiment, the size of the discrimination target image D2 corresponding to the banknote B cut out from the acquired image D1 is enlarged or reduced, so that the size of the discrimination target image D2 is set to the reference images D6 to D6. It can be brought close to the size of D8. Thus, for example, even when the length of the discrimination target image D2 is not constant due to variations in the conveyance speed or when the genuine banknote B is expanded or contracted for some reason, the influence is excluded from the discrimination target image D2. In addition, the discrimination accuracy can be increased by accurately comparing the discrimination target image D2 and the reference images D6 to D8 without reducing the amount of information. Further, since no additional parts such as a rotary encoder are required, the manufacturing cost of the apparatus can be suppressed, and the high resolution of the acquired image D1 obtained by the image line sensor 12 can be maintained.

  Further, as in the present embodiment, in the determination step S20, the determination unit 20 uses the plurality of reference images D6 to D8 stored in the reference image storage area 22c and the determination target image D2 after enlargement or reduction (that is, collation). When the banknote B is discriminated by comparing with the image D5), each of the plurality of reference images D6 to D8 is preferably an image in which the sizes are aligned with respect to each of the plurality of banknotes B having different dimensions. Conventionally, when the dimension of the banknote B as a discrimination target differs depending on the denomination, the denomination is determined based on the dimension of the banknote B as a discrimination target. In such a method, for example, after detecting the outline of the banknote B in the acquired image, the length and area thereof are obtained, and discrimination is performed based on the length and area. However, for example, in foreign banknotes, it may be difficult to accurately obtain the length and area, and there is a problem that the discrimination accuracy is lowered. On the other hand, according to the present embodiment, the sizes of the reference images D6 to D8 related to the plurality of banknotes B having different dimensions are aligned, so that the determination is made based on the pattern or the like regardless of the difference in the dimensions for each denomination. It can be performed. Thereby, even if it is a banknote in which it is difficult to obtain | require a length and an area correctly, the denomination can be discriminate | determined accurately. Moreover, it becomes possible to omit the structure required for the discrimination | determination process based on the length and area of a banknote. Or the discrimination | determination precision based on the length and area of a banknote, and the banknote discrimination | determination apparatus 1A or banknote discrimination method of this embodiment can be used together, and the discrimination | determination precision of the several banknote from which a dimension mutually differs can be improved markedly. Moreover, according to said banknote discrimination | determination apparatus 1A and a banknote discrimination | determination method, since the procedure of the discrimination | determination process regarding each banknote can be made shared, the structure (for example, program structure) of the discrimination | determination part 20 can be simplified. As an example, even when a new denomination is added to a banknote as an identification target, it is only necessary to add a reference image of the denomination, and a large-scale program change can be made unnecessary.

  Further, as in the present embodiment, in the determination step S20, the determination unit 20 cuts out the determination target image D2 from the acquired image D1, adjusts the inclination of the determination target image D2, and then enlarges the determination target image D2. Or it is preferable to reduce. As described above, when the width of the transport path 11 is wider than the width of the bill B, the bill B may be transported in a state of being inclined with respect to the direction orthogonal to the longitudinal direction of the image line sensor 12. In such a case, if the discrimination target image D2 is inclined with respect to the reference images D6 to D8, even if the banknote B is authentic, it does not coincide with the reference images D6 to D8, and the discrimination accuracy decreases. On the other hand, by adjusting the inclination of the discrimination target image D2, the discrimination target image D2 and the reference images D6 to D8 can be compared more accurately, and the discrimination accuracy can be further improved.

  Further, as in the present embodiment, in the determination step S20, the determination unit 20 preferably performs an interpolation process on the pixels included in the determination target image D2 when enlarging or reducing the determination target image D2. Even when the discrimination target image D2 is disturbed (or the definition is reduced) when the discrimination target image D2 is enlarged or reduced, the interpolation processing is performed on the pixels included in the discrimination target image D2. The fineness of the discrimination target image D2 can be maintained, and the discrimination target image D2 and the reference images D6 to D8 can be accurately compared.

  As described above, when the interpolation processing is performed on the pixels included in the determination target image D2, the interpolation processing is preferably processing by bilinear interpolation or bicubic interpolation. According to the bilinear interpolation, the discrimination accuracy is inferior to that of the bicubic interpolation, but it is possible to facilitate the arithmetic processing and shorten the processing time. Moreover, according to bicubic interpolation, the discrimination accuracy can be increased.

