JPWO2010035314A1 - Coin damage determination apparatus and damage determination method - Google Patents

Abstract

The coin damage determination apparatus according to the present invention calculates the luminance frequency by summing up the luminance values of each pixel for each pixel column in a predetermined direction (for example, the horizontal direction) in a predetermined region of the coin image captured by the imaging means. A luminance histogram creating means for creating a luminance histogram, and an absolute difference for calculating an absolute difference value sum that is a sum of absolute values of differences between the luminance frequencies adjacent to each other in the luminance histogram in a predetermined section of the luminance histogram A value sum calculating means. And it is discriminate | determined whether a coin is a dirty coin based on the comparison with the said absolute difference value sum and the predetermined threshold value.

Description

  The present invention relates to a coin damage determination device and a damage determination method for determining whether or not a coin is a fouled coin, and in particular, acquires coin image data and uses the acquired image data to determine whether or not the coin is correct. The present invention relates to a damage determination apparatus and a damage determination method.

  There are various automatic transaction machines that handle coins to collect the damaged coins so that the damaged coins (dirty coins) are not recycled. Inside such an apparatus is provided a discriminating device that examines the state of fouling of a coin and discriminates whether or not the coin is a fouled coin. In such a discriminating device, it is naturally desirable to accurately detect the fouling state of a coin and discriminate the correctness of the coin.

For example, Patent Document 1 proposes a contamination state determination method and apparatus capable of accurately detecting the contamination state of a coin. In the fouling state determination method proposed in Patent Document 1, an entire image of a coin is acquired and a pixel number distribution at each gradation value is obtained. Then, a gradation value in which the area ratio obtained from the obtained pixel number distribution becomes a predetermined ratio value as a whole is calculated from the higher gradation, and the calculated gradation value is a preset reference gradation value. If the price is lower than that of the coin, the coin is judged to be soiled.
Japanese Patent Laid-Open No. 10-11965

  By the way, among the coins in circulation in the world, some of the coins that are judged to be fouled are black and some are white. In the case of the configuration of Patent Document 1, a coin that is black and dirty (in other words, a dark image image) is determined as a dirty coin. However, for example, a coin that is white and dirty (in other words, a bright image) such as a rusted and white-cooked coin may not be judged as a damaged coin.

  Therefore, the applicant acquires the image data of the coin, and uses the matching value based on the template matching between the acquired image data and the reference data prepared in advance to determine whether the coin is a fouled coin. We are also considering to determine However, this method involves a step of normalizing the brightness of image data acquired for the purpose of reducing the influence of variations in image sensors in accordance with the brightness of the template. For this reason, there is a problem that the difference between the dirty currency and the normal currency becomes small, and in fact, what is a dirty currency may be determined as normalization.

  In view of the above points, an object of the present invention is to provide a coin damage determination apparatus and a damage determination method that can accurately determine whether or not a coin is a damaged coin.

  In order to achieve the above object, the present invention is an apparatus for discriminating whether or not a coin is a fouled coin, wherein the coin is conveyed in the conveyance path and the conveyance path. In a predetermined area of the coin image imaged by the imaging means, imaging means for imaging the coin, data digitizing means for converting each pixel data obtained by the imaging means into a digitized luminance value, and the imaging means A luminance histogram generating means for calculating a luminance frequency by summing the luminance values of each pixel for each pixel column in a predetermined direction, and generating a luminance histogram by arranging the luminance frequencies in the order in which the pixel columns are arranged; Absolute difference value sum calculating means for calculating an absolute difference value sum that is a sum of absolute values of differences between adjacent luminance frequencies in a predetermined section of the luminance histogram; Threshold storage means for storing a first threshold value determined in advance for comparison with the difference value sum, and whether the coin is a fouled coin based on a comparison between the absolute difference value sum and the first threshold value It is characterized by comprising a damage discriminating means for discriminating whether or not.

  In this configuration, it is determined whether or not the coin is a defaced coin using an absolute difference sum that is an index of the degree of unevenness of the coin. For this reason, not only when the coin is dirty and dark, but also when the coin is soiled with a bright image, such as when the coin is clouded, it can be determined as a dirty coin. Become.

  In the coin damage determination apparatus having the above-described configuration, the apparatus further includes an average value calculating unit that calculates an average value of the luminance frequencies in the predetermined section of the luminance histogram, and the threshold value storage unit compares the average value with the average value. A second threshold value determined in advance is further stored, and the damage determination means compares the absolute difference value sum with the first threshold value, and compares the average value with the second threshold value. It is preferable to determine whether or not the coin is a damaged coin based on the above.

  According to this configuration, in addition to the above-described sum of absolute difference values, the average value of the luminance frequency that is an indicator of the brightness of the coin is also used to determine the correctness of the coin. For this reason, for example, a coin that is clear and uneven in coins but is black and dirty can be determined as a dirty coin, and the correctness of the coin can be more accurately performed.

  In the coin damage determination apparatus having the above-described configuration, division value calculation means for calculating a division value obtained by dividing the sum of absolute difference values by the average value or a division value obtained by dividing the average value by the sum of absolute difference values. The threshold value storage means further stores a third threshold value that is determined in advance for comparison with the division value, and the fitness determination means includes the absolute difference value sum, the first threshold value, and the like. It is preferable to determine whether or not the coin is a fouled coin based on a comparison between the average value and the second threshold value and a comparison between the division value and the third threshold value.

  In this configuration, a division value obtained by dividing the above-mentioned absolute difference value sum by the above-mentioned average value, or a division value obtained by dividing the above-described average value by the above-mentioned absolute difference value sum, is used to determine the correctness of the coin. It has a configuration adopted in the criteria. Coins suitable for distribution are generally bright images with clear and uneven coins (patterns are clear). In other words, coins with a poor balance, in which the unevenness of the coin is clear but a dark image, or the unevenness of the coin is unclear but bright, are often damaged coins. In this regard, in this configuration, the above-described division value is calculated, and the division value is compared with a preset threshold value by statistical processing to determine whether the loss is correct. It is easy to discriminate it as a coin, and it is possible to more accurately determine the correctness of the coin.

  In the coin damage determination device having the above-described configuration, the threshold storage means stores a plurality of threshold values so that the user can change the setting level of whether or not the coin is a dirty coin. preferable.

  The criterion for determining whether or not a coin is a damaged coin is not necessarily a fixed criterion, and may vary depending on the situation in which the apparatus is used. For this reason, a user who can change the setting of the judgment level as in this configuration is a user-friendly device.

  In the coin damage determination apparatus having the above-described configuration, the coin further includes a denomination identifying unit that identifies a denomination of the coin conveyed through the conveyance path, and the threshold storage unit stores a different threshold for each denomination. It is preferable.

  Coins generally have a plurality of denominations, which differ in size and material. For this reason, in order to deal with a plurality of types of denominations, it is necessary to change the criteria for determining the damage for each denomination, and this configuration is such a configuration. Therefore, the coin damage determination apparatus of this configuration has a wide application range.

