JP2010244386A - Image reading device, method for controlling image reading device, control program and storage medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image reading device for preventing the erroneous recognition of a magnetic ink character when the magnetic ink character passes through a magnetic head in an inclined state. <P>SOLUTION: An image reading device includes: a magnetic reading means (S101) for reading magnetic recognition data from a magnetic ink character printed on an original; a magnetic character recognition means (S105, S106) for recognizing the magnetic ink character printed on an original based on magnetic similarity to be obtained by comparing the read magnetic recognition data with MICR dictionary information; an image reading means (S102) for optically reading the magnetic ink character printed on the original as image data; and a calculation means (S103) for calculating the angle of inclination of the character based on the image data of the read magnetic ink character. The magnetic character recognition means changes the MICRO dictionary information to be used for recognizing the magnetic ink character based on the angle of inclination of the magnetic ink character to be calculated by the calculation means (S104). <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an image reading apparatus that reads magnetic ink characters printed on an original such as a check or bill, a control method for the image reading apparatus, a control program, and a storage medium.

  In the check settlement process, magnetic ink characters such as an account number and money amount printed in the lower column of the check are read, and the read data is referred to a predetermined organization to confirm the validity of the check.

  There is a MICR (Magnetic Ink Character Recognition) device as a device for reading magnetic ink characters. The MICR device reads magnetic information contained in magnetic ink characters with a magnetic head and recognizes the characters based on the read magnetic information (Patent Document 1).

  In addition to the MICR device, an OCR (Optical Character Recognition) device is provided, and when the recognition result of the magnetic ink characters in the MICR device is unreliable, an image reading device that recognizes the magnetic ink characters again in the OCR device is proposed. (Patent Document 2).

JP 49-49545 A JP-A-2004-259254

  However, in the device of Patent Document 2, the degree of similarity with dictionary data for recognizing magnetic ink characters in the MICR device is sufficiently reliable, but may not match the actual magnetic ink characters. .

  This will be specifically described with reference to FIGS. 16 is a perspective view of the image reading apparatus, and FIG. 17 is a block diagram showing an internal configuration of the image reading apparatus shown in FIG.

  As shown in FIGS. 16 and 17, the image reading apparatus 1 includes a placement unit 2, a conveyance path 3, an MICR device 7, and a paper discharge unit 4.

  As shown in FIG. 17, the check placed on the placement unit 2 is conveyed along the conveyance path 3 in the direction of arrow D, and the magnetic ink characters are read as magnetic information by the magnetic head 5 of the MICR device 7. The paper is discharged to the paper discharge unit 4.

  The magnetic ink characters printed on the check are mixed with a magnetic material, so when the check passes the magnetic head 5, the magnetic flux changes according to the shape of the magnetic ink character, and the waveform corresponding to the shape of the character is magnetic. The data is output from the head 5 to the processing unit 7a. Then, the processing unit 7a analyzes the waveform output from the magnetic head 5.

  18A is an external view of “2” that is a part of the E13B font, and FIG. 18B is an external view of “5” that is a part of the E13B font. FIG. 18C is a graph showing waveforms when magnetic data is read by the magnetic head 5 with respect to “2” in FIG. 18A and “5” in FIG.

  In FIG. 18C, the waveform indicated by the solid line is the waveform “2”, the waveform indicated by the dotted line is the waveform “5”, the vertical axis represents the output voltage value, and the horizontal axis represents the passage of time. . The processing unit 7a of the MICR device 7 recognizes the magnetic ink character by analyzing the waveform shown in FIG.

  “2” and “5” shown in FIGS. 18A and 18B are known as characters having very similar waveforms in the E13B font, and “2” shown in FIG. 18C. In the two waveforms “5” and “5”, the only difference is the difference in the time axis between the peak positions (ridges and valleys). The processing unit 7a of the MICR device 7 can compare the time axis differences and correctly recognize the magnetic ink characters.

  The waveforms of “2” and “5” shown in FIG. 18C are examples when the magnetic ink characters pass without being inclined with respect to the magnetic head 5, but depending on the state of the conveyance path 3 and the check. In some cases, the magnetic ink characters pass while being tilted with respect to the magnetic head 5.

  FIG. 19A is an external view when “2” in FIG. 18A is inclined and passes through the magnetic head 5, and FIG. 19B is a part of the E13B font similar to FIG. 18B. It is an external view of "5" which is.

  FIG. 19C is a graph showing a waveform when magnetic data is read by the magnetic head 5 with respect to “2” in FIG. 19A and “5” in FIG. The waveform indicated by “2” is the waveform “2”, and the waveform indicated by the dotted line is the waveform “5”.

