JP3885885B2 - Image processing apparatus and image processing method for securities - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus and an image processing method. In particular, the present invention relates to an image processing apparatus and an image processing method for acquiring an image necessary for specifying a securities used for settlement in settlement processing using securities such as checks and cash vouchers.
[0002]
[Prior art]
In commercial transactions, shopping at stores, payments at restaurants, etc., payments may be made using securities such as checks and cash vouchers. Hereinafter, these settlement processes will be described using checks as securities. In general, an account number or the like is printed as magnetic ink character data on the surface of the check, and the validity of the check is confirmed by referring to these data.
[0003]
Conventionally, when a check is received at a store or the like, after confirming the validity of the check, the date, amount, signature of the check maker, etc. are written and endorsed. These fronts and endorsements are often printed on printers except for signatures. Usually, a check that has been processed at a store or the like is brought into a settlement institution such as a bank and final settlement is performed. However, in recent years, check processing by electronic payment has been proposed to improve the efficiency of payment processing by transmitting electronic data of transaction contents, check image data, and the like to a payment organization.
[0004]
In addition, securities such as checks and cash vouchers usually have a background pattern on the surface. Therefore, when the image resolution is low, it may be difficult to distinguish important information such as account number, payment amount, payer, recipient, signature, and the like. On the other hand, when an image such as a check is acquired as high-resolution image data or color image data, the scanning speed for image acquisition is slow. In addition, since the amount of information of the image data increases, a load such as acquisition of image data and transmission to the storage device increases. Therefore, a lot of time is required for the image acquisition process. Therefore, as a method of acquiring image data of securities necessary for electronic payment with a small amount of information, binary image data obtained by scanning image of securities and binarizing image data acquired based on density is acquired. I thought about how to do it. For example, there is a method in which a threshold for binarizing image data based on density is a value that can be dynamically changed, and the threshold is calculated for each scanned image.
[0005]
[Problems to be solved by the invention]
However, when the threshold value is set to a value that can be dynamically changed and the threshold value is calculated for each scanned image, the threshold value is calculated from the background pattern of the temporarily scanned region. If it is not uniform, for example, if there is a background pattern area that is significantly different from the background pattern of the temporarily scanned area, important information such as account number, payment amount, payer, payee, signature, etc. in the binarized image data And background pattern could not be distinguished.
[0006]
Accordingly, an object of the present invention is to provide an image processing apparatus and an image processing method capable of accurately acquiring image data of securities necessary for electronic settlement with a small amount of information. Furthermore, one of the objects of the present invention is to provide an image processing apparatus and an image processing method capable of determining a method for converting image data acquired by scanning into image data necessary for electronic payment for each image of securities to be acquired. Is to provide a method.
[0007]
Further, an object of the present invention is to provide an image processing apparatus and image processing capable of reading MICR characters, printing necessary information, and efficiently acquiring image data such as checks for electronic payment of check payments. Is to provide a method.
[0008]
[Means for Solving the Problems]
In the present invention, a binarization processing method for binarizing an image of the entire securities is determined by analyzing a part of the image of the securities at the time of settlement processing, and based on the determined binarization processing method. The above-mentioned problems have been solved by accurately acquiring important information such as account number, payment amount, payer, recipient, signature, etc. as image information with a small amount of information.
[0009]
One aspect of the image processing apparatus of the present invention includes an image acquisition unit that converts first image data acquired by scanning the entire securities into binarized image data, and a partial area of the securities. A binarization processing method determining unit that determines a binarization processing method for converting the first image data into binarized image data based on the acquired second image data; The image processing apparatus is characterized in that the region is a character region including magnetic ink characters preprinted on the securities and one or a plurality of background regions including at least part of a background pattern.
[0011]
Another aspect of the image processing apparatus of the present invention further includes an edge pixel detection unit that detects an edge pixel from the background region described above, and the binarization processing method determination unit is based on the number of detected edge pixels. An image processing apparatus characterized by defining the binarization processing method.
[0012]
According to another aspect of the image processing apparatus of the present invention, the binarization processing method determination unit described above is based on the density distribution of pixels indicating the background of characters in the character area and the density distribution of pixels constituting the background area. An image processing apparatus characterized by defining a binarization processing method.
[0013]
According to another aspect of the image processing apparatus of the present invention, the image acquisition unit described above binarizes the first image data based on the first threshold value calculated based on the density distribution of the second image data. First binarization processing means for converting to the converted image data, and the first image data is subjected to the sharpening process for enhancing the edge pixels, and then the first image data is obtained based on the predetermined second threshold value. An image processing apparatus comprising: a second binarization processing unit that converts the image data into binarized image data.
[0014]
One aspect of the image processing method of the present invention is (a) a step of analyzing an image of a partial area of securities and determining a binarization processing method, wherein the partial area is included in the securities. A step that is a character region including a preprinted magnetic ink character and one or a plurality of background regions including at least a part of a background pattern; and (b) an image of the entire securities defined in step (a) 2 And a step of converting to binarized image data based on the binarization processing method.
[0016]
According to another aspect of the image processing method of the present invention, there is further provided a binary value based on (c) a step of detecting edge pixels from the background region, and (d) the number of edge pixels detected in step (c). An image processing method characterized by comprising a step of determining an image processing method.
[0017]
According to another aspect of the image processing method of the present invention, (e) binarization processing is further performed based on the density distribution of pixels indicating the background of characters in the character area and the density distribution of pixels constituting the background area. An image processing method comprising a step of defining a method.
[0018]
Another aspect of the image processing method of the present invention is further characterized in that one of the following binarization processing steps (f) and (g) is performed based on the result of the step (a). This is an image processing method.
(F) calculating a first threshold value based on the density distribution of the image of the partial region, and converting an image of the entire securities into binarized image data based on the calculated first threshold value
(G) A step of performing sharpening processing for emphasizing edge pixels on an image of the entire securities and then converting the image of the entire securities into binary image data based on a predetermined second threshold value.
[0019]
One aspect of the information recording medium of the present invention is a computer-readable information recording medium characterized by recording a program for causing a computer to execute each step of the image processing method described above.
