JPH08212414A - Coin recognition device - Google Patents

Abstract

PURPOSE: To constitute the coin recognition device at low cost and to accurately recognize coins by analyzing binarized data obtained at a binarization part with reference data in a registered pattern storage part. CONSTITUTION: A coin 1 is conveyed on a conveyer path 2 as shown by A and the light of a light source 4 passed through a slit 5 is reflected by a half- mirror 6 to irradiate the reverse side of the coin 1. The reflected light from the irradiated coin 1 is read out as an analog image at a read part 7 and converted into digital image data at an image input part 8. The digital image data are stored in an image storage part 9 and respective image pixels of the digital image data are converted into binarized data at the binarization part 10. The binarized data obtained at the binarization part 10 are analyzed and compared with the reference data in the registered pattern storage part 12, and the decision part 11 decides the coin. Consequently, the device is constituted at low cost and the coin can correctly be recognized.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coin recognition device used for recognizing types and authenticity of coins and the like.

[0002]

2. Description of the Related Art Conventionally, many devices have been used for receiving and handling coins, and particularly in financial institutions, automated equipment such as an automatic transaction machine (ATM) has been used to automate the handling of deposits and savings and the transfer processing. Widely used. In such an automated device, when a customer inserts a coin, the type and authenticity of the coin is recognized.
This processing is performed by a coin recognition device, which is a coin discriminating unit in the automated equipment, and the following is known to perform the recognition.

For example, as described in Japanese Examined Patent Publication No. 63-67714, a device for recognizing by the outer diameter and material of the coin, and as described in Japanese Examined Patent Publication No. 3-63780, depending on the thickness of the coin. There is a recognition device. Further, as a device for detecting the outer diameter of a coin, there is, for example, one described in Japanese Patent Publication No. 2-48951, and as a device for detecting the thickness of a coin, Japanese Patent Publication No. 4-12.
No. 518 is available.

[0004]

However, in the conventional coin recognition device as described above, when the outer diameter and the material of the coins are similar to each other, for example, foreign coins of 500 won and 500 yen coins have outer diameters. And the material is almost similar, and the thickness is 10
Since the difference is only about 0 μm, in order to detect this difference, a thickness detection sensor or a complicated determination circuit is required, which causes a cost increase.

Further, since the detection characteristic of the material sensor changes due to the influence of temperature, an auxiliary circuit for correcting this is required, which also causes a cost increase. Therefore, there is a demand for a coin recognizing device that can accurately recognize coins at low cost.

[0006]

SUMMARY OF THE INVENTION The present invention provides a light source for illuminating the surface of a coin to be conveyed, a reading unit for reading the specularly reflected light of the surface of the coin as an analog image of the coin, and an analog image of the coin. An image capturing unit that captures and converts the digital image data, an image storage unit that stores the digital image data, a binarization unit that represents each pixel in the digital image data with binarized data, and to determine a coin A registration pattern storage unit that stores the reference data of, and a determination unit that analyzes the binarized data obtained by the binarization unit and compares the reference data of the registration pattern storage unit to determine a coin. It is characterized by that.

[0007]

[Function] The light source irradiates an uneven pattern forming a pattern on the coin surface, and when the irradiated portion is perpendicular to the irradiation light, the irradiation light is reflected by the coin surface as specular reflection light and read. Incident on the area. The reading unit outputs the incident light as an analog image to the image capturing unit, and the image capturing unit performs analog-to-digital conversion on the analog image to write it as digital image data in the image storage unit.

The binarizing unit binarizes the digital image data and outputs the binarized data to the judging unit. The determination unit analyzes the binarized data and compares it with the reference data stored in the registered pattern storage unit to determine the coin.

[0009]

Embodiments of the present invention will be described below with reference to the drawings. First Embodiment FIG. 1 is a configuration explanatory diagram of the first embodiment. Figure 1 (a)
Indicates a state in which the coin 1 being conveyed is viewed from above,
(B) shows the state in which the coin 1 is viewed from the side.