  In the coin damage determination device having the above-described configuration, the coin further includes a coin center detection unit that detects a center position of the coin, and the brightness histogram created by the brightness histogram creation unit is the center position by the coin center detection unit. It may be used to detect.

  The detection of the center position of the coin is often required for performing image processing, and this configuration includes a detection means for the center position. In this configuration, the luminance histogram used for detecting the center position is the same as the luminance histogram used for calculating the absolute difference value sum and average value described above for determining the damage. It has become. In other words, in determining the correctness of coins, the same brightness histogram as that used in other processes is used, and the correctness of coins can be determined efficiently.

  Further, in the coin damage determination device having the above-described configuration, an image processing unit that cuts out and processes a ring-shaped image based on the center position detected by the coin center detection unit for the coin image obtained by the imaging unit. And true / false judgment means for judging whether or not the coin is genuine based on a collation value between an image obtained by the image processing means and a template prepared in advance. The means may determine whether or not the coin is a fouled coin only when the authenticity determining means determines that the coin is genuine.

  According to this structure, it is a structure which does not perform damage determination about a forged coin. In other words, the configuration is such that unnecessary discrimination processing is omitted, and a device with high processing efficiency can be obtained.

  Further, in the coin damage determination device having the above-described configuration, the threshold value storage unit further stores a fourth threshold value previously determined for comparison with the reference value, and the damage determination unit stores the absolute difference. Separately from the determination using the sum of values, it is determined whether or not the coin is a defaced coin based on the comparison between the collation value and the fourth threshold value. If it is determined that the coin is a dirty coin, the coin may be determined to be a dirty coin.

  According to this configuration, the template is used separately from the determination of the damage using the sum of absolute difference values described above (this may include the determination using the average value of the luminance frequency). The collation value is also used for determining the fitness. In other words, it is a configuration that determines the correctness of coins in a two-stage manner, and it is possible to determine the correctness more accurately.

  In order to achieve the above object, the present invention provides a coin damage determination method for determining whether or not a coin is a fouled coin, the imaging step for imaging the coin conveyed on a conveyance path, and the imaging A data digitizing step for converting each pixel data obtained in the step into a digitized luminance value, and a predetermined region of the coin image captured in the imaging step, for each pixel column in a predetermined direction. A luminance histogram creating step of calculating a luminance frequency by summing up the luminance values and generating a luminance histogram, and a value obtained by summing up absolute values of differences between the luminance frequencies adjacent to each other in the luminance histogram in a predetermined section of the luminance histogram An absolute difference value sum calculating step of calculating an absolute difference value sum, and an average value calculating step of calculating an average value of the luminance frequencies in the predetermined section of the luminance histogram , For each of the absolute difference sum and the average value, and comprising: the fitness determination step of the coin as compared to the predetermined threshold, it is determined whether the fouling currency, the.

  According to this configuration, the correctness of the coin is determined using the sum of absolute difference values that is an index of the degree of unevenness of the coin and the average value of the luminance frequency that is an index of the brightness of the coin. For this reason, it is possible to discriminate coins with unclear irregularities and dark image coins as fouled coins. That is, not only a dark image of coins, but also, for example, a coin that is clouded as a whole can be determined as a dirty coin, and correct / incorrect determination of the coin can be performed.

  The coin damage determination method having the above-described configuration may further include a coin center detection step of detecting a center position of the coin, and the luminance histogram creation step may be a step included in the coin center detection step.

  According to this configuration, the luminance histogram used for calculating the above-described absolute difference value sum and average value is the same as the luminance histogram used for detecting the center position. That is, in determining the correctness of coins, the same brightness histogram as that used in other processes is used, so that the correctness of coins can be determined efficiently.

  According to the present invention, it is possible to provide a coin damage determination apparatus and a damage determination method capable of accurately determining whether or not a coin is a damaged coin.

It is a block diagram which shows the structure of the coin damage determination apparatus of this embodiment. It is a figure which shows the example of installation of the hole and material measurement part with which the damage determination apparatus of the coin of this embodiment is equipped, and an imaging part. It is the schematic sectional drawing which showed the structure of the imaging part with which the damage determination apparatus of the coin of this embodiment is provided. It is a figure for demonstrating the installation position of the timing sensor with which an imaging part is provided. It is a flowchart which shows the damage determination procedure in the damage determination apparatus of the coin of this embodiment. It is a flowchart which shows the example of the identification procedure of the money type in a money type identification part. It is a figure for demonstrating the method of producing a horizontal histogram from the image image of a coin, and detecting the center position of the horizontal direction of a coin. It is a figure for demonstrating the method of producing a vertical histogram from the image image of a coin, and detecting the center position of the vertical direction of a coin. It is a figure for demonstrating the calculation performed by the calculation part for damage determination. It is a figure for demonstrating the cutting method of the ring image by a ring cutting part. It is a figure which shows the data arrangement | sequence after cutting out a ring image. It is a figure for demonstrating the rotation collation process using the template for matching. It is the figure which showed the image of a current currency (normal currency), and the horizontal histogram corresponding to this image. It is the figure which showed the image of the dirty coin which became black and the horizontal histogram corresponding to this image. It is the figure which showed the image of the dirty dirt which became white and the horizontal histogram corresponding to this image. For 100 yen coins (front side) of new coins, current currencies, and defaced coins, the absolute difference value sum and average value in a predetermined section of the horizontal histogram are calculated and plotted with the horizontal axis as the average value and the vertical axis as the absolute difference value sum. It is a graph. It is the graph which increased the kind of dirty coins and re-plotted to the graph of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Coin damage determination apparatus 2 Coin 3 Carriage path 12 Imaging part (imaging means)
13 A / D converter (data digitization means)
15 Outer diameter / center position detector (coin center detection means)
16 Denomination identification part (denomination identification means)
18 Image processing unit for verification (image processing means)
19 Template verification part (authentication discrimination means)
21 Physical damage determination unit (Perfect determination means)
22 Threshold value storage unit for threshold value determination (threshold value storage means)
151 Luminance histogram creation unit (luminance histogram creation means)
171 Absolute difference value sum calculation unit (absolute difference value sum calculation means)
172 Average value calculation unit (average value calculation means)
173 Division value calculation unit (division value calculation means)

  Hereinafter, embodiments of a coin damage determination apparatus and a damage determination method according to the present invention will be described in detail with reference to the drawings. In the present specification, a damaged coin is expressed as a damaged coin, and a non-stained coin is expressed as a normal coin. Moreover, the correctness determination of a coin refers to determining whether a coin is a dirty coin. Also, authenticity determination refers to determining whether or not a coin is genuine.

(Configuration of coin damage determination device)
First, the configuration of the coin damage determination apparatus according to the present embodiment will be described mainly based on FIGS. 1 and 2. FIG. 1 is a block diagram illustrating a configuration of a coin damage determination apparatus according to the present embodiment. FIG. 2 is a diagram illustrating an installation example of the hole / material measuring unit 11 and the imaging unit 12 provided in the coin damage determination apparatus 1 according to the present embodiment.