  As shown in FIG. 19C, when passing through the magnetic head 5 with “2” tilted, a different waveform from that when passing through the magnetic head 5 without tilting “2” is generated from the magnetic head 5. Is output.

  In the example of FIG. 19C, the waveform when “2” passes through the magnetic head 5 in a tilted state is very similar to the waveform when “5” passes through the magnetic head 5 without tilting. It becomes a waveform.

  The waveform read by the magnetic head 5 is composed of two-dimensional values of “signal level” and “time lapse”. Accordingly, when the situation shown in FIG. 19C occurs, the processing unit 7a of the MICR device 7 can determine whether the waveform is a tilted “2” waveform or a “5” waveform that does not tilt. become unable.

  As a result, while the degree of similarity with dictionary data for recognizing magnetic ink characters by the MICR device 7 is sufficiently reliable, characters that do not actually match the magnetic ink characters printed on the check are erroneously recognized. If you do. In the example of FIG. 19C, the inclined waveform “2” is erroneously recognized as the non-inclined waveform “5”.

  Accordingly, the present invention provides an image reading apparatus, a control method for the image reading apparatus, a control program, and a storage medium that can prevent erroneous recognition of the magnetic ink characters when the magnetic ink characters pass through the magnetic head in a tilted state. The purpose is to provide.

  In order to achieve the above object, an image reading apparatus of the present invention includes a magnetic reading means for reading magnetic recognition data from magnetic ink characters printed on a document, magnetic recognition data read by the magnetic reading means, and MICR dictionary information. And a magnetic character recognition means for recognizing magnetic ink characters printed on a document based on the magnetic similarity obtained by comparing the two, and an image for optically reading the magnetic ink characters printed on the document as image data Reading means; and calculating means for calculating an inclination angle of the magnetic ink character based on image data of the magnetic ink character read by the image reading means, wherein the magnetic character recognition means is printed on a document. The MICR dictionary information used when recognizing a magnetic ink character is changed based on the inclination angle of the magnetic ink character calculated by the calculating means. It is characterized in.

  An image reading apparatus control method according to the present invention includes a magnetic reading unit that reads magnetic recognition data from magnetic ink characters printed on a document, and an image reading unit that optically reads magnetic ink characters printed on a document as image data. A method for controlling an image reading apparatus, comprising: a magnet printed on an original based on a magnetic similarity obtained by comparing magnetic recognition data read using the magnetic reading means and MICR dictionary information; A magnetic character recognizing step for recognizing an ink character, and a calculating step for calculating an inclination angle of the magnetic ink character based on image data of the magnetic ink character read using the image reading means, In the character recognition step, the MICR dictionary information used in recognizing the magnetic ink characters printed on the document is calculated by the magnetic step calculated in the calculation step. Changes based on the inclination angle of the ink characters, and wherein the.

  A control program for an image reading apparatus according to the present invention includes a magnetic reading unit that reads magnetic recognition data from magnetic ink characters printed on a document, and an image reading unit that optically reads magnetic ink characters printed on a document as image data. A control program for an image reading device provided with a magnetic program printed on a document based on a magnetic similarity obtained by comparing magnetic recognition data read using the magnetic reading means with MICR dictionary information Causing a computer to execute a magnetic character recognition step for recognizing an ink character, and a calculation step for calculating an inclination angle of the magnetic ink character based on image data of the magnetic ink character read using the image reading unit. In the magnetic character recognition step, the MICR dictionary information used for recognizing the magnetic ink characters printed on the original is used. Changes based on the inclination angle of the magnetic ink character calculated in serial calculation step, characterized in that.

  The computer-readable storage medium of the present invention stores the control program.

  According to the present invention, it is possible to prevent erroneous recognition of magnetic ink characters when the magnetic ink characters pass through the magnetic head in an inclined state.