In the figure, reference numeral 2 is a conveying path provided with a slit-shaped window 3. The coin 1 is transported on the transport path 2 in the direction of arrow A in the figure, for example. Reference numeral 4 denotes a light source, which comprises a line light source such as a fluorescent lamp or a line filament lamp. The light from the light source 4 is transmitted through the slit 5
Used through. 6 indicates a half mirror,
It is installed so that the irradiation direction of the light from the slit 5 is perpendicular to the surface of the coin 1.

Reference numeral 7 denotes a reading unit, which is composed of a line image sensor such as a CCD, and receives specularly reflected light from the coin surface through the half mirror 6.
It is installed on the vertical axis of the coin surface. In each of the above-mentioned components, the light from the light source 4 is irradiated through the slit 5, and the irradiation direction by the half mirror 6 becomes the vertical axis direction of the coin surface, and the reading unit 7 is positioned on this vertical axis. And is arranged so as to receive the specularly reflected light from the coin surface. Further, the above-described components such as the window 3 and the reading unit 7 are set to a length equal to or longer than the diameter of the coin having the largest diameter among the coins to be recognized, and can scan all the coin types to be recognized. There is.

Reference numeral 8 is an image capturing unit, which is a reading unit 7.
Coin 1 which is analog image data read by
Data of the reflected light intensity is captured, A / D converted to create digital image data, and the digital image data is written and stored in the image storage unit 9 including, for example, a frame memory. The image capturing unit 8 can also extract the outer diameter of the coin 1 and its center point.

Reference numeral 10 denotes a binarization unit which sets a threshold value for binarization from the digital image data, and
By comparing this threshold value with the image data, the binarized image data is output. Reference numeral 11 denotes a determination unit, which compares the binarized data from the binarization unit 10 with the registered pattern which is the reference data stored in the registered pattern storage unit 12 to determine the denomination and true value of the coin 1. It has the function of judging false. The registration pattern storage unit 12 stores, for example, R
It is composed of a storage means such as an OM, a RAM, or a magnetic disk device.

Next, the operation of the coin recognition device having the above structure will be described. When the coin 1 is inserted into the coin recognition device, the light source 4 is turned on, and the light passing through the slit 5 is reflected by the half mirror 6 to illuminate the lower surface of the coin 1. The light specularly reflected from the coin surface passes through the half mirror 6 and enters the reading unit 7. Here, the uneven pattern on the surface forming the pattern of the surface of the coin 1 is clearly extracted as described below.

FIG. 2 is an explanatory diagram of the uneven pattern extraction, showing the uneven portion on the surface forming the pattern of the coin surface. In FIG. 1 (b), the light reflected by the half mirror 6 is shown to irradiate the lower surface of the coin from the bottom to the top, but in FIG. Is expressed. That is, in FIG. 2, the irradiation light 13 from the half mirror 6 is reflected on the surface of the coin 1 and reflected as the reflected light 14 from the coin 1.

First, in FIG. 2A, since the irradiation light 13 irradiates the flat portion of the coin 1, the reflected light 14a goes in the direction opposite to the direction of the irradiation light 13. As a result, the reflected light 14a passes through the half mirror 6 and enters the reading unit 7 as regular reflection light, as shown in FIG. On the other hand, in the case shown in FIG. 2B, since the irradiation light 13 irradiates the convex portion on the coin surface to the upper left in FIG. 2, the reflected light 14b moves in a direction different from the direction of the regular reflection light. Go to Therefore, the reflected light 14b does not go to the half mirror 6 and therefore does not enter the reading unit 7.

Further, in FIG. 2 (c), the irradiation light 13 irradiates a flat portion on the convex portion on the coin surface, so that the reflected light 14c is positive as in the case of FIG. 2 (a). The reflected light is transmitted through the half mirror 6 and is incident on the reading unit 7. Further, in the case shown in FIG. 2D, since the irradiation light 13 irradiates the lower left portion in FIG. 2 on the coin surface, the reflected light 1
4d goes in a direction different from the direction of the specularly reflected light. Therefore, the reflected light 14d does not enter the reading unit 7 as in the case of FIG. 2B.

