JP7018129B2 - Paper leaf processing equipment and paper leaf processing method - Google Patents

Description

The present invention relates to a paper leaf processing apparatus and a paper leaf processing method. More specifically, the present invention relates to a paper leaf processing apparatus and a paper leaf processing method for performing processing such as authenticity determination on paper sheets such as banknotes and gift certificates printed using magnetic ink mixed with a fine magnetic substance. ..

Currently, in a paper leaf processing apparatus equipped with a paper leaf identification unit, identification of the type of paper leaf, correctness / loss determination, authenticity determination, serial number recognition, etc. are performed in the process of transporting the paper leaf. .. For example, by analyzing the image of the paper leaf read by the optical line sensor, the type of the paper leaf can be identified, the authenticity of the paper leaf can be determined, and the serial number written on the paper leaf can be recognized. It has been.

For example, Patent Document 1 discloses a paper leaf processing apparatus that recognizes characters of serial numbers by using an optical image acquired by an optical line sensor.

On the other hand, Patent Document 2 discloses a securities discrimination device that determines the authenticity of a banknote by a magnetic signal obtained from a banknote having a serial number written in magnetic ink.

Patent Document 3 discloses a high-resolution magnetic line sensor composed of 128 magnetic sensors and an imaging device using the magnetic line sensor. In this magnetic line sensor, the pitch in the width direction is 0. A large number of magnetic sensor elements are arranged at 5 mm.

Patent No. 5762830 Patent No. 5206159 Japanese Unexamined Patent Publication No. 2016-57125

However, when image recognition is performed using an optical image acquired by an optical line sensor, an accurate image cannot be acquired because an image of stains on paper leaves or graffiti can also be acquired, and the image cannot be recognized accurately. There is. In particular, when character recognition of a serial number (optical character recognition: OCR) is performed using an optical image acquired by an optical line sensor as in the paper sheet processing apparatus disclosed in Patent Document 1, the serial number to be recognized is configured. Since the numbers and letters to be printed are very thin lines, they may be erroneously recognized due to the influence of dirt, graffiti, background, etc., and the recognition rate of serial numbers may decrease. In addition, in paper sheets with serial numbers in multiple places, when detecting those serial numbers, if one of the serial numbers is detected as a different number from the other serial numbers due to dirt, etc., the OCR result alone will not be used. It is not possible to determine which is the correct number. Further, in the optical image, the information of the magnetic thread having the magnetic information is missing. Furthermore, it is necessary to detect both the front and back sides of the paper leaf, two opposing optical line sensors are required, and the optical line sensor occupies when trying to reliably detect the type and authenticity of the paper leaf. The ratio of price and space will be high.

On the other hand, as described in Patent Document 2, the serial number printed with magnetic ink is not easily affected by stains and the like. However, since the securities discrimination device disclosed in Patent Document 2 performs authenticity determination based only on the detection level (large amount of magnetic charge) of the detection signal measured by the magnetic sensor, the accuracy of authenticity determination is accurate. There is room for improvement.

Further, Patent Document 3 discloses a high-resolution magnetic line sensor and an imaging device using the magnetic line sensor, but does not disclose any technique related to image recognition.

The present invention has been made in view of the above situation, and provides a paper leaf processing apparatus and a paper leaf processing method capable of recognizing an image printed on paper leaves using magnetic ink with high accuracy. Is the purpose.

The present invention is a paper leaf processing apparatus that processes paper leaves printed using magnetic ink, and is a transport unit that transports the paper leaves one by one along a transport path, and a width of the transport path. A magnetic line sensor that is provided in a line shape in the direction and acquires the magnetic data of the conveyed paper sheets, and a magnetic image that generates a magnetic image that is an image based on the magnetic data of the paper sheets acquired by the magnetic line sensor. The generation unit, the storage unit that stores the reference magnetic image that is an image based on the magnetic data of genuine paper sheets, and the magnetic image generated by the magnetic image generation unit are recognized by comparing with the reference magnetic image. It is characterized by having an image recognition unit.

Further, in the above invention, the magnetic line sensor is characterized in that the resolution in the main scanning direction is 25 dpi (Dots Per Inch) or more and the resolution in the sub scanning direction is 50 dpi or more.

Further, in the present invention, in the above invention, the magnetic image generation unit binarizes an image based on magnetic data of paper sheets acquired by the magnetic line sensor to generate a binarized magnetic image as a magnetic image. The reference magnetic image having a digitization unit is a reference binarization magnetic image obtained by binarizing an image based on the magnetic data of genuine paper sheets, and the image recognition unit is the binarization unit. It is characterized in that the binarized magnetic image of the generated paper leaves is recognized by comparing it with the reference binarized magnetic image.

Further, in the present invention, in the above invention, the storage unit stores a plurality of character images used for serial numbers as the reference magnetic image and position information of serial numbers corresponding to the types of paper sheets. The image recognition unit is characterized in that the serial number written on the paper leaf is recognized by using the plurality of character images and the position information of the serial number.

Further, in the present invention, in the above invention, the storage unit stores a character image used for the amount of money as the reference magnetic image and the position information of the amount of money corresponding to the type of paper leaves, and the image recognition unit is used. , The character image and the position information of the amount of money are used to recognize the amount of money written on the paper leaf.

Further, in the present invention, in the above invention, the storage unit stores the symbol image described on the genuine paper leaf as the reference magnetic image and the position information of the symbol corresponding to the type of the paper leaf. The image recognition unit is characterized in that the symbol image and the position information of the symbol are used to recognize the symbol described on the paper leaf.

Further, in the present invention, in the above invention, the storage unit includes the magnetic pattern of the magnetic thread provided on the genuine paper leaf as the reference magnetic image, and the position information of the magnetic thread corresponding to the type of the paper leaf. The image recognition unit is characterized in that it recognizes the magnetic pattern of the magnetic thread provided on the paper sheet by using the magnetic pattern of the magnetic thread and the position information of the magnetic thread.

Further, in the present invention, in the above invention, the paper leaf processing apparatus is provided in a line shape in the width direction of the transport path, and has an optical line sensor for acquiring optical data of the paper leaves to be transported, and the optical optics. The storage unit further includes an optical image generation unit that generates an optical image that is an image based on the optical data of the paper leaves acquired by the line sensor, and the storage unit is an optical image that is an image based on the optical data of the genuine paper leaves. The reference image is further stored, and the image recognition unit generates the optical recognition result recognized by comparing the optical image generated by the optical image generation unit with the optical reference image, and the magnetic image generation unit. It is characterized in that image recognition is performed by complementing each other with the magnetic recognition results recognized by comparing the magnetic image with the reference magnetic image.