[0020]
DETAILED DESCRIPTION OF THE INVENTION
An embodiment of the present invention will be described with reference to the drawings. In addition, the embodiment described below is for explanation, and does not limit the scope of the present invention. Accordingly, those skilled in the art can employ embodiments in which each or all of these elements are replaced by equivalents thereof, and these embodiments are also included in the scope of the present invention.
[0021]
FIG. 1 is a functional block diagram for explaining the basic concept of the image processing apparatus of the present invention. Here, in order to explain the present invention in a concise manner, the characteristic portions are mainly shown, and the description of the transport unit for transporting securities is omitted. The present invention can be applied not only to checks but also to processing of cash vouchers such as gift certificates and other securities. Hereinafter, a description will be given using checks which are typical examples of securities. .
[0022]
As illustrated in FIG. 1, the image processing apparatus 10 includes an image reading unit 15, a binarization processing method determination unit 20, and an image acquisition unit 30. In addition, the binarization processing method determination unit 20 includes a provisional scanning control unit 22, a binarization processing determination unit 23, a threshold calculation unit 24, and a binarization processing control unit that comprehensively controls the units 22, 23, and 24. 21 is provided. The image acquisition unit 30 includes an image acquisition control unit 31, a scan control unit 32, a binarization processing unit 33, an image storage unit 34, and a communication unit 35. Since the binarization processing method determination unit 20 is a part that performs preprocessing for image acquisition, 20 may be referred to as a preprocessing unit, and 30 may be referred to as a main processing unit.
[0023]
When the customer wishes to pay by check, the check is set in a check insertion portion (not shown) of the image processing apparatus 10. When the check is set, the provisional scanning control unit 22 of the binarization processing method determination unit 20 drives a conveyance control unit (not shown) so that a part of the check is read by the image reading unit 15 and the image reading unit 15 A part of the image data of the check is output to the value processing determination unit 23. In the present embodiment, the image reading unit 15 is configured by a monochrome image sensor, and the output image data includes, for example, luminance (hereinafter referred to as “density”) in which each pixel value is represented by 0 to 255. ) Is grayscale data represented by The binarization processing determination unit 23 determines a binarization processing method for binarizing the image data of the entire check from the grayscale data of a part of the check acquired by provisional scanning. Here, the binarization processing method to be determined is one of the following two methods.
[0024]
One method is to calculate a threshold value for a part of grayscale data of a check acquired by provisional scanning by analyzing a density distribution of each pixel constituting the data, and based on the calculated threshold value, This is a “threshold calculation method” for binarizing image data. The other method is a “sharpness method” in which the image data of the entire check is subjected to a sharpening process for enhancing edge pixels, and the sharpened image data is binarized based on a predetermined threshold value set in advance. .
[0025]
When the binarization processing method determined by the binarization processing determination unit 23 is the “sharpness method”, the binarization processing method determined by the binarization processing determination unit 23 is changed to the image acquisition unit 30. This is transmitted to the binarization processing unit 33. When the binarization processing method determined by the binarization processing determination unit 23 is the “threshold calculation method”, the check acquired from the binarization processing determination unit 23 to the threshold calculation unit 24 by temporary scanning is used. Some grayscale data is output. The threshold calculation unit 24 calculates a threshold for binarizing the image data of the entire check. The threshold calculated by the threshold calculation unit 24 is transmitted to the binarization processing unit 33 of the image acquisition unit 30 together with the binarization processing method determined by the binarization processing determination unit 23. The binarization processing unit 33 stores the received threshold value for subsequent binarization processing. In addition, the binarization processing control unit 21 controls the provisional scanning control unit 22, the binarization processing determination unit 23, and the threshold value calculation unit 24 in association with each other.
[0026]
The image acquisition control unit 31 controls each unit 32, 33, 34, 35 of the image acquisition unit. The scanning control unit 32 controls the check scanning based on the scanning instruction from the image acquisition control unit 31. When there is a scanning instruction from the image acquisition control unit 31, the scanning control unit 32 controls the image reading unit 15 to acquire image data of the entire check. In the present embodiment, the image data of the entire check is also grayscale data, like the image data of a part of the check. The gray scale data of the entire check output from the image reading unit 15 is transmitted to the binarization processing unit 33. In the binarization processing unit 33, based on the binarization processing method determined by the binarization processing determination unit 23, the grayscale data of the entire check received from the image reading unit 15 is converted into binarized image data. To do.
[0027]
The binarized image data of the entire check converted by the binarization processing unit 33 is stored in the image storage unit 34 and transmitted to an image storage device (not shown) via the communication unit 35. The image acquisition unit 30 has shown an example in which the binarized image data is stored in the image storage unit 34. However, the image acquisition unit 30 may be directly transmitted to the image storage device without providing the image storage unit 34. The image storage device stores the received binarized image so as to be searchable. This makes it possible to read and refer to the image of the related check when confirming the settlement of the financial institution or in trouble with the payer. Moreover, it is also possible to transmit image data together with electronic settlement data to a settlement organization as necessary.
[0028]
The processes of the binarization processing method determination unit 20 and the image acquisition unit 30 will be described in more detail with reference to FIG. FIG. 2 is a block diagram illustrating an example of a binarization processing method determination process by the binarization processing method determination unit 20 and an image acquisition process by the image acquisition unit 30.
[0029]
A part of the image of the check is read by the image reading unit 15 by the provisional scanning, and gray scale data of the part of the check is output. In the present embodiment, the grayscale data of a part of the check includes three pieces of data, that is, edge determination area portion reading data 200, MICR character portion reading data 201, and background portion reading data 202, each acquired from a predetermined portion of the check. Consists of The edge determination area reading data 200 output from the image reading unit 15 is subjected to edge pixel detection processing 203 by the binarization processing determination unit 23. In the detection of the edge pixel, for each pixel of the read grayscale data, the density difference between the target pixel and the surrounding pixels is calculated, and a frequency distribution of the density difference is generated. When the frequency distribution of the generated density difference is biased toward the larger absolute value of the density difference, the target pixel is detected as an edge pixel. Here, the frequency represents the number of pixels.