In this way, the pattern on the surface of the coin 1 is
When the coin surface is flat, it becomes brighter due to the specular reflection light, while in the diagonally inclined portion, it becomes dark and appears.
The uneven pattern on the surface of can be read. FIG. 3 is an explanatory diagram of image data and conceptually shows digital image data obtained by the image capturing section.

In the figure, reference numeral 15 is digital image data. A white portion in the image data 15 is an image 16 of a coin, and the image 16 portion of the coin has the above-mentioned concavo-convex pattern forming a pattern on the coin surface. From the image data 15, the outer diameter and the center point position of the coin 1 are extracted. The x-axis and the y-axis are shown in the figure, and the end point coordinates x _s and x _{e of} the coin 1 are obtained corresponding to the x axis, and the end point coordinates y _s and y of the coin 1 are obtained corresponding to the y axis. _e
Is required. Further, from the end point coordinates to the center point coordinates (x
_c , y _c ) is calculated. The outer diameter data and the center point data are output to the determination unit 11.

The binarizing unit 10 firstly includes the image capturing unit 8
The image data obtained by the above is taken out, for example, by reading it from the image storage unit 9, and the density occurrence frequency distribution representing the gray value is extracted. In order to binarize the image of the coin 1, it is advisable to determine an appropriate threshold value so as not to be obsessed with fluctuations in output due to dirt, rust, and the like, and fluctuations in the reading unit 7 over time. For example, by calculating the average value of the concentration distribution and determining the optimum threshold value based on this.

FIG. 4 is an explanatory diagram of threshold value determination. The horizontal axis of the figure shows the density of the target image, and the vertical axis shows the frequency of occurrence of this density. The figure shows the case where the average value of the densities is set as the threshold value 17. However, in addition to this, for example, when the distribution shape of the density occurrence frequency has two peaks, the portion which becomes the valley of the peak is set as the threshold value. There are ways to decide. From these, it is advisable to select the method most suitable for the property of the target coin and determine the threshold value. The binarized data binarized based on this threshold value is output to the determination unit 11.

The determination unit 11 first reads out the binarized data output from the binarization unit 10 based on the center point (x _c , y _c ) data output from the image capturing unit 8, and determines the center point. The maximum continuous length of black (white) pixels existing on the line is obtained by counting the black (white) pixels on the straight line, that is, the diameter line. FIG. 5 is an explanatory diagram of the maximum continuous length detection of pixels, and illustrates the above operation of the determination unit 11.

As shown in the figure, in the circle corresponding to the outer edge of the coin 1 in the binary image, one of the straight lines passing through the center of the circle is determined as the reference axis 18, and passing through the center of the circle, this reference A straight line 19 inclined by an angle θ from the axis 17 is designated, and a binary image on this straight line 19 is extracted. The horizontal axis 20 in the graph portion of the figure corresponds to the straight line 19. The vertical axis 21 corresponds to binary image output, where "1" is black and "
0 "is white.

The horizontal axis 20 has the coordinates of a straight line 19, and the longest continuous portion of the portion of the image on the horizontal axis 20 where black or white pixels are continuously generated is black. The maximum continuous length 22 of (white) is extracted. Next, the determination unit 11 changes the value of the tilt angle θ and extracts the maximum continuous length 22 for each θ value to obtain the maximum continuous length distribution.

FIG. 6 is a maximum continuous length distribution graph. In the graph, the horizontal axis 23 is the inclination angle θ and the vertical axis 24 is the maximum continuous length. 25 shows the maximum continuous length distribution for each inclination angle θ. Further, 26 indicates a peaked portion in the maximum continuous length distribution 25. Usually, since the coin 1 is a circular object, the directionality of the uneven pattern is not stable. For this reason, the maximum continuous length distributions obtained when scanning at different angles are different.