Further, in the present invention, when the image recognition unit cannot recognize at least a part of the optical image generated by the optical image generation unit, the corresponding magnetic image generated by the magnetic image generation unit is supported. It is characterized by adopting the magnetic recognition result of the place to be used.

Further, in the present invention, when the image recognition unit cannot recognize at least a part of the magnetic image generated by the magnetic image generation unit, the optical image generated by the optical image generation unit corresponds to the optical image generated by the optical image generation unit. It is characterized by adopting the optical recognition result of the part to be used.

Further, in the present invention, in the above invention, the paper leaf processing apparatus further includes a display unit, and the magnetic image generated by the magnetic image generation unit is displayed on the display unit so that the operator can visually recognize it. It is characterized by that.

Further, in the present invention, in the above invention, the paper leaf processing apparatus is provided in a line shape in the width direction of the transport path, and has an optical line sensor for acquiring optical data of the paper leaves to be transported, and the optics. It further includes an optical image generation unit and a display unit that generate an optical image that is an image based on optical data of paper sheets acquired by a line sensor, and a magnetic image generated by the magnetic image generation unit and the optics. The optical image generated by the image generation unit is displayed side by side on the display unit so that the operator can see it.

Further, the present invention is a paper leaf processing method for processing paper leaves printed using magnetic ink, and the magnetic line sensor provided in a line shape in the width direction of the transport path is used to connect the transport path to the transport path. The magnetic data acquisition step of acquiring the magnetic data of the paper sheets transported along the line, the magnetic image generation step of generating a magnetic image which is an image based on the magnetic data of the paper sheets acquired by the magnetic line sensor, and the above-mentioned It is characterized by comprising an image recognition step of recognizing a magnetic image generated by the magnetic image generation unit by comparing it with a reference magnetic image which is an image based on the magnetic data of genuine paper sheets.

According to the paper leaf processing apparatus and the paper leaf processing method of the present invention, an image printed on a paper leaf using magnetic ink can be recognized with high accuracy.

(A) is a perspective view showing the appearance of the banknote processing apparatus according to the first embodiment, and (b) is a sectional view showing the internal structure of the banknote processing apparatus according to the first embodiment. It is a plan view which shows an example of the banknote to be processed of the banknote processing apparatus which concerns on Embodiment 1, (a) shows the front surface, (b) shows the back surface. It is a functional block diagram of the bill identification part which concerns on Embodiment 1. FIG. It is sectional drawing of the magnetic line sensor which concerns on Embodiment 1. FIG. (A) to (d) are schematic diagrams for explaining the level conversion process by the magnetic image generation unit which concerns on Embodiment 1. It is a flowchart which shows the processing flow in the case of recognizing a serial number of a magnetic image character in Embodiment 1. FIG. It is a schematic diagram which showed an example of the magnetic image displayed on the display part in Embodiment 1. FIG. It is a schematic diagram which showed an example of the magnetic image displayed on the display part in Embodiment 1, and shows the case where a marking is attached to an abnormal part. It is a flowchart which shows the processing flow of the banknote processing method which concerns on Embodiment 1. It is a functional block diagram of the bill identification part which concerns on Embodiment 2. FIG. It is sectional drawing of the optical line sensor and the magnetic line sensor which concerns on Embodiment 2. FIG. It is a flowchart which shows the processing flow in the case of recognizing a serial number of an optical image in Embodiment 2. It is a schematic diagram for demonstrating the integration process of the optical recognition result of a serial number and the magnetic recognition result of a serial number in Embodiment 2. In the second embodiment, it is a flowchart which shows the flow of the process which integrates the optical recognition result of the serial number or the amount of money written on the banknote, and the magnetic recognition result of the serial number or the amount of money written on the banknote. A judgment parameter table that can be referred to in the integrated processing flow shown in FIG. 14 is shown. It is a schematic diagram which showed an example of the optical image and the magnetic image displayed on the display part in Embodiment 2. It is a flowchart which shows the processing flow of the banknote processing method which concerns on Embodiment 2.

Hereinafter, preferred embodiments of the paper leaf processing apparatus and the paper leaf processing method according to the present invention will be described with reference to the drawings. In the following, the present invention will be described by taking as an example a banknote processing device and a banknote processing method for banknotes. Various paper sheets such as securities and card-like media can be applied. For example, the present invention may be used to read magnetic ink characters on checks.

(Embodiment 1)
The configuration of the banknote processing apparatus according to the present embodiment will be described with reference to FIG. 1.
The banknote processing device 200 according to the present embodiment has, for example, the configurations shown in FIGS. 1 (a) and 1 (b). The banknote processing device 200 shown in FIGS. 1A and 1B has a hopper 210 capable of placing a plurality of banknotes, a payout unit 211 for feeding out the banknotes placed on the hopper 210 one by one, and a payout unit 211. A transport path 212 for transporting banknotes unwound from, a banknote identification unit 100A for identifying banknotes, a collection unit 213 for accumulating normal banknotes identified by the banknote identification unit 100A, and an abnormality that does not satisfy predetermined conditions. A reject unit 214 for accumulating various banknotes, a display unit 215 for displaying information input to the banknote processing device 200, a processing result, and the like, and a transport unit for transporting bills one by one along a transport path 212. , Equipped with. The transport unit has a plurality of transport means such as rollers and a drive device such as a motor that drives the transport means. The banknote processing device 200 further includes a transmission type or reflection type optical sensor at each position indicated by a triangular mark in the figure in order to detect a state of transporting banknotes in the device. From the detection results of these optical sensors, the shape and skew state of the bill may be estimated. By using the banknote identification unit 100A built in such a banknote processing device 200, a plurality of banknotes placed on the hopper 210 can be continuously processed into a fake ticket, a loss ticket, or an uncertain authenticity ticket. The determined banknotes can be returned to the reject unit 214 and sorted.

Here, the banknotes to be processed according to the present embodiment will be described.
The banknote to be processed may be one having at least one feature having magnetic information, but as shown in FIGS. 2A and 2B, it is printed with magnetic ink mixed with a fine magnetic substance. A banknote B having a serial number B1, an amount B2, a pattern (portrait or pattern) B3, and a magnetic thread B4 having magnetic information (preferably denomination information) is suitable.