[0030]
Next, an edge pixel number calculation process 204 for calculating the number of edge pixels detected by the edge pixel detection process 203 is executed. Next, edge pixel determination processing for determining whether the binarization processing method is the “threshold calculation method” or the “sharpness method” based on the number of edge pixels calculated by the edge pixel number calculation processing 204 205 is executed. The determination condition in the edge pixel determination process 205 is whether or not the number of edge pixels is larger than a predetermined value.
[0031]
Next, when the edge pixel determination processing 205 determines that the binarization processing method is “threshold calculation method”, the histogram generation processing 206 uses the MICR character portion reading data 201 and the background portion reading data 202 to execute the histogram generation processing 206. Executed. On the other hand, when it is determined that the binarization processing method is the “sharpness method”, the process proceeds to a process of acquiring an image of the entire check without performing the processes (206 to 209) after the histogram generation process 206.
[0032]
The histogram generation process 206 is an average process for counting the number of pixels for each density from the read gray scale data, calculating the density distribution, and further smoothing the frequency distribution by removing noise. . In the background density determination processing 207 executed after the histogram generation processing 206, it is determined again whether the binarization processing method is the “threshold calculation method” or the “sharpness method” in the MICR character portion read data 201. The determination is made based on the density of the background data of the background portion read data 202.
[0033]
When the binarization processing method is determined to be the “threshold calculation method” by the background density determination processing 207, the threshold value calculation unit 24 determines the upper limit value and the lower limit value for determining the basis range for determining the threshold value. 208, and an appropriate threshold value calculation process 209 is performed within this range. If the threshold cannot be calculated for some reason, such as when the background data is dark, the binarization processing method is set to the “sharpness method”, and the process proceeds to processing for obtaining an image of the entire check. The calculated threshold value is used for processing for binarizing gray scale data of the entire check described later.
[0034]
When the binarization processing method is determined, an image reading process 210 is performed to read the entire check with the payment amount, the payee's description, the payer's signature, and the like. Here, the binarization processing method is determined when it is determined as the “sharpness method” or when a threshold value is calculated.
[0035]
When the binarization processing method is the “threshold calculation method” by the binarization processing method distribution processing 211, the grayscale data of the entire check acquired by the image reading processing 210 is the threshold calculated by the threshold calculation unit 24. Based on the above, a binarization process 212 for binarizing each pixel is performed and converted into binarized image data. The converted binarized image data is temporarily stored as a binarized image of the check by the binarized image storage processing 213 and transmitted to the image storage device.
[0036]
When the binarization processing method is “sharpness method” by the binarization processing method distribution processing 211, the edge pixel is added to the grayscale data of the entire check acquired by the image reading processing 210 by the sharpening image processing 214. Apply enhanced sharpening processing (generate sharpened image data). The image data that has undergone the sharpening process is subjected to a binarization process 215 that binarizes each pixel based on a predetermined threshold value that is set in advance, and is converted into binarized image data. The converted binarized image data is temporarily stored as a binarized image of the check by the binarized image storage processing 213 and transmitted to the image storage device.
[0037]
The image processing apparatus 10 as described above can be realized by a CPU, various storage devices such as a memory, hardware including a scanner device, and various control programs.
[0038]
FIG. 3 is a plan view showing an example of a check 40 used for payment, and FIG. 4 is a diagram showing an example of binarized image data acquired from the check shown in FIG. On the check 40, a payer 41, a check number 42, an issue date 43, a payee 44, a payment amount 45, a payment name 46, a payer signature 47, an MICR character 48, etc. are printed or described. In the figure, “XXX” indicates a numeral, signature, character, or the like. These characters can be printed or signed after provisional scanning, except for the MICR characters 48. The check 40 is also dotted with many background patterns 49. There are many types of check background patterns, and their densities are also different. FIG. 3 shows an example of a background pattern, which is not clearly shown in the figure, but generally has a light gray as the base color of the background, and a slightly dark gray background pattern on the base color. It is assumed that 49 are scattered.
[0039]
In the binarized image data 50 of FIG. 4, each pixel data is binarized to white or black, and the background portion 49 and the light gray of the base are removed. 51 to 58 in the figure correspond to 41 to 48 in FIG. 3, respectively.
[0040]
FIG. 5 is a flowchart illustrating an example of an image acquisition process procedure when acquiring binary image data of a check.
[0041]
First, gray scale data is acquired from a part of a check by primary scanning (hereinafter referred to as “provisional scanning”) (S101). Next, a binarization processing method for converting the image data of the entire check into binarized image data is based on the number of edge pixels detected from a part of the grayscale data of the check acquired by the temporary scan, and the temporary scan. A determination is made based on the background density of the MICR character portion and the background printing portion calculated from the partial grayscale data of the acquired check (S102).
[0042]
Here, the area to be temporarily scanned is an area including the MICR character 48 and the background pattern 49 on the check. Since the print position in the width direction of the MICR character 48 is determined by the type of check, the image data of the MICR character portion can be specified based on the output from the image reading unit 15. FIG. 3 shows the MICR character print area as “c”. It can be determined that the MICR character is detected when scanning from the right side of the check 40 and a predetermined image output is obtained from the portion corresponding to the MICR print area c. In addition, the temporary scanning area can be a position further away from the initial detection position of the MICR character by a predetermined distance. Further, when the image processing apparatus 10 is a composite processing apparatus provided with a MICR character reader, the MICR character print position is confirmed based on the output of the MICR character reader, and the provisional scanning area is determined based on the print position. Is also possible.
[0043]
FIG. 3 illustrates the provisional scanning region T. The temporary scanning region T is a region where the width of “b” from the first MICR character is used as the temporary scanning region. FIG. 3 also shows a temporary scanning region T ′ as another example. The temporary scanning region T ′ has a position that is a predetermined distance from the first MICR character as the temporary scanning region. The temporary scan area preferably includes a background pattern that includes the MICR characters and best represents the features of the check. In addition, an area E having a width of “b” in which no MICR character exists in the MICR print area c can be added to the temporary scan area T as a temporary scan area.