FIG. 7 is another maximum continuous length distribution graph.
As shown in the figure, even if the same coin 1 is used, the coin 1
The maximum continuous length distribution 27 having a distribution shape different from that of FIG. 6 is obtained depending on the angle of. 28 is the maximum continuous length distribution 2
7 shows the peaked portion. The maximum continuous length distributions shown in FIGS. 6 and 7 are compared by retaking the peaked portions as a reference.

FIG. 8 is a maximum continuous length distribution graph based on the peak value portion. In the distributions shown in FIGS. 6 and 7, each peak value portion is used as the reference 29,
This shows the same distribution as the maximum continuous length distribution 30 of FIG. In this way, by taking the maximum continuous length distribution with the peak value portion as a reference, it is possible to always extract stable data regardless of the directionality of the coin 1.

The determining unit 11 also determines the denomination of the coin based on the outer diameter data output from the image capturing unit 8. That is, using the outer diameter data of the coin 1, it is checked whether or not the coin 1 is a coin to be accepted. For example, domestic 500 yen, 100 yen, 50 yen, 10
When accepting yen, 5 yen, and 1 yen, determine which of these denominations the coin 1 corresponds to, and if it does not correspond to the outer diameter data of any denomination, select it. exclude.

When the denomination of the coin 1 is discriminated in this way, the discriminating unit 11 determines the outermost diameter among the reference distributions of some maximum continuous length distributions stored in the registered pattern storage unit 12 in advance. The reference distribution of the maximum continuous length distribution to which the data is approximated is extracted, and the registered pattern, which is the reference data, is compared with the maximum continuous length distribution of the coin 1 to determine the authenticity.

Therefore, even coins having similar outer diameters and thicknesses can be discriminated with high accuracy by comparing the data from the uneven pattern of the coin surface.
For example, even when foreign coins having similar outer diameters and thicknesses are thrown into a place where domestic coins are to be accepted, accurate determination can be performed and the coins can be excluded.

Further, since the image of the entire area of the coin surface requiring the recognition of the coin 1 is read and a certain statistic is extracted from the image to determine the binarization threshold value, the recognition target is determined. The discrimination data of will not be affected by local stains or rust on the surface of the coin, and stable recognition will always be possible. Furthermore, the feature that black or white pixels are distributed from the binarized image can be extracted regardless of the rotation angle of the coin, so that the coin 1 is always stable and highly accurate regardless of the insertion angle and the transport condition. Can be recognized.

Further, since the denomination is preliminarily assumed based on the outer diameter data of the target coin and the registered pattern corresponding to the denomination is selected as the reference data to perform the authenticity discrimination processing, it is unnecessary. Therefore, the processing time can be shortened by omitting the determination processing. Since the processing time is shortened in this way, it is possible to suppress variations in the coin transport speed.

The present invention is not limited to the device having the above-mentioned structure, and may have any structure as long as it can scan and analyze the coin surface. FIG. 9 is an explanatory diagram of another configuration example, and in this example, an optical lens 31 is used instead of the slit 5 of FIG. The optical lens 31 is composed of, for example, an aspherical lens, a self-fog lens, or the like, and its focal position is the coin surface and the reading unit 7.
Is the light receiving surface of. Therefore, when the light is converged by the optical lens 31, the light from the light source 4 is reflected by the half mirror 6 to irradiate the surface of the coin 1, and then the reading unit 7
And is analyzed and discriminated in the same manner as described above.

Second Embodiment FIG. 10 is a structural explanatory view of the second embodiment. 10A shows a state in which the coin 1 being conveyed is viewed from above, and FIG. 10B shows a state in which the coin 1 is viewed from the side. In the figure, 2 is a conveyance path, and a slit-shaped window 3
Is provided. The coin 1 is transported on the transport path 2 in the direction of arrow A in the figure, for example.