  The hole / material measuring unit 11 is provided to identify the type of coin (denomination), and specifically includes a hole sensor 111 and a material sensor 112. A photo interrupter is used for the hole sensor 111, and a light emitting portion and a light receiving portion that are arranged to face each other are provided so as to sandwich the conveyance path 3. When the coin 2 passes through the hole sensor 111, if the coin 2 has no hole, the light receiving unit remains blocked. However, if the coin 2 has a hole, light enters the light receiving unit. For this reason, the presence or absence of a hole in the coin 2 can be detected.

  The material sensor 112 is arranged at the rear stage of the hole sensor 111 (on the downstream side of the transport path 3). A magnetic sensor is used as the material sensor 112, and the material of the coin is identified from the magnetic characteristics of the coin 2 detected by the magnetic sensor.

  The imaging unit 12 is arranged at the subsequent stage of the material sensor 112 and can capture the entire image of the coin 2 conveyed on the conveyance path 3. FIG. 3 is a schematic cross-sectional view illustrating a configuration of the imaging unit 12 included in the coin damage determination apparatus 1 according to the present embodiment. As shown in FIG. 3, the imaging unit 12 includes a first LED (Light Emitting Diode) 121, a second LED (Light Emitting Diode) 122, a first mirror 123, a second mirror 124, a timing sensor 125, and a lens 126. An imaging device 127 and a camera substrate 128.

  A plurality of first LEDs 121 are arranged in a ring shape, and are used as illumination for the Low angle. A plurality of second LEDs 122 are arranged in a ring shape at a position different from the first LED 121, and are used as illumination for a high angle. Here, the Low angle means that the angle formed by the optical axis of the illumination light and the transport surface is small, and the High angle means that the angle formed by the optical axis of the illumination light and the transport surface is large. . The reason why the two illuminations for the low angle and the high angle are arranged in this way is to clearly image the stamped portion (pattern) of the coin 2.

  The first mirror 123 reflects the light emitted from the first LED 121 and causes the light from the first LED 121 to enter the coin 2 at a low angle. The second mirror 124 reflects the light emitted from the second LED 122 and causes the light from the second LED 122 to enter the coin 2 at a high angle.

  The timing sensor 125 is a sensor for detecting the timing at which the imaging unit 12 captures a coin image. The timing sensor 125 is composed of a photo interrupter, and is turned on / off depending on the position of the coin 2 conveyed on the conveyance path 3. The timing sensor 125 detects the imaging timing based on the on / off change. When the timing sensor 125 detects that the coin 2 is at a position to be imaged, the imaging unit 12 images the coin 2 irradiated with illumination light.

  FIG. 4 is a diagram for explaining an installation position of the timing sensor 125 included in the imaging unit 12. As shown in FIG. 4, in this embodiment, the coin 2 is shifted to one of the two width regulation guides 3 a and 3 b (width regulation guide 3 a side) of the conveyance path 3 and conveyed by a conveyance belt (not shown) or the like. The position of the timing sensor 125 is determined so that the coin 2 is imaged when the entire coin 2 conveyed in a collocated state is present in the imaging area 30 with respect to all denominations. .

  The lens 126 is a lens that captures reflected light reflected by the surface of the coin 2 irradiated with illumination light. The imaging device 127 is a two-dimensional image sensor. Specific examples of the two-dimensional image sensor include a CCD (Charge-Coupled Devices) image sensor and a CMOS (Complementary Metal-Oxide Semiconductor) image sensor. The number of pixels of an image captured by the imaging device 127 is 256 × 240 pixels, and one pixel has a size of about 0.13 mm × 0.13 mm. A circuit for controlling the imaging unit 12 such as control of the timing sensor 125 and the imaging device 127 is formed on the camera substrate 128.

  Returning to FIG. 1, the A / D conversion unit 13 has 256 gradations (0 to 255) of brightness for each pixel of the image data (analog data) of the coin 2 output from the imaging device 127 of the imaging unit 12. (Luminance value) is allocated and converted to digital data. The A / D converter 13 is an embodiment of the data digitizing means of the present invention.

  The memory 14 stores the image data of the coin 2 digitized by the A / D conversion unit 13.

  The outer diameter / center position detection unit 15 detects the outer diameter and the center position of the coin 2 using the image data of the coin 2 stored in the memory 14. The outer diameter and center position of the coin 2 are used, for example, for performing template verification (described later) for determining the denomination of the coin 2 or determining the authenticity of the coin 2.

  The outer diameter / center position detection unit 15 is provided with a luminance histogram creation unit 151. The luminance histogram creation unit 151 creates two luminance histograms, a horizontal histogram and a vertical histogram, which will be described later in detail. The outer diameter / center position detection unit 15 calculates the outer diameter value D and the center coordinates (Cx, Cy) of the coin 2 based on the two luminance histograms. Details of the method of creating the luminance histogram and the method of detecting the outer diameter / center position will be described later. The outer diameter / center position detection unit 15 is an embodiment of the coin center detection means of the present invention.

  The denomination identifying unit 16 identifies the denomination of the coin 2 that is the object of discrimination based on the information obtained by the hole / material measuring unit 11. The obtained denomination information is output to the outer diameter / center position detection unit 15, the verification image processing unit 18, the template verification unit 19, and the fitness determination unit 20. Details of the denomination identification by the denomination identification unit 16 will be described later. Note that the determination here is a primary determination, and the final denomination is determined by template matching performed later. Moreover, in this embodiment, although it has the structure which performs money type identification based on the information obtained by the hole and material measurement part 11, it is comprised by the sensor (for example, magnetic sensor) which measures the outer diameter and thickness of a coin. May be arranged separately or as an integrated composite sensor, and these types of information may be combined to identify the denomination.

  The damage determination calculation unit 17 calculates a calculation value used to determine the damage of the coin 2 based on the horizontal histogram created by the luminance histogram creation unit 151 provided in the outer diameter / center position detection unit 15. . In detail, the damage determination calculation unit 17 includes an absolute difference value sum calculation unit 171, an average value calculation unit 172, and a division value calculation unit 173.

  The absolute difference value sum calculation unit 171 calculates absolute values of differences between adjacent luminance frequencies of the horizontal histogram in a predetermined section of the horizontal histogram, and calculates the absolute difference value sum by summing up the absolute values of the obtained difference values. . The average value calculation unit 172 calculates the average value of the luminance frequencies in a predetermined section of the horizontal histogram (the same as the predetermined section when the absolute difference value sum is calculated). The division value calculation unit 173 obtains a division value by dividing the absolute difference value sum obtained by the absolute difference value sum calculation unit 171 by the average value obtained by the average value calculation unit 172. The obtained absolute difference value sum, average value, and division value are output to the fitness judgment unit 21, and these are used to judge the fitness of the coin 2.

