JP2021012587A - Serial number reader, paper sheet processing device, and serial number reading method - Google Patents

Abstract

To provide a serial number reader capable of accurately reading a serial number even when the background of the serial number is colored, and to provide a paper sheet processing device and a serial number reading method.SOLUTION: A serial number reader for reading serial numbers of paper sheets is provided, comprising a character recognition unit configured to character-recognize a serial number on a paper sheet based on a brightness image of the paper sheet in HSV color space.SELECTED DRAWING: Figure 1

Description

The present invention relates to a serial number reading device, a paper leaf processing device, and a serial number reading method. More specifically, the present invention relates to a serial number reading device suitable for reading a serial number written on paper sheets, a paper leaf processing device, and a serial number reading method.

At present, in the paper leaf processing device, in the process of transporting the paper leaves, the banknote identification device is used to identify the type of paper leaf, correct / loss determination, authenticity determination, and the like. Of these, the authenticity determination of paper sheets is performed, for example, by analyzing the image of the paper sheets read by the optical line sensor. In this case, for example, the authenticity of the paper leaves is determined based on the arrangement of the serial numbers of the paper leaves and the number of digits.

As a method for reading the serial number of paper sheets, for example, in Patent Document 1, for each pixel from the image in the serial number region, each brightness of R, G, and B and a unique vector set for each denomination are obtained. A method of calculating a linearly combined value, selecting a pixel if the value is within the threshold range, and recognizing a character for an image consisting of the selected pixel is disclosed.

JP-A-2010-225013

High-resolution image data is required to accurately read the serial number of the paper leaf, and for this purpose, the paper leaf is irradiated with light of a specific color such as green light (G). It is conceivable to acquire a kind of reflected image with high resolution. However, in this case, depending on the background color of the serial number, the contrast between the serial number and its surroundings becomes small, and it may be difficult to recognize the serial number as characters. For example, when green light is used, if the serial number is printed in a dark color and the background color is a complementary color of green such as red or orange, the contrast deteriorates.

Further, in the method described in Patent Document 1, the threshold range is widened in order to extract the serial number of a dirty banknote such as a loss ticket, and as a result, the background pattern may also be extracted. In that case, it is still difficult to recognize the serial number. In particular, if colored stains due to pigments or the like occur in the serial number region, character recognition becomes difficult.

The present invention has been made in view of the above situation, and is a serial number reader, a paper leaf processing apparatus, and a description capable of reading a serial number with high accuracy even when the background of the serial number is colored. It is an object of the present invention to provide a number reading method.

In order to solve the above-mentioned problems and achieve the object, the present invention is a serial number reading device for reading serial numbers of paper leaves, and the paper leaves are based on a brightness image in the HSV color space of the paper leaves. It is characterized by including a character recognition unit that recognizes similar serial numbers.

Further, the present invention is characterized in that, in the above invention, the serial number reading device further includes a brightness image generation unit that converts an RGB image of the paper sheets to generate the brightness image.

Further, in the present invention, in the above invention, the character recognition unit determines the brightness image, the RGB image, the red image of the paper leaves, and the paper leaves, which are determined according to the type of the paper leaves. It is characterized in that the serial number of the paper leaf is recognized as a character based on either the green image or the blue image of the paper leaf.

Further, the present invention is characterized in that, in the above invention, the brightness image generation unit generates the brightness image in the serial number region including the serial number in the RGB image.

Further, the present invention is a paper leaf processing apparatus, characterized in that the serial number reading apparatus is provided.

Further, the present invention is a serial number reading method for reading a serial number of a paper leaf, and a character recognition step of recognizing the serial number of the paper leaf as a character based on a brightness image in the HSV color space of the paper leaf. It is characterized by including.

According to the serial number reader, the paper sheet processing apparatus, and the serial number reading method of the present invention, the serial number can be read with high accuracy even when the background of the serial number is colored.

It is a figure for demonstrating the outline of the serial number reading apparatus, the paper leaf processing apparatus, and the serial number reading method which concerns on Embodiment 1, and shows an example of the reflection image which imaged the serial number region of a banknote, (a). Indicates an RGB image, (b) indicates a G image, and (c) indicates a brightness image. It is a perspective schematic diagram which shows the appearance of the banknote processing apparatus which concerns on Embodiment 1. FIG. It is a block diagram explaining the structure of the banknote identification device included in the banknote processing device according to the first embodiment. FIG. 5 is a schematic cross-sectional view illustrating the configuration of an image acquisition unit included in the bill identification device according to the first embodiment. It is a flowchart which shows the processing procedure in the serial number reading apparatus, the paper leaf processing apparatus, and the serial number reading method which concerns on Embodiment 1.