The serial number is a serial number for distinguishing bills one by one, and is composed of characters such as phonetic characters and numbers, and includes information on the characters themselves and color information on those characters. And each banknote is distinguished by this information.

Further, the magnetic thread is a thin strip made of metal or resin, and is provided by being attached to or squeezed into the base material of a banknote. As the magnetic thread, for example, a bar code having a pattern in which magnetized portions and unmagnetized portions appear alternately along the length direction of the magnetic thread is preferably used. A plurality of magnetic threads may be provided in the bill, and the position and shape of the magnetic threads are not particularly limited.

Next, the internal configuration of the bill identification unit 100A will be described.
As shown in FIG. 3, the bill identification unit 100A amplifies the magnetic line sensor 110 that acquires the magnetic data (magnetic characteristics) of the bill transported through the transport path 212, and the output signal of the magnetic line sensor 110. A magnetic data processing unit 120 that performs processing such as A / D conversion, and a magnetic image generation unit 121 that generates an image (hereinafter referred to as a magnetic image) based on the magnetic data of the bill processed by the magnetic data processing unit 120. It has an image recognition unit 101 that performs image recognition processing on a magnetic image, an image output unit 102 that outputs a magnetic image to a display unit 215, and a storage unit 130 that stores a magnetic image or the like.

The magnetic image generation unit 121, the image recognition unit 101, and the image output unit 102 include, for example, a software program for realizing various processes, a CPU for executing the software program, and various hardware controlled by the CPU. , FPGA (Field Programgate Gate Array) and the like. A storage device 130, a separately provided volatile or non-volatile memory, a storage device such as a hard disk, or the like is used for storing software programs and data necessary for the operation of each unit.

As shown in FIG. 4, the magnetic line sensor 110 is arranged on one side of the transport path 212 in a line shape in the width direction of the transport path 212. The magnetic line sensor 110 is sufficiently long with respect to the width of the transport path 212, and detects the magnetic ink printed on the banknote and the magnetic information of the magnetic thread provided on the banknote on the entire surface of the banknote.

The magnetic line sensor 110 includes, in the magnetic head 111, a plurality of magnetic sensors 112 arranged in a line in a direction orthogonal to the transport direction of the bill, and a magnet 113 for generating a bias magnetic field, each of which is provided. The magnetic data acquired by the magnetic sensor 112 is output. That is, it outputs magnetic signals having the same number of channels as the number of magnetic sensors 112. Each magnetic sensor has at least one (preferably two) magnetic detection elements. As each magnetic detection element, a differential magnetic detection element that outputs a change in the magnetic flux density of the magnetic material as a voltage fluctuation is preferably used, and specific examples thereof include a magnetoresistive element (MR element). Examples of the MR element include an anisotropic magnetoresistive element (AMR element), a giant magnetoresistive element (GMR element), a tunnel magnetoresistive element (TMR element), and the like, but the TMR element is preferable from the viewpoint of sensitivity. On the opposite side of the magnetic line sensor 110 across the transport path 212, a hair roller 103 provided with a hair-like material on the outer peripheral surface is arranged so that the bill can be brought into close contact with the magnetic head 111. Since the magnetic line sensor 110 is transparent to the magnetic characteristics of the banknote, that is, can simultaneously detect the magnetic characteristics of the front surface, the back surface, and the inside of the banknote, the magnetic line sensor 110 can be arranged only on one side of the transport path 212.

The resolution of the magnetic line sensor 110 in the main scanning direction (direction orthogonal to the transport direction of bills) is set according to the pitch in the main scanning direction of the magnetic sensor 112, but is preferably 25 dpi or more, preferably 50 dpi. The above is more preferable. The upper limit of the resolution in the main scanning direction of the magnetic line sensor 110 is preferably 200 dpi, more preferably 100 dpi.

The resolution of the magnetic line sensor 110 in the sub-scanning direction (banknote transport direction) can be appropriately set according to the bill transport speed, but is preferably 50 dpi or more, and more preferably 100 dpi or more. The upper limit of the resolution in the sub-scanning direction of the magnetic line sensor 110 is preferably 200 dpi.

The magnetic image generation unit 121 generates a magnetic image from the magnetic data of banknotes acquired by the magnetic line sensor 110 and processed by the magnetic data processing unit 120. More specifically, the magnetic image generation unit 121 first performs a level conversion process for converting a differential waveform, which is an output signal of each magnetic sensor 112 (each channel), into a level waveform. In the level conversion process, first, as shown in FIG. 5A, the intermediate point of the differential waveform is calculated. Next, as shown in FIG. 5B, the peaks and bottoms within the predetermined noise cut upper limit and lower limit are replaced with intermediate points, and the negative value is multiplied by -1 to convert it to a positive value. Next, as shown in FIG. 5 (c), the waveform is shifted so that the intermediate point becomes 0 (V), and the adjacent peaks are interpolated. Then, as shown in FIG. 5D, the interpolated data is replaced with the data between the original peaks.

Then, the magnetic image generation unit 121 quantizes the signal converted into the level waveform of each channel between the maximum value and the minimum value to generate gradation data, and converts the signal into gradation data of all channels as a magnetic image. A grayscale image based on the image (hereinafter referred to as a light and shade magnetic image) is generated.

Further, the magnetic image generation unit 121 has a binarization unit 122 that binarizes a light and shade magnetic image to generate a binarized image (hereinafter referred to as a binarized magnetic image).

The magnetic image (shade magnetic image and binarized magnetic image) generated by the magnetic image generation unit 121 is stored in the storage unit 130. Further, a magnetic image (at least one of a light and shade magnetic image and a binarized magnetic image) is displayed on the display unit 215 so that the operator can see it.

The storage unit 130 is a storage device including a volatile or non-volatile memory, a hard disk, or the like. In addition to the magnetic image generated by the magnetic image generation unit 121, the storage unit 130 stores determination data 131 necessary for discriminating the denomination, authenticity, correctness / loss, and the like. The determination data 131 includes various templates 132 and various thresholds 133.