  The collation image processing unit 18 processes the image data stored in the memory 14 so that the template collation unit 19 can perform collation processing with the basic template. The collation image processing unit 18 includes a ring cutout unit 181, an edge enhancement unit 182, a compression unit 183, and a luminance conversion unit 184.

  The ring cutout unit 181 cuts out the image of the coin 2 in a ring shape around the center coordinates detected by the outer diameter / center position detection unit 15, and the rotation angle θ direction is set to the X direction and the radius R is cut out. It is converted into two-dimensional array data with the direction as the Y direction. The edge emphasizing unit 182 performs a differentiation process on the image data (two-dimensional array data) cut out by the ring cut-out unit 181 and emphasizes the edge portion where the luminance change is abrupt. Thereby, features, such as a pattern of the coin 2, can be emphasized.

  The compression unit 183 compresses the data so that the image data edge-enhanced by the edge enhancement unit 182 has the same size as the denomination candidate basic template (stored in the template storage unit 20). The luminance conversion unit 184 reduces the influence of variations in the imaging device 127 (see FIG. 3), changes in illumination conditions, and the like, and enables recognition under the same conditions, so that the luminance value of each pixel is based on the luminance conversion template. Change brightness.

  The template matching unit 19 performs rotation matching using the image data image-processed by the matching image processing unit 18 and the matching basic template. Specifically, the pattern matching value (collation value) at each rotation angle is calculated for both the front and back basic templates of the denomination candidate. Then, the template matching unit 19 performs denomination determination and authenticity determination based on the obtained pattern matching value.

  The template storage unit 20 stores a basic template for each denomination so that the template collation unit 19 can determine the denomination (including authenticity determination). For each denomination, basic templates on the front and back sides of the coin 2 are stored.

  The correctness determination unit 21 determines the correctness of the coin 2 when it is determined that the coin 2 is genuine as a result of the authenticity determination in the template matching unit 19. The fitness discriminating unit 21 recognizes a coin based on the computation value computed by the fitness discriminating computation unit 17 and the pattern matching value obtained as the pattern matching value at the best rotational position by the template matching unit 19. 2 is determined. A specific discrimination method will be described later.

  The threshold value storage unit 22 for discriminating damages stores a plurality of types of threshold values necessary for discriminating the fitness of the coin 2 for each denomination and further separately on the front and back of the coin 2. In addition, a plurality of types of threshold values determined in advance for each denomination described above are further stored in a plurality of stages so that the user can use different discrimination levels.

(Operation of coin damage detection device)
Next, an example of the operation of the coin damage determination apparatus 1 of the present embodiment configured as described above will be described with reference to the flowchart of FIG. FIG. 5 is a flowchart showing a fitness determination procedure in the coin fitness determination apparatus 1 of the present embodiment.

  The coin 2 conveyed through the conveyance path 3 included in the coin damage determination apparatus 1 according to the present embodiment is conveyed with the front or the back as an upper surface, passes through the hole sensor 111 and the material sensor 112 in this order, and is measured by each sensor. Is performed (step S1). The presence / absence of a hole in the coin 2 conveyed by the hole sensor 111 is measured, and the magnetic characteristic of the coin 2 conveyed by the material sensor 112 is measured. The measured result is output to the denomination identifying unit 16.

  When the measurement results in the hole sensor 111 and the material sensor 112 are input, the denomination identifying unit 16 identifies the denomination of the coin 2 conveyed through the conveyance path 3 (step S2). It should be noted that the denomination identified at this stage is a denomination candidate (which can be said to be a primary determination in this sense), and is not a determination of the denomination (secondary determination). The denomination identifying operation in the denomination identifying unit 16 will be described with reference to the flowchart shown in FIG. FIG. 6 is a flowchart illustrating an example of a denomination identification procedure in the denomination identification unit 16 included in the coin damage determination apparatus 1 according to the present embodiment.

  The denomination identifying unit 16 first checks whether or not there is a hole in the coin 2 (step S101). The confirmation of the presence or absence of the hole in the coin 2 uses the measurement result of the hole sensor 111. When it is determined that the coin 2 has a hole (Yes at Step S101), it is confirmed from the measurement result of the material sensor 112 whether or not the coin 2 can be determined to be 5 yen (Step S102).

  If it can be determined from the material of the coin 2 that the coin is a 5-yen coin (Yes at step S102), the target coin 2 is identified as a 5-yen coin (step S103), and the denomination of the denomination is terminated.

  On the other hand, when the material of the coin 2 is not a 5-yen coin (No at Step S102), the target coin 2 is identified as a 50-yen coin (Step S104), and the denomination is terminated. In this case, it may be confirmed whether the coin 2 is a 50-yen coin from the result measured by the material sensor 111.

  When it is determined that there is no hole in the coin 2 (No in step S101), it is confirmed from the measurement result of the material sensor 112 whether or not the coin 2 can be determined as 500 coins (step S105). When it can be determined from the material of the coin 2 that the coin is a 500 yen coin (Yes at Step S105), the target coin 2 is identified as a 500 yen coin (Step S106), and the identification of the denomination is terminated.

  When it is determined from the measurement result of the material sensor 112 that the coin 2 is not a 500 yen coin (No at Step S105), it is confirmed from the measurement result of the material sensor 112 whether the coin 2 can be determined as a 100 yen coin. (Step S107). If it can be determined from the material of the coin 2 that it is a 100 yen coin (Yes at Step S107), the target coin 2 is identified as a 100 yen coin (Step S108), and the identification of the denomination is terminated.

  If it is determined from the measurement result of the material sensor 112 that the coin 2 is not a 100 yen coin (No at Step S107), it is confirmed from the measurement result of the material sensor 112 whether the coin 2 can be determined as a 10 yen coin. (Step S109). If it can be determined from the material of the coin 2 that it is a 10-yen coin (Yes at step S109), the target coin 2 is identified as a 10-yen coin (step S110), and the denomination is terminated. On the other hand, if the coin 2 cannot be determined from the material of the coin 2 (No at Step S109), the target coin 2 is identified as a 1-yen coin (Step S111), and the denomination is terminated. In this case, it may be confirmed whether the coin 2 is a one-yen coin from the result measured by the material sensor 111.

  The coin 2 that has passed through the hole sensor 111 and the material sensor 112 then passes through the imaging unit 12. The imaging unit 12 images the coin 2 based on the timing signal from the timing sensor 125 (see FIG. 3), and captures the entire image of the coin 2 (step S3). The captured image data is converted into one of 256 gradation luminance values for each pixel by the A / D conversion unit 13 and stored in the memory 14. Note that the image data of the coin 2 stored in the memory 14 has a data size of 256 × 240.