Hereinafter, preferred embodiments of the serial number reading device, the paper leaf processing apparatus, and the serial number reading method according to the present invention will be described with reference to the drawings. As the paper leaves to be the subject of the present invention, various paper leaves such as banknotes, checks, gift certificates, bills, forms, securities, card-like media, etc. can be applied, but in the following, banknotes are targeted. The present invention will be described by exemplifying the apparatus and method described above. Further, in the following embodiment, a bill identification device having a function of the number reading device will be described as a preferred embodiment of the number reading device according to the present invention. The following description is an example of a serial number reading device, a paper leaf processing device, and a serial number reading method.

In the present specification, the reflected image means an image based on the intensity distribution of the light reflected by the paper leaves by irradiating the paper leaves with light. Further, the transmitted image means an image based on the intensity distribution of light transmitted through the paper sheets by irradiating the paper sheets with light.

<Outline of this embodiment>
First, the outline of the present embodiment will be described with reference to FIG. This embodiment is suitable when the background of each character of the serial number is color-printed, particularly when it is color-printed in a plurality of colors. Specifically, for example, as shown in FIG. 1A, the serial number SN is printed in a dark color, and the background around the serial number SN is a red region Rr on which a red-based pattern is printed and green. A case will be described in which a green region Rg on which the keynote pattern is printed and a blue region Rb on which the blue keynote pattern is printed are included. In this case, as shown in FIG. 1 (b), when a green image (G image) is used for character recognition of the serial number SN, the output of the red region Rr becomes low, and the character "AB" of the serial number SN and its character. The contrast with the surrounding red region Rr becomes small, and it becomes difficult to recognize the character "AB". Therefore, it is difficult to recognize the entire serial number SN as characters. Similarly, when a red image (R image) or a blue image (B image) is used, the output of the green region Rg or the blue region Rb becomes low, and it is difficult to recognize the entire serial number SN as characters.

On the other hand, in the present embodiment, the RGB image (red / green / blue color image) as shown in FIG. 1 (a) is converted into the HSV space, and the brightness image as shown in FIG. 1 (c) ( V image) is generated, and the generated brightness image is used for character recognition of the serial number SN. When generating a brightness image, the maximum value of the output values (luminance values) of each RBG is set as the value of the pixel.
Therefore, as shown in FIG. 1C, the brightness image outputs the background around the serial number SN, that is, the red region Rr, the green region Rg, and the blue region Rb, rather than the green image, the red image, and the blue image. The image has a high value. As a result, in the brightness image, the contrast between each character of the serial number SN and their background becomes large, and each character of the serial number SN can be recognized with high accuracy.

If the background of the serial number is printed in a dark color, the contrast between the serial number and the background may be reduced depending on the brightness image. However, in general, even if the background of the serial number is a color-printed banknote, it is necessary for a person to be able to see the serial number, so the difference in brightness between the serial number and the background is large to some extent. Usually, the serial numbers are relatively dark and the background is relatively light. Therefore, it can be said that the brightness image is usually useful for all banknotes in which the background of the serial number is printed in color.

<Structure of banknote processing device>
The configuration of the banknote processing apparatus according to the present embodiment will be described with reference to FIG. The banknote processing device according to the present embodiment may have, for example, the configuration shown in FIG. The bill processing device 300 shown in FIG. 2 is a small bill processing device installed and used on a table, and includes a bill identification device (not shown in FIG. 2) that performs bill identification processing and a plurality of processing targets. When the hopper 301 on which the banknotes are placed in a laminated body and the banknotes drawn out from the hopper 301 into the housing 310 are rejected banknotes such as counterfeit banknotes and uncertain authenticity tickets, the rejected banknotes are discharged. Two reject units 302, an operation unit 303 for inputting instructions from the operator, and four for sorting and accumulating banknotes whose denomination, authenticity, and correctness are identified in the housing 310. It includes a stacking unit 306a to 306d and a display unit 305 for displaying information such as a bill identification count result and an accumulation status of each of the stacking units 306a to 306d. Of the four collection units 306a to 306d, the collection units 306a to 306c store the regular tickets, and the collection unit 306d stores the dirty tickets, based on the result of the positive / loss determination by the bill identification device. The method of distributing banknotes to the accumulation units 306a to 306d can be arbitrarily set.