[0044]
FIG. 7 is a schematic diagram illustrating an example of a temporary scanning region used for determination of the binarization processing method. FIG. 7A shows an example of the temporary scanning region T shown in FIG. 3 and a region selected for edge pixel detection (hereinafter referred to as “edge determination region”), and FIG. And a region (hereinafter referred to as “specific region”) selected for creating a histogram of the MICR character portion and the background printing portion. In the example of FIG. 7A, an area that does not include the MICR character area 74 is selected from the temporary scanning area T as the edge determination area 75. Since the MICR character portion always includes character information, the number of edge pixels increases. Therefore, the determination of the edge pixel in the region excluding the MICR character portion is an accurate determination for distinguishing the background pattern. Further, it is also possible to add the area E where no MICR characters exist in the MICR print area c of FIG. 3 to the edge determination area.
[0045]
In the example of FIG. 7B, the MICR character area 74 and the three background areas 71 to 73 are selected from the temporary scanning area T as the specific area. In this way, it is possible to select a non-continuous area in the temporary scanning area as a specific area for creating a histogram. It is arbitrary how much range is selected as the specific region. However, it is desirable to include a portion where the MICR character and the background print density are high, and that the number of pixels is the minimum number of pixels necessary to determine an appropriate threshold value. It is also possible to select a character printing portion such as the recipient 44 instead of the MICR character portion.
[0046]
Next, it is determined whether or not the determined binarization processing method is the “threshold value calculation method” (S103). When the determined binarization processing method is the “threshold value calculation method” (S103: Yes), the secondary scan (hereinafter, “main scan” is performed based on the grayscale data of a part of the check acquired by the temporary scan. The threshold value for binarizing the image data of the entire check obtained by “)” is calculated (S104). On the other hand, when the binarization processing method is the “sharpness method” (S103: No), the process proceeds to step S107.
[0047]
Next, it is determined whether or not a processing abnormality such as a calculation error has occurred in the threshold value calculation process (S105). When processing abnormality occurs in the threshold value calculation processing (S105: Yes), the binarization processing method is changed to “sharpness method” (S106). On the other hand, if no processing abnormality has occurred in the threshold calculation process (S105: No), that is, if the threshold calculation process ends normally and the threshold is calculated, the process proceeds to step S107.
[0048]
Next, a main scan of the check is executed (S107), and the determined binarization processing method is determined (S108). When the binarization processing method is the “threshold calculation method” (S108: “threshold calculation method”), the threshold value calculated in step S104 is acquired (S109), and the binarization processing based on the acquired threshold value is performed. Binarized image data is generated by applying the grayscale data of the entire check input from the image reading unit 15 (S112). The generated binarized image data is stored in the image storage unit 34 or transmitted to the image storage device (S113).
[0049]
When the binarization processing method is “sharpness method” (S108: “sharpness method”), sharpening processing in which edge pixels are emphasized is applied to the grayscale data of the entire check input from the image reading unit 15, Sharpened image data is generated (S110). Next, a predetermined threshold value set in advance for binarizing the image data of the entire check is acquired (S111), and binarization processing based on the acquired threshold value is applied to the sharpened image data. Binarized image data is generated (S112). The generated binarized image data is stored in the image storage unit 34 or transmitted to the image storage device (S113).
[0050]
The above-described sharpened image data is generated by applying a sharpness filter coefficient to grayscale data in order to enhance edge pixels. Hereinafter, conversion from grayscale data to sharpened image data is referred to as “sharpness conversion”. Also, the sharpness conversion process is called “sharpening image processing”. FIG. 13 shows an example of the sharpness filter coefficient. The value obtained by multiplying the density value of the target pixel by k1 and the density value of surrounding pixels by k2 is set as the density value of the target pixel. Similar processing is performed on each pixel constituting the grayscale data to obtain sharpened image data. For example, the surrounding pixels are surrounding pixels separated by k3 pixels in the vertical and horizontal directions. That is, the following sharpness conversion formula is obtained. In FIG. 13, k1 = 5, k2 = −1, and k3 = 5.
[0051]
Density of target pixel after sharpness conversion p =
(Target pixel density × k1)
+ (Density of surrounding pixel 1 × k2)
+ (Density of surrounding pixel 2 × k2)
+ (Density of surrounding pixel 3 × k2)
+ (Density of surrounding pixel 4 × k2)
Here, when the density p after sharpness conversion is p <0, p = 0, and when p> 255, p = 255.
[0052]
Using the above sharpness conversion formula, sharpening image processing is performed on all the pixels of the grayscale data acquired by the main scanning. FIG. 14 is a diagram showing a histogram before and after sharpness conversion. FIG. 14A is a diagram showing a histogram before sharpness conversion, and FIG. 14B is a diagram showing a histogram after sharpness conversion. As shown in FIG. 14, the density of the pixel after sharpness conversion is concentrated at “0” or “255”. Therefore, it can be seen that the sharpened image data may be binarized with a preset threshold value of 1 to 254.
[0053]
Here, the threshold value to be set is preferably a small value. For example, 30. This is to reduce error pixels (hereinafter referred to as “noise pixels”) due to the binarization process and further increase the compression efficiency after the binarization process. FIG. 15 is a diagram showing density values before and after sharpness conversion of one line of the MICR character part. The x-axis shows the number of dots and the y-axis shows the density.
[0054]
As shown in FIG. 15, in the sharpened image data after the sharpness conversion, the density value of the black part of the image changes to 0 and the density value of the white part of the image changes to 255, and the difference in light and shade is emphasized. Further, when there is a difference in density before sharpness conversion between adjacent pixels, the density difference after sharpness conversion is further increased and the density is enhanced. For example, the black ellipse 61 in the figure. Depending on the threshold value, this large density difference portion becomes a noise pixel. For example, when the threshold value is set to 100, in the elliptical part 61, before sharpness conversion, a part represented as white is represented as black after sharpness conversion and becomes a noise pixel. Therefore, in order to reduce this noise pixel, it is better to set the threshold value to a low value (for example, 30).
[0055]
FIG. 6 is a flowchart illustrating an example of the determination processing procedure of the binarization processing method.