As the template 132, for example, a reference image for comparison with a magnetic image generated by the magnetic image generation unit 121 in order to identify a denomination, authenticity, positive loss, etc. (hereinafter referred to as a reference magnetic image). Etc. are preserved. As the threshold value 133, values for determining various feature quantities acquired from the banknote in order to discriminate the denomination, authenticity, correctness / loss, etc. of the banknote are stored. The predetermined template 132 and the predetermined threshold value 133 are prepared in advance for the denomination of the bill to be processed by the bill identification unit 100A.

The reference magnetic image is an image generated in advance from the magnetic data of genuine banknotes, and may be prepared by the banknote processing device 200 or may be prepared by another device. Further, as the reference magnetic image, the storage unit 130 includes a grayscale image based on the gradation data (hereinafter referred to as a reference shade magnetic image) and an image obtained by binarizing the reference shade magnetic image (hereinafter referred to as reference binarization). It is called a magnetic image.)

In addition, the storage unit 130 also stores setting data and the like in which various data measurement methods for identifying banknotes are set. The storage unit 130 is also used for storing banknote identification results and the like.

The image recognition unit 101 performs image recognition on the magnetic image generated by the magnetic image generation unit 121. This image recognition processing can be performed in the same manner as general image recognition processing (OCR, pattern matching, etc.) for an image based on optical data (hereinafter, optical image). That is, the image recognition unit 101 recognizes the magnetic image by comparing the magnetic image generated by the magnetic image generation unit 121 with the reference magnetic image stored in the storage unit 130, and various information contained in the magnetic image. To read.

The recognition target by the image recognition unit 101 is not particularly limited, but suitable ones include, for example, a serial number, an amount or a pattern printed on a banknote with magnetic ink, a magnetic thread to which magnetic information is attached, and the like. .. Since these magnetic information can be detected by the magnetic line sensor 110 through transmission (from one side of the bill), the image recognition unit 101 may perform the recognition process after inverting the magnetic image.

When recognizing the serial number, the storage unit 130 stores as a reference magnetic image all kinds of characters that can be used for the serial number of the bill to be processed as a character image, and for each denomination of the bill to be processed. , Stores the position information of the serial number indicating in which position of the bill the serial number is written. Then, the image recognition unit 101 first identifies a region in which the serial number is described in the magnetic image based on the position information. Subsequently, each constituent character of the serial number in the area is compared with the character image to identify each constituent character, and finally, the serial number is recognized by combining the specified characters.

When recognizing the amount of money, the storage unit 130 stores all kinds of characters (usually numbers) that can be used for the amount of money of the bill to be processed as a reference magnetic image as a character image, and the money of the bill to be processed. For each type, the position information of the amount indicating the position of the amount written on the bill is stored. Then, the image recognition unit 101 first identifies a region in which the amount of money is described in the magnetic image based on the position information. Subsequently, each constituent character of the amount of money in the area is compared with the character image to identify each constituent character, and finally, the amount of money is recognized by combining the specified characters. When recognizing the amount of money, the image recognition unit 101 uses the entire amount of money as one image instead of each constituent character of the amount of money as a reference magnetic image (a pattern image of all kinds of amount of money that can be used for the bill to be processed). May be matched with.

When recognizing a pattern such as a portrait or a pattern, the storage unit 130 stores all kinds of patterns that can be used for a genuine banknote as a reference magnetic image as a pattern image, and for each denomination of the banknote to be processed. , Stores the position information of the symbol indicating the position of the symbol on the bill. Then, the image recognition unit 101 first identifies a region in which the symbol is described in the magnetic image based on the position information. Subsequently, the symbol is recognized by comparing the symbol in the area with the symbol image.

When recognizing the magnetic pattern of the magnetic thread, the storage unit 130 stores the magnetic pattern of all kinds of magnetic threads that can be used for a genuine bill as a reference magnetic image as a magnetic pattern image, and also stores the magnetic pattern of the bill to be processed. For each denomination, the position information of the magnetic thread indicating the position of the magnetic thread on the bill is stored. Then, the image recognition unit 101 first identifies a region in which the magnetic thread is provided in the magnetic image based on the position information. Subsequently, the magnetic pattern of the magnetic thread is recognized by comparing the magnetic pattern in the region with the magnetic pattern image. When the magnetic pattern of the magnetic thread includes denomination information, the denomination of the banknote can be identified from the recognition result of the magnetic pattern of the magnetic thread.

The denomination identification by the magnetic pattern can be performed by the same method as the denomination identification by the optical image. That is, a magnetic image as shown in FIG. 7, which will be described later, is stored in the storage unit 130 as a reference magnetic image in four directions in which the front and back sides and orientations of the bills are different for each denomination, and the image recognition unit 101 stores the magnetic image in four directions. The denomination can be identified by collating the magnetic image with all reference magnetic images. Further, in the case of a magnetic thread, the output of the magnetic line sensor 110 changes greatly depending on the presence or absence of the magnetic thread. Therefore, the image recognition unit 101 determines that the position where the output is the largest in the magnetic image is the position of the magnetic thread, and the position thereof is determined. The denomination can also be specified by comparing the magnetic pattern of the position with the magnetic pattern image.

These recognition processes may be performed by comparing the light and shade magnetic image with the reference light and shade magnetic image, or by comparing the binarized magnetic image with the reference binarized magnetic image. However, from the viewpoint of recognition accuracy, it is preferable to perform the recognition process using a grayscale image for the serial number, the amount of money, and the pattern, and it is preferable to perform the recognition process using a binarized image for the magnetic thread. ..

The position information of these features is prepared in each of the four directions in which the front and back sides of the bill and the orientations are different from each other, and is stored in the storage unit 130.

Here, with reference to FIG. 6, a processing flow in the case of character recognition of the serial number of the magnetic image will be described.

First, a level conversion process for converting a differential waveform output from the magnetic line sensor 110 into a level waveform is performed to generate a magnetic image (level conversion step S11).

Next, the serial number area including the serial number character string is specified from the magnetic image and cut out (segment processing image cutout processing step S12).

Next, a character range is specified character by character in the cut-out serial number area (segment processing step S13). Next, a character image is cut out character by character from the serial number area (character recognition processing image cutting out processing step S14).

Then, character recognition processing of each cut out character image is performed (character recognition processing step S15).

When the serial numbers are described in a plurality of places on the banknote, the recognition results obtained in the plurality of serial number areas are integrated (recognition result integration processing step S16).