1 is a block diagram for explaining an image reading apparatus which is an example of an embodiment of the present invention. FIG. 6 is a flowchart for explaining a process for recognizing magnetic ink characters printed on a check in the image reading apparatus. It is a figure which shows an example of a reading image. It is a figure which shows the list of the character of E13B font. It is a flowchart for demonstrating the inclination angle calculation process of the character in step S103 of FIG. (A) is a figure which shows an example of a template, (b) is an enlarged view of the template of character "1". It is a figure which shows the result of having performed the optical character recognition process with respect to the read image shown in FIG. It is a figure which shows an example of the dictionary data of a MICR dictionary table. The result of comparing “dictionary data used when the character is tilted” and “dictionary data used when the character is not tilted” with respect to one of the dictionary data of each character included in the MICR dictionary table is shown. FIG. It is a flowchart figure for demonstrating the magnetic character recognition process in FIG.2 S105. It is a flowchart figure for demonstrating the cutting-out process of the character position in step S301 of FIG. It is a graph which shows an example of the result of having detected the peak position signal from the digital signal. It is a figure which shows an example of a sample table. It is the graph which displayed the waveform data of the MICR dictionary table, and the waveform data of the sample table in an overlapping manner. It is a flowchart for demonstrating the output process of the recognition result in step S106 of FIG. It is a perspective view for demonstrating the conventional image reading apparatus. It is a block diagram which shows the internal structure of the image reading apparatus shown in FIG. (A) is an external view of “2” that is a part of the E13B font, and (b) is an external view of “5” that is a part of the E13B font. (C) is a graph showing the waveforms when magnetic data is read by a magnetic head for “2” in (a) and “5” in (b). 18A is an external view when “2” in FIG. 18A is tilted and passes through the magnetic head, and FIG. 18B is “5” which is a part of the E13B font similar to FIG. 18B. FIG. (C) is a graph showing the waveforms when magnetic data is read by a magnetic head for “2” in (a) and “5” in (b).

  Hereinafter, an example of an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram for explaining an image reading apparatus which is an example of an embodiment of the present invention.

  As shown in FIG. 1, in the image reading apparatus 1A of the present embodiment, a plurality of checks as an example of a document are placed on a placement unit (not shown), and the plurality of checks placed on the placement unit are The sheets are separated one by one by unshown separation feeding means and fed to the transport path 3.

  An MICR (magnetic character recognition) device 10 and an OCR (optical character recognition) device 20 are arranged in the transport path 3, and the MICR device 10 and the OCR device 20 are connected via a data / address bus 33. It is controlled by the control unit 31. The data / address bus 33 is connected to a memory 34 for storing various data and the external interface 32, and the external interface 32 is connected to a host device 40.

  The MICR device 10 includes a magnetic head (magnetic reading means) 1000, a magnetic data processing unit 11, a MICR dictionary table (MICR dictionary information) storage unit 12, and a MICR processing unit 13.

  The magnetic head 1000 converts magnetic information of magnetic ink characters printed on a check conveyed in the direction of arrow D through the conveyance path 3 into an electric signal. The magnetic data processing unit 11 performs processing such as A / D conversion and smoothing on the electric signal from the magnetic head 1000 and outputs magnetic data to the MICR processing unit 13.

  The MICR processing unit 13 performs magnetic character recognition processing on the magnetic data output from the magnetic data processing unit 11 with reference to the MICR dictionary table stored in the MICR dictionary table storage unit 12.

  The OCR device 20 includes a reading sensor 2000, an image data processing unit 21, an OCR dictionary table storage unit 23, and an OCR processing unit 22.

  The reading sensor 2000 optically reads magnetic ink characters printed on a check conveyed in the direction of arrow D along the conveyance path 3. The image data processing unit 21 performs various processes on the image data read by the reading sensor 2000 and outputs the processed data to the OCR processing unit 22.

  The OCR processing unit 22 performs optical character recognition processing on the image data output from the image data processing unit 21 with reference to the OCR dictionary table stored in the OCR dictionary table storage unit 23.

  Next, a magnetic ink character recognition process printed on a check in the image reading apparatus 1A of the present embodiment will be described with reference to FIG. Each process in FIG. 2 is executed by a CPU or the like of the main control unit 31 by loading a control program stored in a storage unit such as a ROM (not shown) into a RAM (not shown).

  In the following description, a case where a check on which magnetic ink characters are printed is conveyed along the conveyance path 3 while being inclined with respect to the magnetic sensor 1000 and the reading sensor 2000 is taken as an example.

  The magnetic ink characters read by the magnetic head 1000 and the reading sensor 2000 are, for example, E13B font characters “2”, “5”, and “8” as shown in FIG. Represents a part. FIG. 4 shows a list of characters of the E13B font, but the font to be recognized is not particularly limited.

  In FIG. 2, in step S <b> 101, the main control unit 31 reads magnetic recognition data from magnetic ink characters printed on a check conveyed on the conveyance path 3 by the magnetic head 1000 of the MICR device 10.

  The main control unit 31 controls the magnetic data processing unit 11 to perform processing such as A / D conversion and smoothing on the magnetic recognition data read by the magnetic head 1000 and outputs the processed data to the MICR processing unit 13. Then, the process proceeds to step S102.

  In step S <b> 102, the main control unit 31 optically reads the magnetic ink characters printed on the check conveyed on the conveyance path 3 by the reading sensor 2000 of the OCR device 20.