[0056]
First, edge pixels in the edge pixel determination area 75 where the background pattern 49 exists and where the MICR character 48 does not exist are detected from a part of the grayscale data of the check acquired by provisional scanning, The edge pixel number Ne is calculated (S201). Here, the edge pixel is a pixel having a large absolute value of a density difference from surrounding pixels. For detection of edge pixels, as shown in FIG. 7, the edge pixel determination area 75 is divided into a plurality of blocks 76, and edge pixel determination is performed for each block.
[0057]
FIG. 12 is a diagram for explaining edge pixel determination. In the edge pixel determination, first, a density difference between a pixel to be determined (hereinafter referred to as “determination target pixel”) and surrounding pixels is calculated. FIG. 12A is a diagram illustrating determination target pixels for determining edge pixels in one block and peripheral pixels to be used for calculation of density difference for the determination. Assume that a set of m1 × m2 pixels is one block, a pixel at the center of the block is a determination target pixel A, and pixels other than the determination target pixel A are peripheral pixels. Here, m1 = 5 and m2 = 5 will be described as an example. Further, if all the peripheral pixels included in one block are subject to calculation, it takes a long time to calculate. For example, every other thinned pixel is a peripheral pixel to be calculated (hereinafter referred to as “calculation target peripheral pixel”). Said).
[0058]
Accordingly, as shown in FIG. 12A, the calculation target peripheral pixels for calculating the density difference from the determination target pixel are twelve. Next, the density difference between each of the 12 calculation target peripheral pixels and the determination target pixel is calculated, and the calculated density difference is classified into a range m3 of density differences divided into m3, and each range is determined. The number of pixels classified into (2) is integrated. Here, m3 = 5 will be described as an example. The density difference boundary values n1 and n2 (n1 and n2 are positive integers) are set in advance, and the range 1 is set when the density difference Pi is −255 <= Pi <−n2, and −n2 <= Pi <−n1. Is a range 2 when -n1 <= Pi <= n1, a range 4 when n1 <Pi <= n2, and a range 5 when n2 <Pi <= 255. As described above, the density difference between each of the twelve calculation target peripheral pixels and the determination target pixel is classified into the range 1 to the range 5 and integrated for each range. Here, for example, n1 = 10 and n2 = 50.
[0059]
12B and 12C are diagrams showing the frequency distribution of the density difference between the determination target pixel and the calculation target peripheral pixel. As shown in FIG. 12B, when the density difference distribution is biased toward the larger absolute value of the density difference, that is, when the density difference distribution is biased toward the range 1 or 5, the determination target. It can be determined that the density difference between the pixel and the calculation target peripheral pixel is large. Therefore, in this case, the determination target pixel is determined as an edge pixel. On the other hand, as shown in FIG. 12C, when the density difference distribution is not biased, it cannot be determined that the density difference between the determination target pixel and the calculation target peripheral pixel is large. Accordingly, in this case, it is determined that the determination target pixel is not an edge pixel. For example, the following conditions can be cited as conditions for determining whether the density difference distribution is biased toward the larger absolute value of the density difference. Here, Ci is the frequency in the range i, and Te is a preset value.
[0060]
(C5> Te and C1 = 0)
Or
(C1> Te and C5 = 0)
As described above, one edge pixel is determined with 5 × 5 pixels as one block. Therefore, when the number of blocks included in the edge pixel determination area 75 is Nb, the number Ne of detected edge pixels is 0 <= Ne <= Nb.
[0061]
Returning to FIG. 6, it is next determined whether or not the calculated edge pixel number Ne is larger than a predetermined value Ne0 (S202). That is, it is determined whether or not (Ne> Ne0). Here, the predetermined value Ne0 is a value derived from an experiment or the like and can be changed. For example, the value is 10% of the number of blocks in the edge pixel determination area. That is, Ne0 = Nb × 0.1.
[0062]
When the calculated edge pixel number Ne is larger than the predetermined value Ne0, that is, when (Ne> Ne0) (S202: Yes), the process proceeds to the next step S203. On the other hand, when the calculated edge pixel number Ne is equal to or less than the predetermined value Ne0, that is, when (Ne <= Ne0) (S202: No), the binarization processing method is set to “sharpness method” (S209). ).
[0063]
Next, a frequency distribution is generated for the pixels of the portion where the MICR character 48 exists from a part of the image data of the check acquired by the temporary scanning (S203). Next, a frequency distribution is generated for the pixels of the portion where the background pattern 49 exists from the image data of a part of the check acquired by the temporary scanning (S204). Next, in order to remove the noise portion of the frequency distribution from the frequency distribution of the generated MICR character portion and the frequency distribution of the background portion, each frequency distribution is averaged (S205). Here, the result of averaging the density distribution is called a histogram. The averaging process is, for example, a process in which an average value of a total of 9 points including a point i and 4 points before and after the point i is set as the value of the point i. That is, point (i-4), point (i-3), point (i-2), point (i-1), point i, point (i + 1), point (i + 2), point (i + 3) and point ( Let the average value of i + 4) be the value of point i.
[0064]
FIG. 8 shows a frequency distribution (histogram) obtained by averaging the MICR character portion 74 and the background portions 71 to 73 in the temporary scanning region T of the check 40 shown in FIG. The horizontal axis indicates density, and the vertical axis indicates frequency (number of pixels). The density (brightness) on the horizontal axis is represented by 0 to 255, and the smaller the number, the higher the density (the brightness is low: dark). A first peak 80 having a density of 20 to 60 in FIG. 8 is a set of pixels constituting the MICR character, and a second peak 81 having a density of 100 to 160 is a MICR character part 74 (FIG. 7). This is a set of pixels representing the background pattern 49. A third peak 82 having a density of 160 to 210 is a set of pixels representing the background pattern 49 of the background portions 71 to 73 (FIG. 7). The histogram of FIG. 8 is composed of a MICR character part 74 and background parts 71 to 73. FIG. 9 shows a histogram of the MICR character part 74, and FIG. 10 shows a histogram of the background parts 71 to 73, respectively.