  Further, as in the present embodiment, when the size of the discrimination target image D2 included in the acquired image D1 is outside the predetermined range in the discrimination step S20, the discrimination unit 20 enlarges or reduces the discrimination target image D2, and It is preferable not to compare the discrimination target image D2 with the reference images D6 to D8. Thereby, it is possible to quickly discriminate bills B that are not clearly discriminated, and to save useless processing.

  In the present embodiment, the size of the discrimination target image D2 before enlargement or reduction is not limited. For example, the sizes L1 and L2 of the discrimination target image D2 may be two times, four times, or 0.5 times the standard sizes La and Lb. In these cases, the size of the discrimination target image D2 is the standard size La. , Lb may be enlarged or reduced so as to approach (preferably match). When the size of the discrimination target image D2 is larger than the standard size, a sufficient amount of information for discrimination can be secured. Further, when the size of the discrimination target image D2 is smaller than the standard size, the processing time for discrimination can be shortened.

  The paper sheet discriminating apparatus and the paper sheet discriminating method according to the present invention are not limited to the above-described embodiments, and various other modifications are possible. For example, although the banknote was shown as an example of paper sheets in embodiment mentioned above, the paper sheets used as discrimination | determination object in this invention are not restricted to a banknote.

DESCRIPTION OF SYMBOLS 1A ... Banknote discrimination | determination apparatus, 10 ... Scanning input part, 11 ... Conveyance path, 12 ... Image line sensor, 13 ... Passing detection sensor, 14 ... Control circuit, 20 ... Discrimination part, 21 ... Processing part, 21a ... Scanning processing part, 21b: Discrimination target image cutout unit, 21c: Inclination calculation unit, 21d ... Discrimination target image rotation unit, 21e ... Discrimination target image expansion / contraction unit, 21f ... Pixel interpolation unit, 22 ... Storage unit, 22a ... Acquired image storage area, 22b ... Discrimination target image storage area, 22c... Reference image storage area, 22d .. Collation image storage area, 23... Determination section, 30... Control section, A ... Conveyance direction, B ... Banknote, D1 ... Acquired image, D2 , D3 ... background image, D4 ... inclination data, D5 ... collation image, D6-D8 ... reference image.

Claims (8)

A transport path for transporting the paper sheet to be identified;
An image line sensor that images the paper sheet that is disposed on the conveyance path as a longitudinal direction and intersects the conveyance direction of the conveyance path, and
A storage area for storing a reference image prepared in advance, and a determination unit that determines the paper sheet by comparing the acquired image output from the image line sensor with the reference image;
The discriminating unit cuts out a discrimination target image corresponding to the paper sheet from the acquired image, and enlarges or reduces the discrimination target image so that the size of the discrimination target image approaches the size of the reference image. After that, the paper sheet discrimination device that compares the discrimination target image after enlargement or reduction with the reference image. The determination unit determines the paper sheet by comparing the plurality of reference images stored in the storage area and the determination target image after the enlargement or reduction,
2. The paper sheet discrimination apparatus according to claim 1, wherein each of the plurality of reference images is an image having a uniform size with respect to each of a plurality of paper sheets having different dimensions.   The discrimination unit adjusts an inclination of the discrimination target image after cutting out the discrimination target image from the acquired image, and then enlarges or reduces the discrimination target image. The paper sheet discrimination device according to 1.   The paper sheet according to any one of claims 1 to 3, wherein the determination unit performs an interpolation process on pixels included in the determination target image when the determination target image is enlarged or reduced. Discriminator.   The paper sheet discriminating apparatus according to claim 4, wherein the interpolation processing is processing by bilinear interpolation or bicubic interpolation.   The discrimination unit performs enlargement or reduction of the discrimination target image and comparison between the discrimination target image and the reference image when the size of the discrimination target image included in the acquired image is outside a predetermined range. The paper sheet discriminating apparatus according to claim 1, wherein the paper sheet discriminating apparatus is not provided. An imaging step for imaging a sheet to be discriminated that is transported on the transport path using an image line sensor arranged with the direction intersecting the transport direction of the transport path as a longitudinal direction;
A determination step of determining the paper sheet by comparing the acquired image output from the image line sensor with a reference image prepared in advance;
In the discrimination step, a discrimination target image that is a portion corresponding to the paper sheet is cut out from the acquired image, and the discrimination target image is enlarged or reduced so that the size of the discrimination target image approaches the size of the reference image. Then, the paper sheet discrimination method comprising comparing the discrimination target image after enlargement or reduction with the reference image. In the determination step, the paper sheets are determined by comparing a plurality of the reference images and the determination target images after the enlargement or reduction,
8. The paper sheet discrimination method according to claim 7, wherein each of the plurality of reference images is an image having a uniform size with respect to each of a plurality of paper sheets having different dimensions.

Images