<Structure of bill identification device>
The configuration of the bill identification device (serial number reading device) included in the bill processing device according to the present embodiment will be described with reference to FIG. As shown in FIG. 3, the bill identification device 1 included in the bill processing device according to the present embodiment includes a control unit 10, a detection unit 20, and a storage unit 30.

The control unit 10 is composed of, for example, a program for realizing various processes stored in the storage unit 30, a CPU (Central Processing Unit) that executes the program, various hardware controlled by the CPU, and the like. Has been done. The control unit 10 controls each unit of the bill identification device 1 based on the signal output from each unit of the bill identification device 1 and the control signal from the control unit 10 according to the program stored in the storage unit 30. Further, the control unit 10 has the functions of the identification unit 11, the brightness image generation unit 12, the image determination unit 13, and the character recognition unit 14 by the program stored in the storage unit 30.

The detection unit 20 includes an image acquisition unit 21, a magnetic detection unit 22, and a UV detection unit 23. The image acquisition unit 21 acquires an image of a banknote. The magnetic detector 22 includes a magnetic sensor (not shown) that measures magnetism, and detects the magnetism of magnetic ink, security threads, etc. printed on banknotes by the magnetic sensor. The magnetic sensor is a magnetic line sensor in which a plurality of magnetic detection elements are arranged in a line. The UV detection unit 23 includes an ultraviolet irradiation unit (not shown) and a light receiving unit (not shown), and receives fluorescence generated when the banknotes are irradiated with ultraviolet rays by the ultraviolet irradiation unit and ultraviolet rays transmitted through the banknotes. Detected by.

The configuration of the image acquisition unit 21 will be described with reference to FIG. As shown in FIG. 4, the image acquisition unit 21 includes optical line sensors 110 and 120 arranged so as to face each other. A gap for transporting the bill BN is formed between the optical line sensors 110 and 120, and this gap constitutes a part of the transport path 311 of the bill processing device. The optical line sensors 110 and 120 are located above and below the transport path 311 respectively.

The optical line sensor 110 includes two reflection light sources 111, a condenser lens 112, and a light receiving unit 113. Each reflection light source 111 has a predetermined wavelength of light (invisible light such as infrared light, monochromatic light such as red, green, blue, and white light visible on the main surface (hereinafter referred to as surface A) of the bill BN. Light). The condenser lens 112 collects the light emitted from the reflection light source 211 and reflected by the bill BN. The light receiving unit 113 includes a plurality of image pickup elements (light receiving elements, not shown) arranged in a line in a direction (main scanning direction) orthogonal to the transport direction (secondary scanning direction) of the bill BN, and collects light. The light collected by the lens 112 is converted into an electric signal. Then, after the electric signal is amplified, it is A / D converted into digital data and then output as an image signal.

The optical line sensor 120 includes two reflection light sources 121, a condenser lens 122, a light receiving unit 123, and a transmission light source 124. The reflection light source 121 and the transmission light source 124 have light of a predetermined wavelength (invisible light such as infrared light and monochromatic light such as red, green, and blue) on the main surface (hereinafter, B surface) of the bill BN, respectively. Or visible light such as white light). The condenser lens 122 collects the light emitted from the reflection light source 121 and reflected by the bill BN. The light receiving unit 123 includes a plurality of image pickup elements (light receiving elements, not shown) arranged in a line in a direction orthogonal to the transport direction of the banknote BN, and electrifies the light collected by the condensing lens 122. Convert to a signal. Then, after the electric signal is amplified, it is A / D converted into digital data and then output as an image signal.

The transmission light source 124 is arranged on the optical axis of the condenser lens 112 of the optical line sensor 110, and a part of the light emitted from the transmission light source 124 passes through the bill BN and is transmitted by the optical line sensor 110. It is condensed by the condensing lens 112 and detected by the light receiving unit 113.