[0065]
Next, the peak value density Pn1 and standard deviation σ1 of the background pattern 49 in the MICR character area 74 and the peak value density Pn2 and standard deviation σ2 of the background pattern 49 in the background areas 71 to 73 are calculated (S206). Next, based on the peak value density Pn1 and standard deviation σ1 in the MICR character area 74 and the peak value density Pn2 and standard deviation σ2 in the background areas 71 to 73, the background density and background areas 71 to 73 in the MICR character area 74 are displayed. It is determined whether the background density is substantially the same (S207). For example, it is determined whether or not the peak value concentration Pn1 and the peak value concentration Pn2 are substantially the same, and the standard deviation σ1 and the standard deviation σ are substantially the same.
[0066]
When the background density of the MICR character area 74 and the background density of the background areas 71 to 73 are substantially the same (S207: Yes), the binarization processing method is set as the “threshold value calculation method” (S208). On the other hand, when the background density of the MICR character area 74 and the background density of the background areas 71 to 73 are not substantially the same (S207: No), the binarization processing method is set to “sharpness method” (S209).
[0067]
FIG. 11 is a flowchart illustrating an example of a threshold value calculation processing procedure (“threshold value calculation method”). First, from the histogram of the MICR character portion 74, a minimum density limit value PMin (FIGS. 8 and 9) for calculating a threshold value is calculated (S301). Since the threshold value distinguishes the background and the character, the threshold value must be at least between the density of the MICR character and the background portion. Therefore, the minimum threshold value PMin of the threshold needs to be brighter than the MICR character. For this purpose, first, the minimum limit value PMin is calculated based on the density of the MICR characters. When the minimum limit value PMin cannot be obtained from the histogram of the MICR character portion 74 (S302: Yes), the binarization processing method is changed to “sharpness method” (S309).
[0068]
When the minimum limit value PMin is calculated (S302: No), the MICR maximum limit value PMMax (FIG. 9), which is the maximum limit value of the MICR character portion 74, is obtained (S303). The MICR maximum limit value PMMax is a brighter limit value obtained from the histogram of the MICR character part 74. If the threshold value is further increased, the background pattern 49 of the MICR character part 74 and the MICR character cannot be distinguished. This is the limit value. When the MICR maximum limit value PMMax cannot be obtained (S304: Yes), the binarization processing method is changed to “sharpness method” (S309).
[0069]
When the MICR maximum limit value PMMax is calculated (S304: No), the background maximum limit value PbMax (FIG. 10), which is the maximum limit value of the background portions 71 to 73, is obtained (S305). The maximum background limit value PbMax is the brighter limit value of the background portions 71 to 73. If the threshold value is further increased, the limit value cannot be distinguished from the background pattern 49 of the background portions 71 to 73 and the MICR characters. It will be. When the background maximum limit value PbMax cannot be obtained (S306: Yes), the binarization processing method is changed to “sharpness method” (S309). Accordingly, a threshold value cannot be calculated by the “threshold value calculation method”, and a check processed with an unreadable error can also obtain a binary image by changing the binarization processing method to “sharpness”.
[0070]
When the background maximum limit value PbMax is calculated (S306: No), the threshold maximum limit value PMax (FIG. 8) is obtained (S307). The maximum limit value PMax indicates the maximum limit value of the threshold for eliminating the background pattern 49 of the MICR character portion 74 and the background portions 71 to 73, and is a small value of the MICR maximum limit value PMMax and the background maximum limit value PbMax. Becomes the maximum limit value PMax. When the minimum limit value PMin and the maximum limit value PMax are obtained, the threshold value is calculated from these values (S308).
[0071]
(Embodiment applied to a composite processing apparatus)
Checks are also used for shopping at retail stores such as supermarkets. In a supermarket, a POS system is usually installed in a checkout counter. In the case of payment by check, in addition to conventional processing such as reading magnetic ink characters (MICR characters) on checks and printing necessary items, image data acquisition processing of printed surfaces such as checks, transaction details Or image data transmission processing is required. These operations are not only complicated for the operator, but also require a long time for processing. In addition, when the printer, the MICR reader, the scanner (image processing apparatus), and the like are separate apparatuses, a problem arises in that an installation space for each apparatus must be secured around the cash register (settlement counter). This problem can be solved by a composite processing apparatus having a printer and a MICR reader.
[0072]
That is, it is possible to continuously perform a series of processing such as checks from magnetic ink character reading processing to printing surface scanning processing by a composite processing apparatus provided with a MICR reader, scanner, and the like in a POS printer. As a result, it is possible to reduce the operation burden on the operator in payment by check, and to shorten the time required for processing.
[0073]
For this purpose, a composite processing apparatus having a printing function includes a conveyance path that guides a print medium such as a check, a magnetic head that is arranged along the conveyance path and reads magnetic ink characters recorded in advance on the print medium, and a conveyance path. A first print head disposed along the path and performing printing on the first surface of the print medium; and a second print head disposed along the transport path and performing printing on the second surface of the print medium; It is desirable to have a configuration including a scanner (image reading unit) that is arranged along the conveyance path and scans (scans) the first or second surface of the print medium. Such a composite processing apparatus can also be considered as a POS printer having a check processing function.
[0074]
A composite processing apparatus according to an embodiment of the present invention having a POS printer function and a MICR function will be described below with reference to the drawings.
[0075]
FIG. 16 is a perspective view of the composite processing apparatus 110 according to an embodiment of the present invention having a POS printer function and an MICR character reading function. As shown in this figure, the composite processing apparatus 110 is covered with a resin cover 111, and an insertion port 112 for manually inserting the check P is formed on the front surface portion thereof, while the upper surface portion thereof is formed. Is formed with a discharge port 113 through which the check P is discharged. Furthermore, the composite processing apparatus 110 of the present embodiment includes a roll paper storage unit (not shown) that stores roll paper at the rear part thereof, and the roll paper stored in the roll paper storage unit passes through the printing unit. It is pulled out from the roll paper discharge port 114 on the upper surface of the apparatus.