  The main control unit 31 controls the image data processing unit 21 to perform various processes such as shading processing and gamma correction on the image data read by the reading sensor 2000, and outputs the processed data to the OCR processing unit 22. Proceed to step S103.

  In step S103, the main control unit 31 controls the OCR processing unit 22 to refer to the OCR dictionary table for the optical ink character image data output from the image data processing unit 21 in step S102, and to read the optical character. Perform recognition processing.

  The main control unit 31 calculates the tilt angle of the magnetic ink character based on the optical character recognition processing result.

  FIG. 5 is a flowchart for explaining the character inclination angle calculation processing in step S103 of FIG.

  In step S201, the main control unit 31 controls the OCR processing unit 22 to refer to the OCR dictionary table for the magnetic ink character image data output from the image data processing unit 21 in step S102 of FIG. Then, optical character recognition processing is performed, and the process proceeds to step S202.

  Here, the OCR dictionary table stored in the OCR dictionary table storage unit 23 has dictionary data for all E13B font characters, that is, 14 characters.

  The dictionary data included in the OCR dictionary table is a template for image data of each character. Hereinafter, the dictionary data for each character included in the OCR dictionary table is referred to as a template.

  FIG. 6A is a diagram illustrating an example of a template, and FIG. 6B is an enlarged view of the template of the character “1”. In FIG. 6A, as an example, the height direction of the character is set as the x direction, and the direction orthogonal to the x direction is set as the y direction.

  In the process of step S201, optical character recognition processing is performed with reference to this template for the magnetic ink character image data output from the image data processing unit 21 in step S102.

  Specifically, starting from the upper left coordinates of the acquired image data, the image for the width and height of the template in the OCR dictionary table is compared with all the templates in the OCR dictionary table, and the pixel mismatch rate is counted. .

  The comparison between the template and the OCR dictionary table is then repeatedly performed by shifting one pixel at a time in the vertical and horizontal directions, and is performed for all the pixels of the image data.

  If a template having a mismatch rate count smaller than a predetermined number exists, the position is recognized as a character position, and a recognition result is output.

  For example, the result of performing the above optical character recognition process on the read image shown in FIG. 3 is as shown in FIG.

  A range surrounded by a rectangle R in FIG. 7A is a range actually recognized as a character. FIG. 7B shows a recognition result finally output by the OCR processing unit 22. As shown in FIG. 7B, the optical character recognition result by the OCR processing unit 22 is composed of the recognition result character, the mismatch rate with the template, and the recognized position (upper left coordinate, lower right coordinate), and is acquired here. The recognized result is stored in the memory 34.

  In step S202, the main control unit 31 calculates the inclination angle of the magnetic ink character based on the recognition result calculated in step S201.

  In general, since magnetic ink character strings are arranged at the same height in the horizontal direction of the document, if the inclination angle of the magnetic ink characters is S, S is, for example, the upper left coordinate information and the second character. Can be calculated by the following equation (1) using the upper left coordinate information.

Inclination angle S (absolute value) of magnetic ink character = | (first character y coordinate-2 character y coordinate) / (first character x coordinate-2 character x coordinate) | (1)
Then, the main control unit 31 writes the inclination angle S of the magnetic ink character calculated by the above equation (1) in the memory 34, and proceeds to step S104 in FIG.

  Returning to FIG. 2, in step S104, the main control unit 31 controls the MICR processing unit 13 to create an MICR dictionary table corresponding to the character inclination angle calculated in step S103 from the MICR dictionary table storage unit 12. select.

  As described above, the waveform when the magnetic ink character string passes through the magnetic head 1000 with the magnetic ink character string tilted is different from the waveform when the magnetic ink character string passes through the magnetic head 1000 without being tilted.

  Accordingly, the MICR dictionary table storage unit 12 has two types of reference data for all characters of the E13B font: “reference data used when characters are tilted” and “reference data used when characters are not tilted”. The MICR dictionary table is stored. In the present embodiment, the case where two types of MICR dictionary tables are used is illustrated for convenience, but three or more types of MICR dictionary tables may be used.

  Here, the reference data is dictionary data having two elements of “peak position time” and “amplitude level”, and FIG. 8 shows an example of the dictionary data.

  Further, FIG. 9 shows “dictionary data used when the character is tilted” and “dictionary data used when the character is not tilted” for one of the dictionary data of each character included in the MICR dictionary table. The result of having compared is shown.

  FIG. 9 is a graph schematically showing the time and amplitude level of the peak position possessed by the dictionary data as waveforms. The horizontal axis represents the passage of time and the vertical axis represents the amplitude level. Data indicated by solid lines is “dictionary data used when characters are not tilted”, and data indicated by dotted lines is “dictionary data used when characters are tilted”.