Each of the light sources 111, 121, 124 may be a line-shaped light guide body (not shown) extending in a direction perpendicular to the paper surface (main scanning direction) of FIG. 4 and both end portions (one end portion) of the light guide body. ) Is provided with a plurality of light emitters (not shown). The light sources 111, 121, and 124 may not be provided with a light guide, but may be provided with a large number of light sources arranged in a direction perpendicular to the transport direction of the banknote BN. Each of the light sources 111, 121, and 124 includes, for example, one or more LED elements capable of irradiating light having a predetermined wavelength as a light emitting body.

The optical line sensors 110 and 120 each repeatedly take an image (exposure of the image sensor) on the banknote BN being conveyed in the conveying direction at regular time intervals, and output an image signal so that the image acquisition unit 21 can perform the image acquisition unit 21. , Acquire various images of the entire banknote BN (image acquisition step). Specifically, the image acquisition unit 21 acquires the reflection image of the A side of the bill BN and the transmission image of the bill BN based on the output signal of the optical line sensor 110, and based on the output signal of the optical line sensor 120. The reflection image of the B side of the bill BN is acquired. Further, the image acquisition unit 21 acquires at least an RGB image, a red image, a green image, and a blue image of each of the A side and the B side of the bill BN as a visible reflection image of the bill BN. Each image acquired by these image acquisition units 21 is stored in the storage unit 30 as image data 31.

In the optical line sensors 110 and 120, the reflection light sources 111 and 121 irradiate the bill BN with white light, and the light reflected by the bill BN is collectively emitted by an image sensor equipped with red, green, and blue color filters. By receiving light, an image signal for an RGB image may be output. On the other hand, in the optical line sensors 110 and 120, the reflection light sources 111 and 121 sequentially irradiate the bill BN with monochromatic light of red, green, and blue, and the light of each color reflected by the bill BN is an image sensor without a color filter. The image signal for an RGB image may be output by sequentially receiving light with.

The storage unit 30 is composed of a non-volatile storage device such as a semiconductor memory or a hard disk, and stores various programs and various data for controlling the bill identification device 1.

The identification unit 11 compares the feature pattern of the image of the banknote acquired by the image acquisition unit 21 with the denomination identification template, which is a feature pattern for each denomination of the banknote, and determines the denomination of the banknote.

The brightness image generation unit 12 converts the RGB image of the banknote stored in the storage unit 30 into an HSV color which is a color space composed of three components of hue (Hue), saturation (Saturation), and brightness (Value). Generate a brightness image in space. More specifically, an image is generated in which the maximum value of the output values (luminance values) of each RBG of the RGB image is the value of the pixel. For example, if the pixels are R = 200, G = 150, B = 20, the pixel value of R is 200, and if the pixels are R = 100, G = 150, B = 20, the pixel value of G is 150. If the pixels have R = 20, G = 150, and B = 200, 200 of B is set as the pixel value.

At this time, the brightness image generation unit 12 generates a brightness image in the serial number region (for example, a rectangular region) including the serial number of the RGB image. The storage unit 30 stores position information indicating the position of the serial number area in the RGB image of the banknote for each denomination of the banknote, and the brightness image generation unit 12 stores the banknote determined by the identification unit 11. Based on the denomination and the position information, the serial number area in the RGB image of the banknote is specified.

Further, the storage unit 30 stores type information indicating the type of the image used for character recognition of the serial number of the banknote for each denomination of the banknote, and the image determination unit 13 stores the type information indicating the type of the image. Based on the denomination of the bill determined by 11 and this type information, one image to be used for character recognition of the serial number is one of a brightness image, an RGB image, a red image, a green image, and a blue image. Select and determine the image to use for character recognition of the serial number. As for the type information, for each of the banknotes of all denominations to be processed, the above-mentioned five types of images are suitable for character recognition of the serial number according to the printing status of the serial number background in the serial number area. It is prepared in advance by deciding from among.

The character recognition unit 14 is based on the image of the banknote (either a brightness image, an RGB image, a red image, a green image, or a blue image) determined by the image determination unit 13 according to the denomination of the banknote. Character recognition (optical character recognition: OCR) is performed on the serial number of the bill. That is, when the brightness image is selected by the image determination unit 13, the character recognition unit 14 recognizes the serial number of the bill based on the brightness image generated by the brightness image generation unit 12. The storage unit 30 stores all kinds of characters that can be used for the serial number of the bill to be processed as a character image, and the character recognition unit 14 is within the serial number area of the image determined by the image determination unit 13. Each constituent character of the serial number is compared with a character image to identify each constituent character, and finally, the serial number is recognized by combining the specified characters.