[0076]
FIG. 17 is a side sectional view showing the internal structure of the composite processing apparatus 110. As shown in this figure, a conveyance path 115 of a check P from the insertion port 112 to the discharge port 113 is formed inside the combined processing apparatus 110. The conveyance path 115 has the insertion port 112 side facing the horizontal direction, while the discharge port 113 side faces the vertical direction, and bends in an L shape in a side view. On the conveyance path 115, in order from the insertion port 112 side, a paper rear end detector 116, MICR head (magnetic head) 117, first feed roller pair 118, paper front end detector 119, paper positioning member 120, back print head A (second print head) 121, a second feed roller pair 122, a front print head (first print head) 123, a paper discharge detector 124, and a scanner 125 are arranged. A scanner feed roller (press feed roller) 126 is provided.
[0077]
The paper trailing edge detector 116, the paper leading edge detector 119, and the paper discharge detector 124 are configured by, for example, a transmissive or reflective photosensor, and the presence or absence of the check P is not contacted at each position in the conveyance path 115. To detect. The sheet positioning member 120 is used to temporarily stop the check P inserted from the insertion port 112 at a predetermined position. For example, the sheet positioning member 120 protrudes into the transport path 115 according to driving of an actuator such as a solenoid, and the transport path. It is configured to perform a transforming operation to a posture of retreating from 115. The first feed roller pair 118 and the second feed roller pair 122 are each composed of a pair of roller members facing each other with the transport path 115 interposed therebetween, and transport the check P in both forward and reverse directions by driving one of the rollers. Further, one of the roller members is configured to be movable back and forth with respect to the other roller member, and opens and closes the transport path 115 by an advance / retreat operation according to driving of an actuator such as a solenoid.
[0078]
The MICR head 117 is for reading magnetic ink characters recorded on the surface of the check P, and the validity / invalidity of the check P is determined based on the read data of the MICR head 117. As shown in FIG. 18, the magnetic ink characters are recorded in the MICR recording area 127 on the surface of the check P, and the recorded data includes the account number of the check P and the like. A pressing member 117a that presses the check P against the MICR head 117 during a reading operation is provided at a position opposite to the MICR head 117. However, the pressing member 117a is normally retracted from the MICR head 117, and the conveyance path 115 is opened.
[0079]
The front print head 123 is for printing a written item such as a payee, a date, and an amount of money on the surface of the check P, and this written item is printed in a written region 128 shown in FIG. The front print head 123 is a serial print head supported by a carriage, and realizes dot matrix printing in one or more columns while moving in the width direction of the check P. In this embodiment, a dot impact type print head for transferring the ink on the ink ribbon to the check P is used as the front print head 123, but another type of print head may be used.
[0080]
The back printing head 121 is for printing the endorsement items necessary for the store side such as the shopper's authentication number, date, and usage amount on the back side of the check P. The endorsement items are shown in FIG. Printed in the endorsement area 129. The back printing head 121 is of a shuttle type and includes a plurality of heads with a predetermined interval in the width direction of the check P, and one or a plurality of rows of dot matrix printing is performed by moving the head within the interval width. Realize. In this embodiment, a dot impact type print head that transfers the ink on the ink ribbon to the check P is employed as the back print head 121, but another type of print head may be employed.
[0081]
The scanner 125 corresponds to the image reading unit 15 shown in FIG. 1 and scans the surface of the printed check P to obtain image data. The read image data is stored in an image storage device ( (Not shown) and used for electronic settlement or transaction verification. The acquired image data may be stored in the image storage device after being compressed. In addition, as described in the above example, a temporary scan is first performed to calculate a threshold value, and image data such as a check acquired by the subsequent main scan is converted into binarized image data based on the threshold value. The In the present embodiment, a contact image sensor (CIS: Contact Image Sensor) is used as the scanner 125, and the scanning operation is performed with the check P in close contact with the reading surface 125a.
[0082]
The scanner feed roller 126 is for conveying the check P during the scanning operation, and conveys the check P to the discharge port 113 side while pressing the check P against the reading surface 125a of the scanner 125. At this time, as shown in FIG. 19, the scanner feed roller 126 presses the check P to a position slightly shifted from the scanner focus position A without pressing the check P to the focus position A of the scanner 125. That is, the scanner focal position A is offset to the upstream side or downstream side with respect to the scanner contact position B of the scanner feed roller 126. For example, in this embodiment, the scanner focus position A is offset from the scanner contact position B of the scanner feed roller 126. The scanner focal position A is set at a position offset by 0.8 mm on the downstream side of the conveyance (discharge port 113 side). Thereby, it is avoided that the pressing force of the scanner feed roller 126 acts directly on the scanner focal position. Therefore, when the check P is scanned immediately after printing, the amount of ink adhering to the scanner focal position A can be reduced, and it is possible to prevent the deterioration of the quality of the scanned image due to the ink adhering as much as possible. Further, in the present embodiment, as described above, the scanner focal position A is offset downstream of the scanner contact position B of the scanner feed roller 126, so that the scanable area on the front end side of the check P is widened. Is possible. When the scanner focal position A is largely offset with respect to the scanner contact position B of the scanner feed roller 126, the check P may float from the reading surface 125a at the scanner focal position A. However, as described above, Since the amount is about 0.8 mm, the float of the check P with respect to the reading surface 125a is suppressed to 0.2 mm or less, and there is no possibility that the quality of the scanned image is deteriorated.
[0083]
In the above scanning operation, first, provisional scanning (primary scanning) is performed, and part of the image data of the check is acquired for determination of the binarization processing method and / or calculation of the threshold value. When the binarization processing method is determined as the “sharpness method” from the image data acquired by the temporary scanning, or when the threshold value is calculated when the binarization processing method is the “threshold calculation method”, the main scanning is started. . In the main scan (secondary scan), an image is read while the check P is conveyed upward by the scan feed roller 126, and the check P is discharged as it is from the discharge port 113. At this time, the end portion of the discharged check P is held by the downstream conveyance path 115 of the scan feed roller 126. That is, the end portion of the conveyance path 115 (between the scan feed roller 126 and the discharge port 113) is oriented substantially vertically and has a length of about L / 6 (L: the length of the check P). Therefore, it is possible to hold the discharged check P, and the discharged check P does not fall from the combined processing apparatus 110.