  Next, based on the image data stored in the memory 14, a brightness histogram is created by the brightness histogram creation unit 151 included in the outer diameter / center position detection unit 15 (step S4). The purpose of creating the luminance histogram here is to detect the outer diameter and the center position of the coin 2 conveyed on the conveyance path 3, and as described above, there are two types of luminance histograms of the vertical histogram and the horizontal histogram. Created.

  Hereinafter, an example of creating two types of luminance histograms will be described with reference to FIGS. 7A and 7B. FIG. 7A is a diagram for explaining a method of creating a horizontal histogram from the image image of the coin 2 and detecting the center position of the coin 2 in the horizontal direction. FIG. 7B is a diagram for explaining a method of creating a vertical histogram from the image image of the coin 2 and detecting the center position of the coin 2 in the vertical direction.

  First, a method of creating a horizontal histogram (Hhist) will be described with reference to FIG. 7A. The luminance histogram creation unit 151 calculates the luminance value of each pixel for each pixel column in the vertical direction (direction parallel to the conveyance direction A of the coin 2) in the coin image window region WDh (the region hatched in FIG. 7A). The luminance frequency is calculated by summing up. In the window region WDh, there are 256 vertical pixel columns from the left edge to the right edge in FIG. 7A, and the luminance frequency is calculated for all these pixel columns. Then, a luminance histogram for each pixel column is arranged in the order in which the pixel columns in the vertical direction are arranged, and a horizontal histogram as shown in FIG. 7A is created. In FIG. 7A, a broken line is a guide surface of the width regulation guide 3a (see FIG. 4).

Note that the window region WDh is a striped region of the section H _T -H _B excluding the region in the vicinity of both ends so as not to be affected by the contact edge between coins that are continuously conveyed.

  The window region WDh corresponds to an example of the predetermined region in the present invention, and the vertical pixel column corresponds to the pixel column in the predetermined direction in the present invention.

  Next, a method for creating a vertical histogram (Vhist) will be described with reference to FIG. 7B. The luminance histogram creation unit 151 calculates the luminance value of each pixel for each pixel column in the horizontal direction (direction perpendicular to the conveyance direction A of the coins 2) in the coin image window region WDv (region hatched in FIG. 7B). The luminance frequency is calculated by summing up. Then, the luminance frequencies for each pixel column are arranged in the order in which the pixel columns in the horizontal direction are arranged to create a vertical histogram as shown in FIG. 7B.

The window area WDv is a stripe area hatched in FIG. 7B so as not to be affected by the reflection of the conveyor belt, the influence of other coins conveyed in a chained state, or the influence of other coin powder or dust. Is set to In FIG. 7B, V _L corresponds to a position where the guide surface of the width regulation guide 3a is present.

  When the horizontal histogram and the vertical histogram are created as described above, the outer diameter / center position detection unit 15 calculates the center position (Cx, Cy) and the outer diameter value D of the coin 2 in the horizontal direction and the vertical direction. (Step S5; see FIG. 5).

  In determining the center position Cx, the edge portion of the coin 2 is searched from the vicinity of the center toward both sides Xa and Xb using a horizontal histogram (see FIG. 7A). The search for the edge portion is performed by, for example, comparing the luminance frequency with the threshold value TH for each X coordinate value (pixel number value), and the coordinate immediately before the coordinate value (pixel number value) at which the luminance frequency is smaller than the threshold value TH. The value is the edge position. In FIG. 7A, XL and XR correspond to left and right edge positions, and in this case, the horizontal center position Cx is Cx = (XL + XR) / 2.

  Similarly, when obtaining the center position Cy, the edge portion of the coin 2 is searched from the vicinity of the center toward both sides Ya and Yb using a vertical histogram (see FIG. 7B). The search for the edge portion is determined in the same manner as when the center position in the horizontal direction is obtained. In FIG. 7B, YT and YB correspond to the upper and lower edge positions, and in this case, the center position Cy in the vertical direction is Cy = (YT + YB) / 2.

  Further, the outer diameter value D of the coin 2 is obtained by using the detected horizontal edge positions XL and XR. Specifically, the outer diameter value D is obtained from the difference value between the two edge positions XL and XR. The reason why the horizontal edge positions XL and XR are used without using the vertical edge positions YT and YB is to make it difficult to be affected by the reflection of the conveyor belt.

  When the outer diameter value and the center position are detected, the calculation by the damage discriminating operation unit 17 is performed, and the absolute difference value sum, average value, and division value are calculated (step S6; see FIG. 5). Details of the calculation performed by the fitness determination calculation unit 17 will be described with reference to FIG. FIG. 8 is a diagram for explaining the calculation performed by the damage determination calculation unit 17, and the histogram in the drawing is a horizontal histogram used for obtaining the outer diameter value and the center position of the coin 2.

  First, the absolute difference value sum obtained by the absolute difference value sum calculation unit 171 will be described. The absolute difference sum is adjacent to the horizontal histogram in a predetermined section (section indicated by an arrow in FIG. 8) of the horizontal histogram used to obtain the outer diameter value D and the center position (Cx, Cy) of the coin 2. The absolute values of the luminance frequency differences are obtained, and the absolute values of the obtained difference values are summed up. The absolute difference value sum is obtained as an index for measuring the degree of unevenness of the coin 2. For this reason, the area for which the sum of absolute difference values is to be obtained must be an area in the coin 2 as a matter of course.

  In the present embodiment, the area for calculating the sum of absolute difference values is a fixed area (the same pixel area) regardless of the denomination. For this purpose, the above-mentioned window region WDh and the predetermined section of the horizontal histogram are set so that the absolute difference value sum can be obtained for any denomination having various sizes. In setting the window area WDh, it is also considered that the influence of the contact edge between coins conveyed continuously as described above does not occur.

More specifically, the the a in the present embodiment the pixel number of the H _T window area WDh shown in FIG. 8 (vertical pixel number) 60, H pixel number _B (vertical pixel number) and 170. Further, the leftmost pixel number (horizontal pixel number) of the predetermined section (the section indicated by the arrow in FIG. 8) of the horizontal histogram is 100, and the rightmost pixel number (horizontal pixel number) is 180. This region corresponds to a region surrounded by a black frame (a thick thick solid line) in FIG. The pixel numbers in the horizontal direction are 1 for the leftmost pixel column and 256 for the rightmost pixel column of the image region (region surrounded by a black square). The pixel numbers in the vertical direction are 1 for the uppermost pixel column and 240 for the lowermost pixel column of the image area.

  The average value calculation unit 172 calculates the average value of the luminance frequencies in the same section (section indicated by an arrow in FIG. 8) where the absolute difference value sum of the horizontal histogram is calculated. Specifically, all the luminance frequencies in the section indicated by the arrow in FIG. 8 are summed, and 1 is added to the difference between the pixel numbers in the section indicated by the arrow (corresponding to 180-100 in this embodiment). Divide by the number to get the average value. The average value of the luminance frequency is obtained as an index for measuring the average brightness of the coin 2.