  The MICR processing unit 13 reads the character inclination angle calculated in step S103 from the memory 34. Then, if the inclination angle of the read character is equal to or smaller than a predetermined threshold, the MICR processing unit 13 selects the MICR dictionary table including “dictionary data used when the character is not inclined” from the MICR dictionary table storage unit 12. .

  Further, if the inclination angle of the read character exceeds a predetermined threshold, the MICR processing unit 13 selects the MICR dictionary table including “dictionary data used when the character is inclined” from the MICR dictionary table storage unit 12. Then, the process proceeds to step S105 in FIG.

  The dictionary data shown in FIGS. 8 and 9 are created with a peak number of 4. However, the dictionary data is not limited to this, and the dictionary data includes an appropriate peak number for each character that the MICR processing unit 13 can recognize. Can be created.

  Returning to FIG. 2, in step S105, the main control unit 31 controls the MICR processing unit 13, and for the magnetic data output from the magnetic data processing unit 11 in step S101, the MICR selected in step S104. Magnetic character recognition processing is performed with reference to the dictionary table.

  FIG. 10 is a flowchart for explaining the magnetic character recognition processing in step S105 of FIG.

  10, in step S301, the main control unit 31 controls the MICR processing unit 13 to cut out the magnetic character position from the magnetic data output from the magnetic data processing unit 11 in step S101 of FIG.

  The character position is cut out by detecting the character start position. In order to detect the character start position, first, the peak of the sampling waveform is detected from the magnetic data. Next, the character start position is obtained based on the peak position detected first, and a predetermined range is cut out as data for one character.

  With reference to FIG. 11, the character position clipping process in step S301 in FIG. 10 will be described. In FIG. 11, D is sampling data. D (X) is sampling data at the position of X. R and K are constants uniquely determined by the character font and sampling rate. N is a character count indicating the number of characters currently. Length is the length of the sampling data. Also, the position corresponding to the right end portion of the check is set to 0.

  In step S401, the main control unit 31 controls the MICR processing unit 13 to reset X and N to 0 in order to scan magnetic data from the right end of the check, and proceeds to step S402.

  In step S402, the main control unit 31 controls the MICR processing unit 13 to compare D (X) and K. If D (X) exceeds K, the main control unit 31 proceeds to step S404, and D (X ) Is K or less, the process proceeds to step S403.

  In step S404, the main control unit 31 controls the MICR processing unit 13 to compare D (X) and D (X + 1). If D (X) exceeds D (X + 1), step S405 is performed. If D (X) is equal to or less than D (X + 1), the process proceeds to step S403.

  In step S405, the main control unit 31 determines that D (X) is the first vertex position in the waveform of the magnetic data, and controls the MICR processing unit 13 so that the data from X to X + R is the Nth character. The position information is stored in the memory 34, and the process proceeds to step S406.

  In step S406, the main control unit 31 controls the MICR processing unit 13, advances X by R (X = X + R), counts N by 1 (N = N + 1), and proceeds to step S407.

  On the other hand, in step S403, the main control unit 31 controls the MICR processing unit 13 to count one X (X = X + 1), and proceeds to step S407.

  In step S407, the main control unit 31 controls the MICR processing unit 13 to determine whether X is less than Length. If X is less than Length, the process returns to step S402, and if X is greater than or equal to Length, Then, the character cutout process is terminated, and the process proceeds to step S302 in FIG.

  Returning to FIG. 10, in step S302, the main control unit 31 controls the MICR processing unit 13 to reset the character count N to 0, and proceeds to step S303.

  In step S303, the main control unit 31 controls the MICR processing unit 13 to acquire the Nth character magnetic data, and proceeds to step S304. Here, the MICR processing unit 13 refers to the position information data stored in the memory 34 and cuts out the waveform of the portion corresponding to the position information data in the magnetic data.

  In step S304, the main control unit 31 controls the MICR processing unit 13 to detect a peak position signal for the N-th magnetic data and create peak position data, and the process proceeds to step S305.

  The E13B font, which is a font for magnetic ink characters in the present embodiment, has a peak read signal obtained when characters are read by the magnetic head 1000 when the width of one character is divided into eight equal parts with respect to the time axis. Are configured to exist in different blocks.

  Therefore, the MICR processing unit 13 divides the character into eight equal parts, detects the peak position of the magnetic signal, and calculates the detected peak position as the peak position signal. Furthermore, the MICR processing unit 13 detects the amplitude level for each peak position, and calculates peak value data indicating the time and amplitude level at the peak position.

  As a method of detecting the peak position, for example, there is a method of detecting, as the peak position, a point where the amplitude level of the sampling waveform has changed from increase to decrease and a point where the amplitude level of the sampling waveform has changed from decrease to increase.