[0084]
【The invention's effect】
As described above, the present invention determines a binarization processing method for binarizing the image data of the entire securities based on the image data acquired by temporarily scanning a part of the image of the securities. Based on the determined binarization processing method, important information such as account number, payment amount, payer, payee, signature, etc. can be obtained as image information accurately with a small amount of information. It has become possible to provide an image processing apparatus and an image processing method capable of accurately obtaining necessary securities image data with a small amount of information.
[0085]
In other words, the “threshold calculation method” can be acquired at high speed with a smaller amount of data, and the “sharpness method” can accurately read image information that could not be read.
[0086]
Further, according to the embodiment of the present invention, it is possible to read MICR characters, print necessary information, and efficiently acquire image data such as a check in order to electronically pay a check. As a result, a series of processes such as checks from the magnetic ink character reading process to the printing surface scanning process can be performed continuously, reducing the operational burden on the operator and reducing the time required for the process. it can.
[Brief description of the drawings]
FIG. 1 is a functional block diagram for explaining a basic concept of an embodiment of an image processing apparatus according to the present invention.
FIG. 2 is a block diagram illustrating an example of a binarization processing method determination process by a binarization processing method determination unit 20 and an image acquisition process by an image acquisition unit 30;
FIG. 3 is a plan view showing an example of a check.
FIG. 4 is a diagram showing an example of binarized image data acquired from the check shown in FIG.
FIG. 5 is a flowchart showing an example of an image acquisition process procedure when acquiring binary image data of a check according to the present invention.
FIG. 6 is a flowchart showing an example of a determination processing procedure of the binarization processing method according to the present invention.
FIG. 7 is a schematic diagram illustrating an example of a provisional scanning region used for determination of a binarization processing method.
8 is a frequency distribution diagram (histogram) obtained by averaging the MICR character portion 74 and the background portions 71 to 73 in the provisional scanning region T of the check 40 shown in FIG.
9 is a histogram of the MICR character part 74 constituting the histogram of FIG.
10 is a histogram of background portions 71 to 73 constituting the histogram of FIG. 8;
FIG. 11 is a flowchart showing an example of a threshold calculation processing procedure according to the present invention.
FIG. 12 is a diagram for explaining edge pixel determination.
FIG. 13 is a diagram illustrating an example of a sharpness filter coefficient.
FIG. 14 is a diagram showing a histogram before and after sharpness conversion.
FIG. 15 is a diagram showing density values before and after sharpness conversion of one line of a MICR character part.
FIG. 16 is a perspective view of a composite processing apparatus 110 having a print function and the like.
17 is a side sectional view showing the internal structure of the composite processing apparatus 110. FIG.
FIG. 18 is a schematic view of a check.
FIG. 19 is a side view showing a scanner and a scanner feed roller.
[Explanation of symbols]
10 Image processing device
15 Image reading unit
20 Binarization processing method determination unit
30 Image acquisition unit
40 Check surface
42 check number
44 Payee
45 Payment amount
47 Signature
48 Magnetic ink characters (MICR characters)
49 Background Pattern
50 Binary image data
110 Combined processing equipment
112 insertion slot
113 outlet
115 Transport route
117 MICR head
118 First feed roller pair
121 Back printing head
122 Second feed roller pair
123 Front print head
125 scanner
126 Scanner feed roller
130 Roller spindle
132 Roller retracting mechanism
133 Solenoid for scanner feed roller
134 Presser bar
135 Presser spring
136 1st gear
137 2nd gear
138 Scanner feed motor
140 Control unit
P check
T Temporary scanning area

Claims (10)

An image acquisition unit that converts first image data acquired by scanning the entire securities into binary image data;
A binarization processing method determination unit that determines a binarization processing method for converting the first image data into binarized image data based on the second image data acquired from a partial area of the securities. And the partial area is a character area including magnetic ink characters preprinted on the securities and one or more background areas including at least a background pattern. Processing equipment.   The image processing apparatus further includes an edge pixel detection unit that detects an edge pixel from the background area, and the binarization processing method determination unit determines the binarization processing method based on the number of detected edge pixels. The image processing apparatus according to claim 1.   The binarization processing method determining unit determines the binarization processing method based on a density distribution of pixels indicating the background of the character in the character area and a density distribution of pixels constituting the background area. The image processing apparatus according to claim 1, wherein the image processing apparatus is characterized.   The image acquisition unit converts a first binarized image data into a binarized image data based on a first threshold value calculated based on a density distribution of the second image data. The image processing apparatus according to claim 1, further comprising a processing unit.   The image acquisition unit performs a sharpening process for enhancing edge pixels on the first image data, and then converts the first image data into a binarized image based on a predetermined second threshold value. The image processing apparatus according to claim 1, further comprising second binarization processing means for converting into data. An image processing method comprising the following steps.
(A) Analyzing an image of a partial area of securities and determining a binarization processing method, wherein the partial area includes a character area and a background including magnetic ink characters printed in advance on the securities The image of the entire securities (step (b)), which is one or more background regions including at least a part of the pattern, is converted into binarized image data based on the binarization processing method defined in step (a). Process to convert The image processing method according to claim 6, further comprising the following steps.
(C) detecting an edge pixel from the background region (d) determining the binarization processing method based on the number of edge pixels detected in the step (c) The image processing method according to claim 6, further comprising the following steps.
(E) determining the binarization processing method based on a density distribution of pixels indicating the background of the character in the character area and a density distribution of pixels constituting the background area The image processing method according to claim 6, wherein one of the following binarization processing steps (f) and (g) is performed based on the result of the step (a).
(F) calculating a first threshold value based on a density distribution of the image of the partial area, and converting an image of the entire securities into binarized image data based on the calculated first threshold value. (G) The image of the entire securities is subjected to a sharpening process for enhancing edge pixels, and then the entire image of the securities is converted into binary image data based on a predetermined second threshold. Process   10. A computer-readable information recording medium on which a program for causing a computer to execute each step of the image processing method according to claim 6 is recorded.

Images