Further, from the result of image recognition, if there is no magnetic reaction at the place where the magnetic reaction should be, the image recognition unit 101 outputs the information of the place as the abnormal place information to the image output unit 102. On the contrary, when there is a magnetic reaction in a place where there should be no magnetic reaction, the information of the place is output to the image output unit 102 as the abnormal place information. That is, the image recognition unit 101 outputs the information of the place where there is a difference from the reference magnetic image in the magnetic image to the image output unit 102 as the abnormality place information.

The image output unit 102 outputs a magnetic image to the display unit 215. As a result, as shown in FIG. 7, a magnetic image is displayed on the display unit 215. Although the presence or absence of magnetism cannot be visually confirmed, by displaying the magnetic image on the display unit 215, it is possible to explain to the operator the reason why the bill was rejected.

Further, when the abnormal portion information is input from the image recognition unit 101, the image output unit 102 generates an image in which the abnormal portion of the magnetic image is marked based on the abnormal portion information, as shown in FIG. Then, it is output to the display unit 215. This makes it easier for the operator to understand the reason for rejection. FIG. 8 shows, as an example, a case where the serial number is printed with ordinary ink containing no magnetic substance, and the magnetic data of the serial number is not detected. In FIG. 8, a marking is attached to a place where the serial number should be in the magnetic image.

Next, the processing flow of the banknote processing method according to the present embodiment will be described with reference to FIG. 9.

First, the magnetic line sensor 110 acquires the magnetic data of the banknotes transported along the transport path 212 (magnetic data acquisition step S21).

Next, a light and shade magnetic image is generated as a magnetic image based on the magnetic data of the banknote acquired by the magnetic line sensor 110, and stored in the storage unit 130 (magnetic image generation step S22). The magnetic image may be related to the entire surface of the bill or a part of the bill.

Next, as another magnetic image, the light and shade magnetic image is binarized to generate a binarized magnetic image, which is stored in the storage unit 130 (binarized image generation step S23).

Next, a shade magnetic image or a binarized magnetic image is selected according to the recognition target, and the selected magnetic image is compared with the corresponding reference magnetic image (reference shade magnetic image or reference binarized magnetic image) to obtain an image. The recognition process is performed (image recognition step S24). The reference magnetic image is created in advance from the magnetic data of genuine paper sheets and stored in the storage unit 130. The reference magnetic image may be related to the entire surface of the banknote or a part of the banknote.

Next, a magnetic image (shading magnetic image or binarized magnetic image) with markings is displayed on the display unit 215 as needed (image display step S25). The displayed magnetic image may be related to the entire surface of the bill or a part of the bill.

The result of the image recognition is used in the bill identification process in the bill identification unit 100A. The content of the identification process is not particularly limited, and for example, in the case of a banknote, various functions such as identification of a denomination and determination of authenticity or correctness of the banknote can be mentioned. As described above, the recognition result of the magnetic pattern of the magnetic thread can be used for denomination identification. For authenticity judgment, the presence or absence of magnetic data (magnetic ink or magnetic thread) can be used from the result of image recognition. If there is magnetic data, it is judged as a genuine ticket, and if there is no magnetic data, it is judged as counterfeit. can. Further, it may be determined whether or not the image recognition is possible, and if the image recognition is possible, it may be determined as a genuine ticket, and if the image recognition is not possible, it may be determined as a forgery. This makes it possible to improve the detection accuracy of counterfeit tickets. Further, the result of character recognition of the serial number may be used for managing banknotes such as ensuring traceability.

As described above, in the present embodiment, a magnetic image is generated based on the magnetic data of the bill acquired by the magnetic line sensor 110, and the magnetic image is recognized by comparing with the reference magnetic image. The influence on the recognition process of graffiti, background, etc. can be reduced, and the recognition accuracy can be improved. This is because the main stains and graffiti are not accompanied by magnetic components. Further, the background of the magnetic feature such as the serial number printed with magnetic ink is usually not accompanied by a magnetic component.

Further, in the present embodiment, it is possible to perform the magnetic image recognition process only by arranging the high-resolution magnetic line sensor 110 on one side of the transport path 212. Therefore, it is not necessary to provide an optical line sensor, and the configuration of the bill identification unit 100A can be simplified as compared with the bill identification unit that has optical line sensors on both sides of the transport path and performs recognition processing by optical images. As a result, the degree of freedom in the layout of the banknote processing device 200 is increased, and the size of the banknote processing device 200 can be reduced.

Infrared light is an important factor in determining the authenticity of banknotes, but infrared and magnetic reactions can often be detected at the same location. Therefore, it is possible to perform authenticity determination without providing an optical line sensor.

(Embodiment 2)
In the present embodiment, the features peculiar to the present embodiment will be mainly described, and the description of the contents overlapping with the first embodiment will be omitted. Further, in the present embodiment and the first embodiment, members having the same or similar functions are designated by the same reference numerals, and the description of the members will be omitted in the present embodiment.

As shown in FIG. 10, in the present embodiment, the bill identification unit 100B uses the optical line sensor 140 for acquiring optical data (optical characteristics) of the bill transported on the transport path 212 and the output signals of the optical line sensor 140. On the other hand, an image (hereinafter referred to as an optical image) is generated based on the optical data of the bill processed by the optical data processing unit 170 that performs processing such as amplification processing and A / D conversion and the optical data processing unit 170. It further has an optical image generation unit 171.

The optical image generation unit 171 is composed of, for example, a software program for realizing various processes, a CPU for executing the software program, various hardware controlled by the CPU, a logical device such as FPGA, and the like. ing. A storage device 130, a separately provided volatile or non-volatile memory, a hard disk, or the like is used to store software programs and data necessary for the operation.

As shown in FIG. 11, the optical line sensor 140 is arranged in a line on the upstream side of the magnetic line sensor 110 (closer to the hopper 210) in the width direction of the transport path 212. The optical line sensor 140 is sufficiently long with respect to the width of the transport path 212, and detects the optical information of the banknote on the entire surface of the banknote.

The optical line sensor 140 has an upper unit 150 and a lower unit 160. The upper unit 150 and the lower unit 160 are located on opposite sides of the transport path 212, and acquire optical data of banknotes. Further, the optical line sensor 140 can acquire the reflection data based on the reflected light reflected by irradiating one surface of the banknote with light and the transmitted data based on the transmitted light transmitted by irradiating the banknote with light. can.

The upper unit 150 includes a light emitting unit 151 and a transparent member 155. Further, the lower unit 160 includes light emitting units 161a and 161b (hereinafter collectively referred to as light emitting units 161), a condenser lens 162, a light receiving unit 163, a light receiving unit substrate 164, and a transparent member 165.