  Note that the peak position detection method may be in other forms. For example, for the magnetic data waveform shown in FIG. 12 (a), the signal resulting from detecting the peak position by the above method is as shown in FIG. 12 (b), and the peak value data is shown in FIG. 12 (c). As shown.

  In step S305, the main control unit 31 controls the MICR processing unit 13, and based on the peak value data obtained in step S304, a sample representing the relationship between the peak position time and the amplitude level (polarity) in a table A table is created and the process proceeds to step S306.

  FIG. 13 shows an example of the sample table. In FIG. 13, for convenience, it is assumed that four peak positions of the N-th magnetic data are detected, but the present invention is not limited to this.

  In step S306, the main control unit 31 controls the MICR processing unit 13 to compare the dictionary data in the MICR dictionary table selected in step S104 with the data in the sample table created in step S305. The comparison processing here is performed by calculating two-dimensional distances between the character peaks of the sample table and all the MICR dictionary tables.

  FIG. 14 is a diagram in which the waveform data of the MICR dictionary table and the waveform data of the sample table are displayed in an overlapping manner.

  In FIG. 14, the vertical axis represents amplitude, the horizontal axis represents time, the waveform represented by a solid line represents the waveform data of the sample table, and the waveform represented by a dotted line represents the waveform data of the MICR dictionary table.

  Here, a comparison process between the sample table shown in FIG. 13 and the MICR dictionary table shown in FIG. 8 will be described.

  The MICR processing unit 13 calculates the two-dimensional distances L0 to L3 of the four terms in the MICR dictionary table with respect to the first term (St_0, Sl_0) of the sample table shown in FIG.

  As shown in FIG. 14, the distance L0 is a distance from the first peak S_peak0 of the sample table waveform to the first peak D_peak0 of the MICR dictionary table waveform. The distance L1 is a distance from the first peak S_peak0 of the sample table waveform to the second peak D_peak1 of the MICR dictionary table waveform.

  The distance L2 is a distance from the first peak S_peak0 of the sample table waveform to the third peak D_peak2 of the MICR dictionary table waveform. The distance L3 is a distance from the first peak S_peak0 of the sample table waveform to the fourth peak D_peak3 of the MICR dictionary table waveform.

These distances L0 to L3 are calculated by the following equations.
L0 = {A × (St — 0−Dt — 0) ² + B × (S1 — 0—D1 — 0) ² } ^1/2
L1 = {A × (St — 0−Dt — 1) ² + B × (Sl — 0—Dl — 1) ² } ^1/2
L2 = {A × (St — 0−Dt — ² ) ² + B × (Sl — 0—Dl — ² ) ² } ^1/2
L3 = {A × (St — 0−Dt — 3) ² + B × (Sl — 0—Dl — 3) ² } ^1/2
(A and B are predetermined constants)
When obtaining the distance using the above equation, it is preferable to appropriately adjust the waveform data so that the aspect ratio of the waveform data does not become extremely long or long.

  Next, the MICR processing unit 13 determines, based on the calculated distance, the brute force which is closest to the first term in the sample table among the four terms in the MICR dictionary table. Memorize near peak distance.

  The term in the dictionary table determined to be closest to the first sample data is omitted thereafter. Then, for the second term (St_1, Sl_1), the third term (St_2, Sl_2),... In the sample table, the two-dimensional distance of the terms in the MICR dictionary table is calculated in the same way, It is judged whether it is near.

  Then, the distance between the nearest peaks is determined up to the fourth term (St_3, S1_3), and the distance to the term determined to be closest to each sample data is added, and the total value is obtained. Is stored in the memory 34.

  Similarly, the sum of the distances determined to be closest such as the next character in the MICR dictionary table and the next character in the sample table of FIG. 13 is calculated, stored in the memory 34, and the process proceeds to step S307.

  In step S307, the main control unit 31 controls the MICR processing unit 13 to determine a recognition result character from the sum of the distances calculated in step S306. Here, the main control unit 31 selects a character in the dictionary table that minimizes the sum of the distances calculated in step S306 as a character having a high magnetic similarity as a recognition result character.

  Then, the main control unit 31 associates the selected character with the total distance (hereinafter referred to as “score”) when the character is recognized, stores it in the memory 34, and proceeds to step S308.

  In step S308, the main control unit 31 controls the MICR processing unit 13 to determine whether the character count N is less than all the extracted characters. If N is less than the number of characters, it is determined that the data is similar to the dictionary data, and the process proceeds to step S309. If N is greater than or equal to the number of characters, the process proceeds to step S106 in FIG.