A light source 4 is composed of a line light source such as a fluorescent lamp or a line filament lamp. This light source 4
The light from is used through the slit 5. Reference numeral 6 denotes a half mirror, which is installed so that the irradiation direction of light from the slit 5 is perpendicular to the surface of the coin 1.
A reading unit 7 is composed of a line image sensor such as a CCD and is installed on the vertical axis of the coin surface so as to receive the specularly reflected light from the coin surface through the half mirror 6.

In each of the above-mentioned components, the light from the light source 4 is irradiated through the slit 5, the irradiation direction by the half mirror 6 becomes the vertical axis direction of the coin surface, and the reading unit 7 is arranged on this vertical axis. Is located so as to receive specularly reflected light from the coin surface. Reference numeral 8 denotes an image capturing unit, which captures data of the reflected light intensity of the coin 1, which is analog image data read by the reading unit 7, and A / D converts it to create digital image data.
This is written and stored in the image storage unit 9 including, for example, a frame memory. The image capturing unit 8 can also extract the outer diameter of the coin 1 and its center point.

Reference numeral 10 denotes a binarization unit which sets a threshold value for binarization from the digital image data, and
By comparing this threshold value with the image data, the binarized image data is output. Reference numeral 32 denotes a determination unit, which compares the binarized data from the binarization unit 10 with the registered pattern, which is the reference data stored in the registered pattern storage unit 33, and determines the denomination and true value of the coin 1. It has the function of judging false. The registration pattern storage unit 33 stores, for example, R
It is composed of a storage means such as an OM, a RAM, or a magnetic disk device.

FIG. 11 is a block diagram of the judging section and the registration pattern storing section, and the judging section 32 will be described in more detail with reference to this figure. A boundary line component extraction unit 34 is for extracting the white / black boundary line of the binarized image.
A compression unit 35 compresses the boundary line component image, which is a binary image output from the boundary line segment extraction unit 34, by a generally known compression method to reduce the data size, Get value image data. 36 is a collating unit,
The compressed binary image data is compared with the reference data stored in the registered pattern storage unit 33 to determine the denomination and authenticity of the coin.

As described above, the judging section 32 is composed of the boundary line component extracting section 34, the compressing section 35 and the collating section 36, and works together with the registration pattern storing section 33.
A coin recognition process will be performed. Next, the operation of the coin recognition device having the above configuration will be described. When the coin 1 is inserted into the coin recognition device, the light source 4 is turned on, and the light passing through the slit 5 is reflected by the half mirror 6 to illuminate the lower surface of the coin 1.

The light specularly reflected from the coin surface passes through the half mirror 6 and enters the reading unit 7. Here, the uneven pattern on the surface forming the pattern of the surface of the coin 1 is clearly extracted as described below. FIG. 2 is an explanatory diagram of the uneven pattern extraction, showing the uneven portion on the surface forming the pattern of the coin surface. In FIG. 1 (b), the light reflected by the half mirror 6 is shown to irradiate the lower surface of the coin from the bottom to the top, but in FIG. Is expressed. That is, in FIG. 2, the irradiation light 13 from the half mirror 6 is reflected on the surface of the coin 1 and reflected as the reflected light 14 from the coin 1.

First, in FIG. 2A, since the irradiation light 13 irradiates the flat portion of the coin 1, the reflected light 14a goes in the direction opposite to the direction of the irradiation light 13. As a result, the reflected light 14a passes through the half mirror 6 and enters the reading unit 7 as regular reflection light, as shown in FIG. On the other hand, in the case shown in FIG. 2B, since the irradiation light 13 irradiates the convex portion on the coin surface to the upper left in FIG. 2, the reflected light 14b is different from the direction of the regular reflection light. Head to. Therefore, the reflected light 14b does not go to the half mirror 6 and therefore does not enter the reading unit 7.