  As described above, the division value calculation unit 173 divides the absolute difference value sum obtained by the absolute difference value sum calculation unit 171 by the average value obtained by the average value calculation unit 172 to obtain a division value.

  As described above, the calculation performed by the damage determination calculation unit 17 uses a luminance histogram (specifically, a horizontal histogram) created to detect the outer diameter value D and the center position of the coin 2. . That is, since it is the structure which uses a brightness | luminance histogram for the several objective, it can be said that the coin damage determination apparatus 1 of this embodiment is performing the process using a coin image efficiently.

  Next, image processing by the matching image processing unit 18 is performed. The image processing by the matching image processing unit 18 will be described with reference to FIGS. FIG. 9 is a diagram for explaining a ring image cutout method by the ring cutout unit 181. FIG. 10 is a diagram showing a data array after cutting out a ring image.

  In the image processing by the collation image processing unit 18, first, the image of the coin 2 is cut out in a ring shape by the ring cutout unit 181 (step S 7; see FIG. 5). Specifically, referring to FIG. 9, center coordinates = (Cx, Cy), extraction base point (radial extraction start position from the central coordinates) = r, extraction width (radial extraction range) = Wy, A parameter such as cutout length (cutout range in the circumferential direction) = Wx is read from a storage unit (not shown). For the center coordinates, the values detected by the outer diameter / center position detector 15 are read, and for the others, the values stored in advance as the cutting parameters for each denomination are read.

  Then, based on the read parameters, a ring image is cut out by 360 degrees at a certain extraction angle with respect to the circular coin image. Each of the extracted ring image data (SDATA) is developed into two-dimensional striped sampling data (PDATA) having the rotation angle θ direction as the X axis and the radius R direction as the Y axis as shown in FIG.

  This expansion is performed by converting the polar coordinates into Euclidean coordinates using Equation 1 shown below, for example, where the center coordinates of the coin image are (Cx, Cy). When cutting out, differentiation processing is performed using a 3 × 3 pixel spatial filter during edge emphasis processing. Therefore, one extra pixel is used for the radius R direction between (r−1) and (r + Wy + 1). Cut out.

(Formula 1)
PDATA (i, j) = SDATA (Px + Cx, Py + Cy)
However, Px = ri × sin θj, Py = ri × cos θj
ri = i + (r−1), θj = j × 2π / Wx
i = 0 to (Wy-1), j = 0 to (Wx-1)

  As described above, when the ring cutout unit 181 converts the ring image data into the two-dimensional stripe data for a predetermined extraction angle, in this embodiment, the edge emphasizing unit 182 characterizes the pattern of the cutout image. Perform edge enhancement processing for emphasis. The edge enhancement processing is performed using 3 × 3 spatial filtering by a “Prewitt operator”, which is a known technique, for example.

  When the ring cutting and edge enhancement processing is performed, data compression is performed by the compression unit 183 so that the matrix has the same size as the basic template stored as the standard image data of the denomination (step S8; FIG. 5). reference). The compression rate in this case is determined for each denomination.

  Next, in order to reduce the influence of variations and the like of the imaging device 127 (see FIG. 3), the luminance conversion processing of the compressed data is performed using the luminance conversion template in which the standard luminance data is set. (Step S9). In the present embodiment, in the luminance conversion, the luminance value of each compressed pixel data is converted into the template luminance value in correspondence with the luminance order.

  As described above, when the image processing of the matching image processing unit 18 is performed, the template image that has been subjected to the image processing by the template matching unit 19 and the basic template image that is the standard image data of the denomination candidate coins. (Hereinafter referred to as a matching template) is executed while rotating the template (step S10). The matching template is a template stored in the template storage unit 20 for each denomination.

  FIG. 11 is a diagram for explaining the rotation collation process using the matching template. The calculation method of coincidence collation is performed by obtaining the luminance difference of the corresponding pixels over the entire area in the coin image data IDATA after the luminance conversion and the matching template MDATA. That is, the luminance difference between fn and tn corresponding to the two data is obtained over the entire area, and the total value Pm of the absolute values of the obtained luminance differences is calculated by the following equation 2, and at the rotation angle. The pattern matching value is used.

(Formula 2)
Pm = Σ | fn−tn |
However, n = 1 to the total number of pixels (= number of array data after compression)

  Then, the pattern matching value Pm is obtained for each rotation angle while relatively rotating the coin image data IDATA after the image processing and the matching template MDATA for one round. Specifically, in the present embodiment, the pattern matching value Pm is calculated while shifting one pixel (one column) in the x direction for the coin image data (two-dimensional array data) IDATA after image processing. Is repeated a number of times corresponding to one rotation.

  At this time, since the front and back of the coin 2 is not determined, the above processing is performed using two types of matching templates.

  The position where the value of the pattern matching value Pm obtained as described above is the minimum is the rotational position with the best matching degree with the matching template. The template collation unit 27 detects this best rotational position. Then, based on the pattern matching value at the best rotational position (for example, by comparing with a predetermined threshold value), the denomination and the front and back of the target coin 2 are determined, and the authenticity of the coin 2 is further determined. It discriminate | determines (step S11; refer FIG. 5).

  When the template collation unit 19 determines that the coin 2 is genuine (Yes at Step S11), the correctness determination unit 21 determines the correctness of the coin 2. The damage determination unit 21 first determines whether or not the target coin is a normal coin by using the calculation result of the damage determination calculation unit 17 (step S12).

  Here, the damage determination method using the calculation result of the damage determination calculation unit 17 will be described by taking a case of 100 yen coin (Japanese coin) as an example. FIG. 12 is a diagram showing an image (100 yen, front side) of a current currency (normal currency) and a horizontal histogram corresponding to the image. FIG. 13 is a diagram illustrating an image (100 yen, front side) of a dirty dirt coin stained black and a horizontal histogram corresponding to the image. FIG. 14 is a diagram showing an image (100 yen, front side) of a dirty coin that is stained white and a horizontal histogram corresponding to the image. The horizontal histogram creation conditions are the same as those described above (the conditions shown in FIG. 7A). Also, the absolute difference value sum and average value calculation conditions shown in the figure are the same as the above-described conditions (conditions shown in FIG. 8). Furthermore, the hatched portions in FIGS. 12 to 14 are originally black, but are shown with hatched lines to make the drawings easier to see.

  As shown in FIG. 12, the coins circulating as normal coins have a large sum (= 25000) of absolute difference values indicating the degree of unevenness of the coins because the coin pattern is clear. Further, the average value of the luminance frequency indicating the brightness of the coin is also a relatively large value (= 6800).

  On the other hand, for the coins that are black and dirty and classified as fouled coins as shown in FIG. 13, since the pattern of the coins is unclear, the sum of absolute difference values indicating the degree of unevenness of the coins is a small value (= 7800). Moreover, since it is black and is a dark image, the average value of the luminance power indicating the brightness of the coin is also a small value (= 4200).