The light sources of the light emitting units 151 and 161 are composed of a light guide body or an LED array. The light emitting unit 161 has a plurality of light sources having different peak wavelengths (for example, an infrared light (IR) light source, a red light (R) light source, a green light (G) light source, and a blue light (B) light source). There is. The light emitted from the light emitting unit 161 is applied to the bill through the transparent member 165, and the light reflected from the bill is collected by the condenser lens 162, received by the light receiving unit 163, and acquired by the light receiving unit 163. The data is transmitted to the optical data processing unit 170 by the light receiving unit substrate 164. The light receiving unit 163 is a line-shaped light receiving sensor extending in a direction perpendicular to the paper surface of FIG. 11, and a large number of pixel unit units are arranged in a line shape. Each pixel unit contains a plurality of light receiving elements corresponding to the number of light source types, and each light receiving element includes a bandpass filter that transmits only light in a predetermined wavelength band.

The light emitting unit 151 has a plurality of light sources having different peak wavelengths (for example, an infrared light (IR) light source and a green light (G) light source). The light emitted from the light emitting unit 151 irradiates the bill through the transparent member 155 made of transparent glass or resin, and the light transmitted through the bill enters the condenser lens 162 through the transparent member 165 made of transparent glass or resin. The light is collected by the condenser lens 162, received by the light receiving unit 163, and the data acquired by the light receiving unit 163 is transmitted to the optical data processing unit 170 by the light receiving unit substrate 164.

The configuration shown in FIG. 11 shows a very simple configuration, and the optical line sensor 140 is configured to be able to acquire reflection data and transmission data on both sides of a banknote in some cases. May be good. For example, instead of the upper unit 150, a unit similar to the lower unit 160 may be arranged, and reflection data may be acquired by the upper and lower units, respectively.

The optical image generation unit 171 generates a grayscale image (hereinafter referred to as a shade optical image) as an optical image from the optical data of the banknote acquired by the optical line sensor 140 and processed by the optical data processing unit 170. More specifically, an optical reflection image and an optical transmission image are generated from the reflection data and the transmission data, respectively.

Further, the optical image generation unit 171 has a binarization unit 172 that binarizes a shade optical image to generate a binarized image (hereinafter referred to as a binarized optical image).

The optical image (shading optical image and binarized optical image) generated by the optical image generation unit 171 is stored in the storage unit 130. Further, at least one of these optical images is displayed on the display unit 215 so that the operator can see it.

In the present embodiment, the storage unit 130 further stores the optical image generated by the optical image generation unit 171. Further, as the template 132, for example, a reference image (hereinafter referred to as a reference optical image) for comparison with an optical image generated by the optical image generation unit 171 in order to identify a denomination, truth, correctness, etc. It has been saved.

The reference optical image is an image generated in advance from the optical data of genuine banknotes, and may be prepared by the banknote processing device 200 or may be prepared by another device. Further, as the reference optical image, the storage unit 130 includes a grayscale image based on gradation data (hereinafter referred to as a reference shade optical image) and an image obtained by binarizing the reference shade optical image (hereinafter referred to as reference binarization). It is called an optical image.)

In the present embodiment, the image recognition unit 101 recognizes not only the magnetic image but also the optical image generated by the optical image generation unit 171. The image recognition process for the optical image can be performed in the same manner as the general image recognition process for the optical image. That is, the image recognition unit 101 recognizes the optical image by comparing the optical image generated by the optical image generation unit 171 with the reference optical image stored in the storage unit 130, and various information contained in the optical image. To read.

The object to be recognized by the image recognition unit 101 in the optical image is not particularly limited, but suitable examples include a serial number, an amount of money, a design, and the like. Since general methods such as OCR and pattern matching can be used for these image recognition, a detailed description thereof is omitted here, but as an example, a processing flow for recognizing a serial number of an optical image is described. It is shown in FIG.

As shown in FIG. 12, it is the same as the processing flow when the magnetic image is used except that there is no level conversion step.

That is, first, the serial number area including the serial number character string is specified from the optical image and cut out (segment processing image cutout processing step S31).

Next, a character range is specified character by character in the cut-out serial number area (segment processing step S32).

Next, a character image is cut out character by character from the serial number area (character recognition processing image cutting out processing step S33).

Then, character recognition processing of each cut out character image is performed (character recognition processing step S34).

When the serial number is described in a plurality of places on the banknote, the recognition results obtained in the plurality of serial number areas are integrated (recognition result integration process step S35).

When character recognition is performed based only on an optical image like a general bill identification unit, in particular, serial numbers written in a plurality of places on the bill are recognized based on the optical image, and as a result, they are recognized. When the serial numbers are recognized as different from each other, there is a possibility that one of them is erroneously recognized, but it cannot be determined from the result of character recognition by the optical image which one is erroneous. However, in the present embodiment, since character recognition by the magnetic image is also performed, it is possible to determine which of the result of the character recognition by the optical image is incorrect from the result of the character recognition by the magnetic image.

Further, in the present embodiment, the image recognition unit 101 mutually compares the optical recognition result recognized by comparing the optical image with the optical reference image and the magnetic recognition result recognized by comparing the magnetic image with the reference magnetic image. Performs image recognition by complementing. That is, when one of the recognition results includes an unread part, the unread part is supplemented with the corresponding part of the other recognition result. This makes it possible to improve the accuracy of image recognition.

For example, as shown in FIG. 13, even if an unreadable character is present in the optical recognition result of the serial number and a non-magnetic character is present in the magnetic recognition result of the serial number, both results are integrated. By doing so, the recognition rate can be improved. The same applies to the amount stated on the banknote.

Here, with reference to FIG. 14, an integrated processing flow of the optical recognition result of the serial number or the amount of money written on the banknote and the magnetic recognition result of the serial number or the amount of money written on the banknote will be described.

First, the character count n is reset to 0 (step S41).

Next, the character count n is incremented by one (step S42).

Next, from the magnetic recognition result of the nth character, it is determined whether or not there is magnetic data of the nth character (step S43).

If there is no magnetic data for the nth character, it is assumed that character recognition by the magnetic image of the nth character is impossible, and the result of character recognition by the optical image is adopted (step S44), and the nth character is confirmed (step S45). ).