Further, in FIG. 2 (c), the irradiation light 13 irradiates a flat portion on the convex portion on the coin surface, so that the reflected light 14c is positive as in the case of FIG. 2 (a). The reflected light is transmitted through the half mirror 6 and is incident on the reading unit 7. Further, in the case shown in FIG. 2D, since the irradiation light 13 irradiates the lower left portion in FIG. 2 on the coin surface, the reflected light 1
4d goes in a direction different from the direction of the specularly reflected light. Therefore, the reflected light 14d does not enter the reading unit 7 as in the case of FIG. 2B.

In this way, the pattern on the surface of the coin 1 is
When the coin surface is flat, it becomes brighter due to the specular reflection light, while when it is slanted, it becomes a shadow and appears darker.
The uneven pattern on the surface of can be read. FIG. 3 is an explanatory diagram of image data and conceptually shows digital image data obtained by the image capturing section.

In the figure, 15 is the digital image data. A white portion of the image data 15 is an image 16 of the coin, and the image 16 portion of the coin is
There is the uneven pattern described above that forms the pattern on the coin surface.
From the image data 15, the outer diameter and the center point position of the coin 1 are extracted. The figure shows the x and y axes,
The end point coordinates x _s and x _{e of} the coin 1 are obtained corresponding to the x axis, and the end point coordinates y _s and y _{e of} the coin 1 are obtained corresponding to the y axis. In addition, center point coordinates (x _c , y _c ) and outer diameter data are calculated from the end point coordinates. The outer diameter data and the center point data are output to the determination unit 32.

The binarization unit 10 firstly includes the image capturing unit 8
The image data obtained by the above is taken out, for example, by reading it from the image storage unit 9, and the density occurrence frequency distribution representing the gray value is extracted. In order to binarize the image of the coin 1, it is advisable to determine an appropriate threshold value so as not to be obsessed with fluctuations in output due to dirt, rust, and the like, and fluctuations in the reading unit 7 over time. For example, by calculating the average value of the concentration distribution and determining the optimum threshold value based on this.

FIG. 4 is an explanatory diagram of threshold value determination. The horizontal axis of the figure shows the density of the target image, and the vertical axis shows the frequency of occurrence of this density. The figure shows the case where the average value of the densities is set as the threshold value 17. However, in addition to this, for example, when the distribution shape of the density occurrence frequency has two peaks, the portion which becomes the valley of the peak is set as the threshold value. There are ways to decide. From these, it is advisable to select the method most suitable for the property of the target coin and determine the threshold value. The binarized data binarized based on this threshold value is output to the determination unit 32.

In the judging section 32, the boundary line component extracting section 34 first extracts the white / black boundary line component image of the binarized data. FIG. 12 is an explanatory diagram of a template for extracting the boundary line component. Here, as an example, 3 ×
The template 37 composed of 3 is listed. This template 37 has a white pixel 38 and a black pixel 39.
The black border is configured to pass through the center of the template. If the data to be recognized matches any of the templates in the figure, it is determined that there is a white / black boundary line, and the black pixel is output corresponding to this, thereby generating the boundary line component. Get the image.

The compression unit 35 converts the boundary line component image, which is the binary image thus output, and applies a generally known compression method to reduce the data size. FIG. 13 is an explanatory diagram of an example of the compression method. In the figure, 40 is a boundary line component image output from the boundary line component extraction unit 34, and 41 is the output image 4 thereof.
A sub-region obtained by dividing 0 is shown. When compressing each of the x direction and the y direction in the drawing to, for example, 1/4, the size of this sub region is set to 4 × 4. If there is a predetermined number of black pixels, such as one or two, among the 16 pixels in this sub-region, a black image is output corresponding to this, and compression is performed. The predetermined number of pixels may be three or four, etc., and should be appropriately determined according to the recognition target.

Although the description has been given here by taking an example in which compression is performed to ¼ in each of the x direction and the y direction in the drawing,
This may be another compression rate, for example, 1/5 or 1/8. The higher the compression rate, the smaller the amount of data and the less the processing amount, and the higher the speed, but the accuracy of the determination decreases, so the optimal compression rate should be set according to the recognition target. .