  In step S309, the main control unit 31 controls the MICR processing unit 13, counts N by 1 (N = N + 1), and returns to step S303.

  Returning to FIG. 2, in step S <b> 106, the main control unit 31 outputs the recognition result in the magnetic ink character recognition process to the host device 40.

  FIG. 15 is a flowchart for explaining the recognition result output process in step S106 of FIG. In FIG. 15, C is a predetermined threshold, N is a character count, and Length is the length of a character string.

  In step S501, the main control unit 31 resets the character count N to 0, and proceeds to step S502.

  In step S502, the main control unit 31 obtains “score (sum of distances)” in the Nth character recognition result from the memory 34, and proceeds to step S503. Here, the “score (total distance)” represents the degree of difference between the sampling data and the MICR dictionary data.

  In step S <b> 503, the main control unit 31 determines whether or not the “score (total distance)” of the Nth character exceeds the threshold value C.

  When the “score (total distance)” of the Nth character is equal to or less than the threshold C, the main control unit 31 determines that the magnetic ink character recognition result in the MICR processing unit 13 is reliable, and the process proceeds to step S505. move on.

  On the other hand, if the “score (total distance)” of the Nth character exceeds the threshold value C, the main control unit 31 determines that the magnetic ink character recognition result in the MICR processing unit 13 is unreliable, and step S504. Proceed to

  In step S504, the main control unit 31 changes the magnetic ink character recognition result to “unread” indicating that the character is an unrecognizable character, and proceeds to step S505.

  In step S505, the main control unit 31 determines whether or not the character count N has reached the length Length of the character string.

  When the character count N reaches the length of the character string, the main control unit 31 transmits the magnetic ink character recognition result to the host device 40 via the external interface 32, and ends the process. On the other hand, if the character count N has not reached the length Length of the character string, the main control unit 31 proceeds to step S506.

  In step S506, the main control unit 31 increments the character count N by 1 (N = N + 1), returns to step S502, and repeats the processing.

  As described above, in this embodiment, the tilt angle of the magnetic ink character is calculated based on the image data read by the reading sensor 2000. Then, the MICR dictionary table having dictionary data suitable for the calculated inclination angle of the magnetic ink character is selected from the MICR dictionary table storage unit 12, and the magnetic ink character recognition process is performed using the MICR dictionary table having the selected dictionary data. Do. Thereby, it is possible to prevent erroneous recognition of the magnetic ink characters when the magnetic ink characters pass through the magnetic head 1000 in a tilted state.

  In this embodiment, the character position information obtained by the optical character recognition process is used only for calculating the inclination of the magnetic ink character. However, the character recognized by the optical character recognition process is recognized by the magnetic ink character. It is also possible to output as a result.

  For example, if the “score” exceeds the threshold value C in step S503 among the magnetic ink characters recognized by the MICR processing unit 13, the recognition result is not changed to “unread” in step S504. May be changed to a recognized character by optical character recognition processing.

  In this embodiment, the position information of each character recognized by the optical character recognition process is used for calculating the tilt angle. However, the present invention is not limited to this. For example, the information on the edge of the document included in the image data is used. The tilt angle of the image may be calculated.

  Further, the image is scanned in the vertical direction from the lower end of the image of the document, and the point where the black pixel (the pixel constituting the magnetic ink character) appears for the first time is detected at a plurality of positions in the horizontal direction. And the above equation (1) may be used to calculate the tilt angle of the character.

  Furthermore, in the present embodiment, separate MICR dictionary tables are used for the case where the magnetic ink characters are inclined and the case where they are not inclined, but the present invention is not limited to this.

  For example, a single MICR dictionary table may be provided, and the peak position of dictionary data in the MICR dictionary table may be changed and adjusted in accordance with the inclination angle of characters. In this way, memory usage can be reduced.

  Further, in this embodiment, the inclination angle of the magnetic ink character is calculated as one value for all the characters of the check. However, the inclination angle is calculated separately for each adjacent character of the check, It may be determined whether or not each character is inclined. If it does in this way, the further improvement of the recognition rate in recognition processing of a magnetic ink character can be expected. Alternatively, optical character recognition may be performed only by detecting inclination, and character determination by optical character recognition may be performed again for a character having a poor MICR recognition result score.

  In addition, this invention is not limited to what was illustrated to the said embodiment, In the range which does not deviate from the summary of this invention, it can change suitably.

  For example, part or all of the magnetic character recognition and optical character recognition processing of the present invention may be performed by an external device such as a host device. In this case, a system including an image reading device and an external device such as a host device corresponds to the image reading device of the present invention. Also, part or all of the processing of the present invention may be changed to processing by hardware.