  In addition, for the coins that are white and dirty and are classified as fouled coins as shown in FIG. 14, since the pattern of the coins is unclear, the absolute difference sum indicating the degree of unevenness of the coins is a small value (= 9900). It has become. Moreover, since it is white and is a bright image, the average value of the luminance frequency indicating the brightness of the coin is a large value (= 10000).

  Conventionally, a method for determining whether or not a coin is a fouled coin by the contrast of the coin has been performed. However, for example, in the case of a dirty soil that is white and dirty as shown in FIG. 14, it is not determined that it is a dirty coin based on the judgment based only on the light and darkness of the coin, and may be distributed in the market. For this reason, in the coin damage determination apparatus 1 of the present embodiment, the absolute difference value sum is introduced as a damage determination index, and the coin damage is determined based on the average value and the absolute difference value sum. .

  FIG. 15 calculates the absolute difference value sum and average value in a predetermined section (section indicated by an arrow in FIG. 8) of the horizontal histogram for 100 yen coins (front side) new coins, current currencies, and damaged coins. It is the graph which plotted the axis | shaft as an average value and the vertical axis | shaft as the absolute difference sum. Here, the new coin refers to a coin (new coin) before being distributed to the market. In addition, there are two types of fouled coins: one with black dirt that is severely classified as a fouled coin (indicated by a cross in the figure) and one that white dirt is severely classified as a fouled coin (indicated by a circle in the figure). Two types are used.

  As shown in FIG. 15, the new coin evaluated as the most beautiful coin is plotted on the upper right of the graph because it is glossy and the pattern is clear. That is, in order to discriminate only as clean coins as possible as normal coins, it is only necessary to set a threshold value so that only coins having a large sum of absolute difference values and average values become normal coins. The threshold value is stored in the damage determination threshold value storage unit 22.

  In FIG. 15, 14500 is a threshold value for comparison with the sum of absolute difference values (first threshold value) obtained by the fitness determination calculation unit 17, and 4200 is a threshold value for comparison with the average value (second threshold value). The case where it is set as is shown. In this case, a case where the sum of absolute difference values is less than 14500 or a case where the average value is less than 4200 is determined as a damaged coin. And when it sets in this way, it turns out that not only the coin which becomes dirty black but the coin which becomes dirty white can be discriminate | determined as a dirty coin.

  Coins with a clear balance between brightness and brightness, such as coins with a clear but uneven image and a dark image, or a bright image with an unclear or uneven coin, are judged as fouled coins. Tend to. In consideration of this point, a threshold (third threshold; the slope of the diagonal line in FIG. 15 is an example) is also set for the divided value obtained by dividing the sum of absolute difference values by the average value. It is also possible to adopt a configuration in which coins with a poor balance are removed more strictly (determined as dirty coins). For this reason, the third threshold may be set to determine the fitness. In this embodiment, the third threshold is also used to determine the fitness.

  Also, as shown in FIG. 16, the state of coins that are determined as fouled coins in the market cannot be classified simply as black or white, and is actually various. Which coin is determined as a damaged coin varies depending on the user, and the apparatus preferably has a plurality of determination criteria. Therefore, as shown in FIG. 16, it is preferable to take statistics about various coins and set the first to third threshold values so that a plurality of levels can be selected.

  In FIG. 16, if the level of the third threshold (the slope of the line shown diagonally in the graph corresponds) is increased, the new coin may be determined as a damaged coin. In such a case, the area in which the new coin is plotted is grasped in advance, and if the coin to be discriminated for damage is in the area of the new coin, it is determined as a normal coin. It may be configured.

  Here, FIG. 16 calculates the absolute difference value sum and the average value in a predetermined section of the horizontal histogram (section indicated by the arrow in FIG. 7) for 100 yen coins (front side) new coins, current currencies, and damaged coins. FIG. 16 is a graph in which the horizontal axis is plotted as an average value and the vertical axis is plotted as a sum of absolute difference values, and shows a case where the types of damaged coins are increased as compared with the case of FIG. 15.

  As described above, when it is determined that the money is normal by performing the damage determination using the calculation result of the damage determination calculating unit 17 (Yes in Step S12; see FIG. 5), the best rotation is further performed. Whether or not the coin is a normal coin is determined using the pattern matching value at the position (step S13). This determination is made by comparison with a threshold value (fourth threshold value) stored in the damage determination threshold value storage unit 22 for comparison with the pattern matching value. That is, for example, when the pattern matching value is smaller than the threshold value, it is determined as a normal coin, and when it is not, it is determined as a dirty coin.

  When it is determined that it is a normal coin using the pattern matching value at the best rotational position (Yes at Step S13), it is determined at the damage determination unit 21 as a normal coin (Step S14) and is a normal coin. Data is output. In this case, conveyance is controlled as normal coins.

  On the other hand, when it is determined that it is a dirty coin by the damage determination using the calculation result of the damage determining calculation unit 17 (No in step S12), and with the dirty coin using the pattern matching value at the best rotational position. When it is determined that there is (No at Step S13), the damage determination unit 21 determines that it is a dirty coin (Step S15), and data indicating that it is a dirty coin is output. For the coins that are determined to be damaged coins, for example, a process such as collection so as not to circulate thereafter is performed.

  When the template matching unit 19 determines that the coin is a fake (No at Step S11), it is determined that the coin is a fake coin (Step S16), and data indicating that the coin is a fake coin is output. Only, it is not possible to determine whether the coins are good or bad. Forged coins are also processed, for example, collected so as not to be distributed thereafter.

  As described above, in the coin damage determination apparatus 1 of the present embodiment, not only the calculation result of the damage determination calculation unit 17 but also the pattern matching value at the best rotational position is used to determine the coin damage. It has a configuration. With such a configuration, it is possible to determine the accuracy of the loss because the loss is determined for a wide range of coins. However, the coin damage determination using the calculation result of the damage determining calculation unit 17 can determine the coin damage with considerably higher accuracy than the conventional configuration. For this reason, it is also possible to adopt a configuration in which the correctness of the coin is determined using only the calculation result of the correctness determination calculation unit 17.

(Other)
The coin damage determination apparatus and the damage determination method according to the present invention are not limited to the embodiment described above. That is, various modifications can be made without departing from the object of the present invention.

  For example, in the above-described embodiment, the absolute difference value sum and the average value used for determining the correctness of coins are obtained using a horizontal histogram. However, the present invention is not limited to this configuration, and for example, a configuration using a vertical histogram may be used.

  Moreover, in embodiment shown above, it was set as the structure using the brightness | luminance histogram used in order to obtain | require the center position of a coin, when obtaining the absolute difference value sum and average value for discriminating the correctness of a coin. However, it is not necessarily limited to this configuration. Even if the center position of the coin is not detected, it is possible to determine the correctness of the coin of the present invention. Therefore, the coin damage determination device of the present invention may be configured not to include a coin center detecting means for detecting the center position of the coin. In such a case, the “predetermined region” set when creating the luminance histogram of the present invention can be configured differently from the present embodiment. That is, since the center of the coin is not obtained, the “predetermined area” can be set only to the area necessary for obtaining the sum of absolute difference values and the average value (for example, an area surrounded by a black frame in FIG. 8). . When the predetermined area is set as described above, the “predetermined section of the luminance histogram” for obtaining the sum of absolute difference values and the like is the entire section of the luminance histogram.