The collating unit 36 first determines the denomination of the coin 1 based on the outer diameter data output from the image capturing unit. That is, using the outer diameter data of the coin 1, it is checked whether or not the coin 1 is a coin to be accepted. For example, domestic 500 yen, 100 yen, 50 yen, 10
When accepting yen, 5 yen, and 1 yen, determine which of these denominations the coin 1 corresponds to, and if it does not correspond to the outer diameter data of any denomination, select it. exclude.

Then, of the reference images stored in the registered pattern storage unit 33 in advance, the reference image having the closest outer diameter data value is taken out, compressed by the compression unit 35, and output. And the coins 1 are compared by matching as described below to determine the authenticity of the coin 1. FIG. 14 is a diagram for explaining the operation of the matching unit. In the figure, 42 indicates compressed binary image data which is an output image from the compression unit 35. Four
Reference numeral 3 is reference data, and the coin is rotated at a predetermined angle, and the corresponding reference data is stored in the registration pattern storage unit 3 respectively.
3 conceptually shows the stored state.

The matching unit 36 matches the image data 42 corresponding to the coin 1 with the reference data 43, and determines that the image data 42 is true when the matching rate is larger than a predetermined reference value. . The registration pattern storage unit 3
The reference image 43 stored in No. 3 has data prepared by rotating the front and back surfaces for each denomination by a predetermined step angle. For example, if this predetermined graduation angle is 5 degrees, then 1 for each 5 deg.
This means that the reference data is created by rotating the circumference.

The predetermined angle is not limited to 5 degrees, but may be 1
It may be 0 degree increments. Although the processing amount decreases as the notch angle increases, the recognition accuracy decreases conversely. Therefore, it is preferable to use the most appropriate notch angle according to the recognition target. As described above, since the uneven patterns of the target coin surface are compared, it is possible to perform highly accurate discrimination even for coins having similar outer diameters and thicknesses. For example, even when foreign coins having similar outer diameters and thicknesses are thrown into a place where domestic coins are to be accepted, accurate determination can be performed and the coins can be excluded.

Since the image of the entire area of the coin surface to be recognized by the coin 1 is read and a certain statistic is extracted from the image, the threshold value for binarization is determined. The discrimination data of the recognition target is not affected by local stains or rust on the coin surface, and stable recognition is always possible. Further, by compressing the binarized image, the image data size is reduced, the processing amount is reduced, and the processing time can be shortened. Since the processing time is shortened in this way, it is possible to suppress variations in the coin transport speed.

Further, by storing the reference image for all rotation angle regions for each fixed angle step in the registered pattern storage unit 33, complicated rotation correction at the time of recognition becomes unnecessary,
The authenticity can be accurately determined by a simple process. In addition, the denomination is assumed in advance based on the outer diameter data of the target coin, and the registered pattern corresponding to the denomination is selected as the reference data to perform the true / false determination process. Processing can be omitted and processing time can be shortened.

The present invention is not limited to the device having the above-mentioned structure, and may have any structure as long as the device can scan and analyze the coin surface. FIG. 15 is an explanatory diagram of another configuration example, and in this example, an optical lens 31 is used instead of the slit 5 of FIG. The optical lens 31 is composed of, for example, an aspherical lens or a self-fog lens, and its focal position is on the coin surface and the light receiving surface of the reading unit 7. Therefore, this optical lens 31
The light from the light source 4 is reflected by the half mirror 6 to illuminate the surface of the coin 1, and then is incident on the reading unit 7 to be analyzed and discriminated in the same manner as described above. become.

[0057]

As described above in detail, since the coin is determined by comparing the uneven pattern of the target coin surface with the reference data, the coins whose outer diameter and thickness are similar to each other are determined. Even in this case, there is an effect that it is possible to perform highly accurate recognition. Moreover, since accurate determination can be performed without using a material sensor, a thickness detection sensor, and other complicated circuits, it is possible to provide a highly accurate device at low cost.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a configuration of a first embodiment.