<Processing flow by banknote processing device>
Next, with reference to FIG. 5, a processing flow by the bill processing apparatus according to the present embodiment, particularly a processing flow related to the serial number reading apparatus will be described. This process is repeated one by one for the bills carried into the bill identification device 1.

As shown in FIG. 5, first, the brightness image generation unit 12 converts the RGB image of the banknote to generate the brightness image (brightness image generation step S11). Next, the image determination unit 13 determines one image for character recognition of the serial number from the RGB image, the red image, the green image, and the blue image based on the denomination of the bill (image determination step). S12). Then, the character recognition unit 14 recognizes the serial number of the banknote based on the image determined by the image determination unit 13 (character recognition S13).

As described above, in the present embodiment, since the serial number of the bill is recognized as characters based on the brightness image in the HSV color space of the bill, the serial number is high even when the background of the serial number is colored. It can be read accurately.

Further, since the serial number reading device (banknote identification device) according to the present embodiment includes a brightness image generation unit 12 that converts an RGB image of a bill to generate a brightness image, it is common to acquire an RGB image of the bill. It is possible to generate a brightness image using a simple banknote processing device.

Further, in the present embodiment, the character recognition unit 14 determines the serial number of the bill based on any of the brightness image, the RGB image, the red image, the green image, and the blue image of the bill, which is determined according to the denomination of the bill. Since the character recognition is performed, an image suitable for character recognition can be used. Therefore, the reading accuracy of the serial number can be further improved.

Further, in the present embodiment, since the brightness image generation unit 12 generates the brightness image in the serial number region including the serial number of the RGB image, the image size of the brightness image can be reduced, and the brightness image can be reduced. It is possible to reduce the burden of processing and memory related to generation.

In the above embodiment, the image used for character recognition of the serial number is selected from one of a brightness image, an RGB image, a red image, a green image, and a blue image according to the denomination of the bill. However, the brightness image may be used for character recognition of the serial number regardless of the denomination of the bill. As a result, the processing and memory related to the image determination unit 13 can be omitted.

Further, in the above embodiment, the case where the brightness image is generated only in the serial number area has been described, but the brightness image may be generated in the entire RGB image of the banknote. As a result, the memory related to the position information of the serial number area can be omitted.

Although the embodiments of the present invention have been described above with reference to the drawings, the present invention is not limited to the above embodiments. Further, the configurations of the respective embodiments may be appropriately combined or modified as long as they do not deviate from the gist of the present invention.

As described above, the present invention is a technique useful for reading the serial number written on paper sheets.

1: Banknote acceptors (number reader)
10: Control unit 11: Identification unit 12: Brightness image generation unit 13: Image determination unit 14: Character recognition unit 20: Detection unit 21: Image acquisition unit 22: Magnetic detection unit 23: UV detection unit 30: Storage unit 31: Image Data 110, 120: Optical line sensors 111, 121: Reflection light source 112, 122: Condensing lens 113, 123: Light receiving unit 124: Transmission light source 300: Bill processing device 301: Hopper 302: Reject unit 303: Operation unit 305 : Display units 306a to 306d: Accumulation unit 310: Housing 311: Transport path BN: Bill SN: Serial number Rr: Red region Rg: Green region Rb: Blue region

Claims (6)

It is a serial number reader that reads the serial numbers of paper sheets.
A serial number reading device including a character recognition unit that recognizes a serial number of the paper leaf based on a brightness image in the HSV color space of the paper leaf. The serial number reading device according to claim 1, further comprising a brightness image generation unit that converts an RGB image of the paper sheets to generate the brightness image. The character recognition unit determines the brightness image, the RGB image, the red image of the paper leaf, the green image of the paper leaf, or the blue color of the paper leaf, which is determined according to the type of the paper leaf. The serial number reading device according to claim 2, wherein the serial number of the paper leaf is recognized as a character based on any of the images. The serial number reading device according to claim 2 or 3, wherein the brightness image generation unit generates the brightness image in a serial number region including the serial number in the RGB image. A paper leaf processing apparatus comprising the serial number reading apparatus according to any one of claims 1 to 4. It is a serial number reading method that reads the serial number of paper sheets.
A serial number reading method comprising a character recognition step of recognizing a serial number of the paper leaf based on a brightness image in the HSV color space of the paper leaf.

Images