  In the embodiment described above, when the division value unit 173 obtains the division value, the absolute difference value sum obtained by the absolute difference value sum calculation unit 171 is divided by the average value obtained by the average value calculation unit 172. To obtain the configuration. However, it is not limited to this configuration. In other words, the average value obtained by the average value calculation unit 172 may be divided by the absolute difference value sum obtained by the absolute difference value sum calculation unit 171 to obtain a division value. In this case, it is necessary to change the third threshold value accordingly.

  Moreover, in embodiment shown above, it has the structure which discriminate | determines the damage of a coin using the absolute difference value sum and average value which are obtained from a horizontal histogram. However, as can be seen from FIG. 15, for example, a coin with a severe level of corruption can be identified as a dirty currency using only the sum of absolute difference values. For this reason, it is good also as a structure which discriminate | determines the damage of a coin only using the absolute difference value sum obtained from a horizontal histogram. However, the absolute difference value sum is an index that represents the degree of unevenness of a coin. For example, a coin that is stained black but has clear unevenness may be determined as a normal coin. For this reason, the structure which discriminate | determines the damage of a coin using the absolute difference value sum and average value which are obtained from a horizontal histogram like this embodiment is preferable.

  Further, in the embodiment described above, the denomination identifying unit 16 is provided to correspond to a plurality of denominations. However, for example, if the purpose is to discriminate only the correctness of a single coin, the denomination identifying unit 16 may be omitted.

  In the embodiment described above, the denomination identifying unit 16 is configured to identify the denomination using the measurement results of the hole sensor 111 and the material sensor 112, but is not limited to this configuration. For example, the outer diameter of the coin may be further obtained, and this may be added to identify the denomination. And it is good also as a structure calculated | required from a magnetic sensor about the outer diameter of a coin or a hole, for example, and it is good also as a structure calculated | required from image data.

  Further, in the above-described embodiment, the case where the object for which damage is determined is a Japanese coin has been described. However, the present invention is not limited to a Japanese coin, and is applicable to a coin of another country. Needless to say, you can.

  According to the coin damage determination apparatus and the damage determination method of the present invention, it is possible to accurately determine a dirty coin. For this reason, this invention is suitable with respect to the apparatus which needs to discriminate | determine the correctness of a coin.

Claims (10)

A coin damage determination device for determining whether a coin is a fouled coin,
A transport path through which the coins are transported;
Imaging means for imaging the coins conveyed in the conveyance path;
Data digitizing means for converting each pixel data obtained by the imaging means into digitized luminance values;
In a predetermined area of the coin image imaged by the imaging means, a luminance histogram creating means for calculating a luminance frequency by summing up the luminance values of each pixel for each pixel row in a predetermined direction, and creating a luminance histogram;
Absolute difference value sum calculating means for calculating an absolute difference value sum that is a sum of absolute values of differences between the luminance frequencies adjacent to each other in the luminance histogram in a predetermined section of the luminance histogram;
Threshold storage means for storing at least a first threshold determined in advance for comparison with at least the absolute difference value sum;
A fitness judging means for judging whether or not the coin is a dirty coin based on a comparison between the absolute difference value sum and the first threshold;
An apparatus for discriminating whether or not the coin is good or bad. An average value calculating means for calculating an average value of the luminance frequencies in the predetermined section of the luminance histogram;
The threshold value storage means further stores a second threshold value determined in advance for comparison with the average value,
The damage determination means determines whether or not the coin is a dirty coin based on a comparison between the absolute difference value sum and the first threshold value, and a comparison between the average value and the second threshold value. The coin damage determination device according to claim 1, wherein: A division value calculation means for calculating a division value obtained by dividing the absolute difference value sum by the average value, or a division value obtained by dividing the average value by the absolute difference value sum;
The threshold value storage means further stores a third threshold value determined in advance for comparison with the division value,
The fitness determining means is based on a comparison between the absolute difference value sum and the first threshold, a comparison between the average value and the second threshold, and a comparison between the division value and the third threshold. It is discriminate | determined whether the said coin is a dirty coin and the damage determination apparatus of the coin of Claim 2 characterized by the above-mentioned.   The threshold value storage means stores a plurality of levels of threshold values so that the user can change the setting level of whether or not the coin is a fouled coin. An apparatus for discriminating whether or not the coins are correct or not. A denomination identifying means for identifying the denomination of the coin conveyed through the conveyance path;
4. The coin damage determination device according to claim 1, wherein the threshold value storage means stores a different threshold value for each denomination. A coin center detecting means for detecting the center position of the coin;
4. The correct coin according to claim 1, wherein the brightness histogram created by the brightness histogram creating means is used for detecting the center position by the coin center detecting means. 5. Loss discrimination device. Image processing means for cutting out and processing a ring-shaped image based on the center position detected by the coin center detection means for the coin image obtained by the imaging means;
Authenticity determination means for determining whether or not the coin is genuine based on a collation value between an image obtained by the image processing means and a template prepared in advance;
The said damage determination means determines whether or not the coin is a damaged coin only when the authenticity determination means determines that the coin is genuine. Coins damage discrimination device. The threshold value storage means further stores a fourth threshold value previously determined for comparison with the verification value,
The damage determination means, apart from the determination using the absolute difference value sum, determines whether the coin is a dirty coin based on a comparison between the verification value and the fourth threshold value, The coin damage determination device according to claim 7, wherein if it is determined that at least one of the coins is a dirty coin, the coin is determined to be a dirty coin. A method for discriminating whether or not a coin is a damaged coin,
An imaging step of imaging the coins conveyed through the conveyance path;
A data digitization step of converting each pixel data obtained by the imaging step into a digitized luminance value;
In a predetermined area of the coin image captured in the imaging step, a luminance histogram creating step of calculating a luminance frequency by summing the luminance values of each pixel for each pixel row in a predetermined direction, and creating a luminance histogram;
An absolute difference value sum calculating step of calculating an absolute difference value sum that is a value obtained by summing up the absolute values of the differences between the luminance frequencies adjacent to each other in the luminance histogram in a predetermined section of the luminance histogram;
An average value calculating step of calculating an average value of the luminance frequencies in the predetermined section of the luminance histogram;
For each of the absolute difference value sum and the average value, a damage determination step of determining whether or not the coin is a dirty coin in comparison with a predetermined threshold value;
A method for discriminating whether or not a coin is good or bad. A coin center detecting step of detecting a center position of the coin,
The method for discriminating whether or not the coin is correct / incorrect according to claim 9, wherein the luminance histogram creating step is a step included in the coin center detecting step.

Images