FIG. 2 is an explanatory diagram of extraction of uneven patterns.

FIG. 3 is an explanatory diagram of image data.

FIG. 4 is an explanatory diagram of threshold value determination.

FIG. 5 is an explanatory diagram of maximum continuous length detection of pixels.

[Figure 6] Maximum continuous length distribution graph

[Figure 7] Other maximum continuous length distribution graph

FIG. 8: Maximum continuous length distribution graph based on the peak value part

FIG. 9 is an explanatory diagram of another configuration example.

FIG. 10 is an explanatory diagram of the configuration of the second embodiment.

FIG. 11 is a block diagram of a determination unit and a registration pattern storage unit.

FIG. 12 is an explanatory diagram of a template for boundary line component extraction.

FIG. 13 is an explanatory diagram of an example of a compression method.

FIG. 14 is an operation explanatory diagram of the collating unit.

FIG. 15 is an explanatory diagram of another configuration example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 coin 4 light source 7 reading unit 8 image capturing unit 9 image storage unit 10 binarizing unit 11 determination unit of the first embodiment 12 registered pattern storage unit 32 of the first embodiment 32 determination unit of the second embodiment 33 second implementation Example registration pattern storage unit 34 boundary component extraction unit 35 compression unit 36 collation unit 41 sub-region

Claims (7)

[Claims] 1. A light source that illuminates the surface of a coin to be conveyed, a reading unit that reads the specularly reflected light from the surface of the coin as an analog image of the coin, and an analog image of the coin is captured and converted into digital image data. And an image storage unit for storing the digital image data, a binarization unit that represents each pixel in the digital image data with binarized data, and reference data for determining the coin. A registration pattern storage unit and a determination unit that analyzes the binarized data obtained by the binarization unit and compares the binarized data with reference data in the registration pattern storage unit to determine a coin. And a coin recognition device. 2. The image capturing unit according to claim 1,
The central point data of the coin is extracted, and the determination unit determines the portion corresponding to the diameter of the coin of the binarized data obtained by the binarization unit based on the central point data extracted by the image capturing unit. Analyze and extract the length of the longest continuous one of the two predetermined values as the maximum continuous length, and the maximum continuous length of each of the diameters that make a predetermined angle with the diameter. A coin recognition device for extracting coins to extract a maximum continuous length distribution, and comparing the maximum continuous length distribution with reference data stored in a registered pattern storage unit in advance to determine a coin. 3. The determination unit according to claim 1, wherein the determination unit detects a boundary between a pixel having one of two values and a pixel having the other value from the binarized data obtained by the binarization unit. A boundary line component extraction unit that extracts the boundary line component image, and a collation unit that compares the boundary line component image with reference data stored in advance in the registered pattern storage unit to make a determination. Coin recognition device. 4. The boundary component image obtained from the boundary component extracting unit according to claim 3, is divided into sub-regions, and in each sub-region, a pixel having one of binary values is When there is a predetermined number or more, one of the values is determined as a value representing the sub-region to extract the compressed binary image data, and a compression unit that outputs the compressed binary image data to the matching unit is provided, A coin recognizing device characterized in that a collating unit compares the compressed binary image data with reference data stored in a registered pattern storage unit in advance to make a determination. 5. The coin recognition device according to claim 3, wherein reference data corresponding to a plurality of coin rotation angles are registered in advance in a registration pattern storage unit. 6. The image capturing unit according to claim 1, claim 2, claim 3, claim 4 or claim 5, extracting the outer diameter data of the coin, and the determining unit based on the outer diameter data. A coin recognizing device characterized by discriminating the type of the coin, and discriminating the authenticity of the coin by comparing only the reference data corresponding to the type of the coin among the reference data in the registered pattern storage unit. 7. In claim 1, claim 2, claim 3, claim 4, claim 5, and claim 6, the binarizing unit sets a threshold value corresponding to the intensity of specular reflection light on the surface of the coin. A coin recognizing device which is determined and binarized based on the threshold.