When it is determined that the magnetic data of the nth character is present, it is determined whether or not an unreadable character is present in at least one of the optical recognition result and the magnetic recognition result of the nth character (step S46).

When there is no unreadable character in the nth character, it is determined whether or not the optical recognition result of the nth character and the magnetic recognition result of the nth character match each other (step S47), and if they match, they match. The character is determined as the nth character (step S45).

If they do not match, the determination parameter table shown in FIG. 15 is referred to (step S48). When the preset determination parameter is 1, the evaluation value in the optical recognition result is compared with the evaluation value in the magnetic recognition result, the result having the higher evaluation value is adopted, and the result is confirmed as the nth character. (Step S45). When the determination parameter is 2, the presence / absence of the flag of the optical recognition result and the presence / absence of the flag of the magnetic recognition result are determined, the result having the flag is adopted, and the result is determined as the nth character (step S45). .. When the determination parameter is 3, it is determined that the nth character is unrecognizable (step S45).

In step S46, when an unreadable character is present in the nth character, it is determined whether or not both the optical recognition result and the magnetic recognition result in the nth character are unreadable (step S49).

If none of them is unreadable, the result of the one that was not unreadable is adopted (step S50) and confirmed as the nth character (step S45).

If all of them are unreadable, it is determined that the nth character is unrecognizable (step S51), and the nth character is determined to be unrecognizable (step S45).

When the nth character is determined by the above step, it is determined whether or not the character count n matches the number of characters N to be recognized (step S52). Ends a series of processes.

In the present embodiment, the image output unit 102 outputs an optical image and a magnetic image to the display unit 215. As a result, as shown in FIG. 16, the optical image and the magnetic image are displayed side by side on the display unit 215.

Next, the processing flow of the banknote processing method according to the present embodiment will be described with reference to FIG.

First, the optical line sensor 140 acquires optical data of banknotes transported along the transport path 212 (optical data acquisition step S61).

Next, an optical image is generated based on the optical data of the banknote acquired by the optical line sensor 140 and stored in the storage unit 130 (optical image generation step S62). The optical image may be the entire surface of the banknote or a part of the banknote. Further, the optical image may be a shading optical image or a binarized optical image.

Next, the magnetic line sensor 110 acquires the magnetic data of the banknotes transported along the transport path 212 (magnetic data acquisition step S63).

Next, a magnetic image is generated based on the magnetic data of the banknote acquired by the magnetic line sensor 110 and stored in the storage unit 130 (magnetic image generation step S64). The magnetic image may be related to the entire surface of the banknote or a part of the banknote. Further, the magnetic image may be a shading magnetic image or a binarized magnetic image.

Next, the optical image recognition process is performed by comparing the optical image with the reference optical image, and the magnetic image recognition process is performed by comparing the magnetic image with the reference magnetic image (image recognition step S65). The reference optical image and the reference magnetic image are created in advance from the optical data and the magnetic data of genuine paper sheets, respectively, and stored in the storage unit 130. These reference images may be related to the entire surface of the bill or may be related to a part of the bill. Further, these reference images may be grayscale images (reference shade optical image and reference shade magnetic image), or images obtained by binarizing grayscale images (reference binarization optical image and reference binarization). It may be a magnetic image). However, if the recognition target is a grayscale image, image recognition is performed by comparing with the grayscale reference image, and if the recognition target is a binarized image, it is compared with the binarized reference image. Perform image recognition. Depending on the recognition target, the same reference binarized image may be used as the reference binarized optical image and the reference binarized magnetic image.

Next, the optical image and the magnetic image with markings, if necessary, are displayed side by side on the display unit 215 (image display step S66). The displayed optical image and magnetic image may be related to the entire surface of the bill or may be related to a part of the bill.

The result of the image recognition is used in the bill identification process in the bill identification unit 100B. The content of the identification process is not particularly limited, and for example, in the case of a banknote, various functions such as identification of a denomination and determination of authenticity or correctness of the banknote can be mentioned. In the present embodiment, the recognition result of the optical image and the recognition result of the magnetic image can be appropriately used together for each identification / determination process. This makes it possible to improve the accuracy of each identification / determination process.

As described above, in the present embodiment, the optical recognition result and the magnetic recognition result are complemented with each other to perform image recognition. More specifically, when at least a part of the optical image cannot be recognized, the magnetic recognition result of the corresponding part of the magnetic image is adopted, and when at least a part of the magnetic image cannot be recognized, the corresponding part of the optical image is adopted. The optical recognition result of is adopted. Therefore, the accuracy of image recognition can be improved. Here, at least a part of the optical image or the magnetic image may be, for example, a part of characters constituting a serial number, an amount of money, or the like as described above, or may be a part of a design.

In this embodiment, the image recognition based on the optical image and the image recognition based on the magnetic image may be changed depending on the denomination of the bill. For example, the method of reading the serial number (image recognition by optical image or image recognition by magnetic image) may be changed depending on the denomination of the bill.

Further, in the present embodiment, both the optical image of the serial number and the magnetic image of the serial number may be saved, but the image of the serial number to be saved may be switched based on the optical recognition result and the magnetic recognition result. ..

Further, in the present embodiment, the image displayed on the display unit 215 may be switched to an optical image or a magnetic image depending on the reason for rejection. For example, if a reject factor occurs in the recognition process or identification process based on the optical image, the optical image is displayed on the display unit 215, and if the reject factor occurs in the recognition process or identification process based on the magnetic image, the magnetic image is displayed. May be displayed on the display unit 215.

Further, in the first and second embodiments, the feature for specifying the shape and position information of the bill may be acquired from the magnetic image. For example, the shape and skew state of the bill can be estimated from the position and skew angle of the magnetic thread and the denomination according to the magnetic pattern of the magnetic thread.

Further, the bill processing apparatus 200 according to the first and second embodiments may be provided with a function of displaying the reason for rejection of the counterfeit ticket prevention element such as thickness, fluorescence, and ultraviolet rays on the display unit 215.

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 recognizing an image printed on paper sheets using magnetic ink.

100A, 100B: Banknote identification unit (paper leaf identification unit)
101: Image recognition unit 102: Image output unit 103: Hair roller 110: Magnetic line sensor 111: Magnetic head 112: Magnetic sensor 113: Magnet 120: Magnetic data processing unit 121: Magnetic image generation unit 122: Binarization unit 130: Storage unit 131: Judgment data 132: Template 133: Threshold 140: Optical line sensor 150: Upper unit 151, 161, 161a, 161b: Light emitting unit 155, 165: Transparent member 160: Lower unit 162: Condensing lens 163: Light receiving Unit 164: Light receiving unit Substrate 170: Optical data processing unit 171: Optical image generation unit 172: Binarization unit 200: Bill processing device (paper leaf processing device)
210: Hopper 211: Feeding unit 212: Transport path 213: Accumulating unit 214: Rejecting unit 215: Display unit B: Banknotes (paper sheets)
B1: Serial number B2: Amount B3: Symbol B4: Magnetic thread

Claims (13)

It is a paper leaf processing device that processes paper leaves on which symbols are printed using magnetic ink.
A transport unit that transports paper sheets one by one along the transport path,
A magnetic line sensor that is provided in a line shape in the width direction of the transport path and acquires magnetic data of the paper sheets to be transported, and a magnetic line sensor.
A magnetic image generation unit that generates a magnetic image that is an image based on the magnetic data of paper sheets acquired by the magnetic line sensor, and a magnetic image generation unit.
An optical line sensor that is provided in a line shape in the width direction of the transport path and acquires optical data of the paper sheets to be transported, and an optical line sensor.
An optical image generation unit that generates an optical image that is an image based on optical data of paper sheets acquired by the optical line sensor, and an optical image generation unit.
A magnetic image generated by the magnetic image generation unit, an image recognition unit that recognizes an optical image generated by the optical image generation unit, and an image recognition unit.
Equipped with
The image recognition unit recognizes the symbol from the optical image generated by the optical image generation unit and recognizes the symbol from the magnetic image generated by the magnetic image generation unit, and when there is an unread portion. , A paper leaf processing apparatus characterized in that the recognition results of both are complemented with each other to recognize the symbol. The paper leaf processing apparatus according to claim 1, wherein the symbol is a serial number or an amount. The paper leaf processing apparatus according to claim 1 or 2, wherein the magnetic line sensor has a resolution of 25 dpi or more in the main scanning direction and a resolution of 50 dpi or more in the sub-scanning direction. The magnetic image generation unit has a binarization unit that binarizes an image based on the magnetic data of paper sheets acquired by the magnetic line sensor and generates a binarized magnetic image as a magnetic image.
The image recognition unit compares the binarized magnetic image of the paper leaves generated by the binarization unit with the reference binarized magnetic image obtained by binarizing the image based on the magnetic data of the genuine paper leaves. The paper leaf processing apparatus according to any one of claims 1 to 3, wherein the paper sheet processing apparatus is characterized in that it is recognized by the operation. Equipped with a storage unit that stores a reference magnetic image, which is an image based on the magnetic data of genuine paper sheets.
The storage unit stores a plurality of character images used for the serial number as the reference magnetic image and the position information of the serial number corresponding to the type of paper leaf.
The invention according to any one of claims 1 to 4, wherein the image recognition unit recognizes the serial number described on the paper leaf by using the plurality of character images and the position information of the serial number. Paper leaf processing equipment. Equipped with a storage unit that stores a reference magnetic image, which is an image based on the magnetic data of genuine paper sheets.
The storage unit stores a character image used for the amount of money as the reference magnetic image and position information of the amount of money corresponding to the type of paper leaves.
The paper leaf according to any one of claims 1 to 5, wherein the image recognition unit recognizes the amount of money described in the paper leaf using the character image and the position information of the amount. Processing equipment. Equipped with a storage unit that stores a reference magnetic image, which is an image based on the magnetic data of genuine paper sheets.
The storage unit stores a symbol image described on a genuine paper leaf as the reference magnetic image and position information of the symbol corresponding to the type of the paper leaf.
The paper leaf according to any one of claims 1 to 6, wherein the image recognition unit recognizes the symbol described on the paper leaf by using the symbol image and the position information of the symbol. Processing equipment. Equipped with a storage unit that stores a reference magnetic image, which is an image based on the magnetic data of genuine paper sheets.
The storage unit stores the magnetic pattern of the magnetic thread provided on the genuine paper leaf as the reference magnetic image and the position information of the magnetic thread corresponding to the type of the paper leaf.
The image recognition unit is characterized in that it recognizes the magnetic pattern of the magnetic thread provided on the paper sheet by using the magnetic pattern of the magnetic thread and the position information of the magnetic thread. The paper leaf processing apparatus according to any one. When the image recognition unit cannot recognize at least a part of the optical image generated by the optical image generation unit, the image recognition unit adopts the magnetic recognition result of the corresponding portion of the magnetic image generated by the magnetic image generation unit. The paper leaf processing apparatus according to any one of claims 1 to 8. When the image recognition unit cannot recognize at least a part of the magnetic image generated by the magnetic image generation unit, the image recognition unit adopts the optical recognition result of the corresponding portion of the optical image generated by the optical image generation unit. The paper leaf processing apparatus according to any one of claims 1 to 9. The paper sheet according to any one of claims 1 to 10, further comprising a display unit, wherein the magnetic image generated by the magnetic image generation unit is displayed on the display unit so that the operator can see it. Kind of processing equipment. With more display
Claims 1 to 1, wherein the magnetic image generated by the magnetic image generation unit and the optical image generated by the optical image generation unit are displayed side by side on the display unit so that the operator can visually recognize them. 11. The paper leaf processing apparatus according to any one of 11. It is a paper leaf processing method for processing paper leaves on which symbols are printed using magnetic ink.
A magnetic data acquisition step of acquiring magnetic data of paper sheets transported along the transport path by a magnetic line sensor provided in a line shape in the width direction of the transport path, and a magnetic data acquisition step.
A magnetic image generation step of generating a magnetic image which is an image based on the magnetic data of paper sheets acquired by the magnetic line sensor, and
An optical data acquisition step for acquiring optical data of paper sheets transported along the transport path by an optical line sensor provided in a line shape in the width direction of the transport path, and an optical data acquisition step.
An optical image generation step of generating an optical image which is an image based on optical data of paper sheets acquired by the optical line sensor, and an optical image generation step.
The symbol is recognized from the optical image generated in the optical image generation step, the symbol is recognized from the magnetic image generated in the magnetic image generation step, and when there is an unread portion, both recognition results are obtained. An image recognition step that complements each other to recognize the symbols, and
A paper leaf processing method characterized by being equipped with.

Images