  The object of the present invention can also be achieved by executing the following processing. That is, a storage medium in which a program code of software that realizes the functions of the above-described embodiments is supplied to a system or apparatus, and a computer (or CPU or MPU) of the system or apparatus is stored in the storage medium. This is the process of reading the code.

  In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the program code and the storage medium storing the program code constitute the present invention.

  Moreover, the following can be used as a storage medium for supplying the program code. For example, floppy (registered trademark) disk, hard disk, magneto-optical disk, CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD + RW, magnetic tape, nonvolatile memory card, ROM or the like. Alternatively, the program code may be downloaded via a network.

  Further, the present invention includes a case where the function of the above-described embodiment is realized by executing the program code read by the computer. In addition, an OS (operating system) running on the computer performs part or all of the actual processing based on an instruction of the program code, and the functions of the above-described embodiments are realized by the processing. Is also included.

  Furthermore, a case where the functions of the above-described embodiment are realized by the following processing is also included in the present invention. That is, the program code read from the storage medium is written in a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer. Thereafter, based on the instruction of the program code, the CPU or the like provided in the function expansion board or function expansion unit performs part or all of the actual processing.

1A Image reader 3 Transport path 10 MICR device 11 Magnetic data processing unit 12 MICR dictionary table storage unit 13 MICR processing unit 20 OCR device 21 Image data processing unit 22 OCR processing unit 23 OCR dictionary table storage unit 31 Main control unit 32 External interface 33 Address bus 34 Memory 40 Host device 1000 Magnetic head 2000 Read sensor

Claims (10)

Magnetic reading means for reading magnetic recognition data from magnetic ink characters printed on a document;
Magnetic character recognition means for recognizing magnetic ink characters printed on a document based on magnetic similarity obtained by comparing magnetic recognition data read by the magnetic reading means and MICR dictionary information;
Image reading means for optically reading magnetic ink characters printed on a document as image data;
Calculating means for calculating an inclination angle of the magnetic ink character based on image data of the magnetic ink character read by the image reading means;
The magnetic character recognizing unit changes the MICR dictionary information used when recognizing the magnetic ink character printed on the document based on the inclination angle of the magnetic ink character calculated by the calculating unit;
An image reading apparatus. A storage unit for storing a plurality of types of MICR dictionary information corresponding to the inclination angle of the magnetic ink characters;
The magnetic character recognizing unit recognizes a magnetic ink character printed on a document from a plurality of types of MICR dictionary information stored in the storage unit, based on the inclination angle of the magnetic ink character calculated by the calculating unit. Selecting the MICR dictionary information to be used when
The image reading apparatus according to claim 1. Based on the magnitude of the magnetic similarity, the magnetic character recognition unit includes a determination unit that determines whether to replace the magnetic ink character recognition result with unread.
The image reading apparatus according to claim 1, wherein The calculating unit calculates an inclination angle of the image from information on an edge of the document included in the image data of the magnetic ink characters read by the image reading unit;
The image reading apparatus according to claim 1, wherein the image reading apparatus is an image reading apparatus. Optical character recognition means for optically recognizing magnetic ink characters contained in the image data read by the image reading means;
The image reading apparatus according to claim 1, wherein Based on the magnitude of the magnetic similarity, there is provided a judging means for judging whether or not to replace the magnetic ink character recognition result by the magnetic character recognition means with the magnetic ink character recognition result by the optical character recognition means. ,
The image reading apparatus according to claim 5. The calculating unit calculates the inclination of the magnetic ink character printed on the document based on the position information of the character recognized by the optical character recognizing unit;
The image reading apparatus according to claim 5, wherein the image reading apparatus is an image reading apparatus. A control method for an image reading apparatus, comprising: a magnetic reading unit that reads magnetic recognition data from magnetic ink characters printed on a document; and an image reading unit that optically reads magnetic ink characters printed on a document as image data. ,
A magnetic character recognition step for recognizing magnetic ink characters printed on a document based on magnetic similarity obtained by comparing magnetic recognition data read using the magnetic reading means and MICR dictionary information;
A calculation step of calculating an inclination angle of the magnetic ink character based on image data of the magnetic ink character read using the image reading unit,
In the magnetic character recognition step, the MICR dictionary information used when recognizing the magnetic ink characters printed on the original is changed based on the inclination angle of the magnetic ink characters calculated in the calculation step.
A control method for an image reading apparatus. A control program for an image reading apparatus comprising magnetic reading means for reading magnetic recognition data from magnetic ink characters printed on a document, and image reading means for optically reading magnetic ink characters printed on a document as image data. ,
A magnetic character recognition step for recognizing magnetic ink characters printed on a document based on magnetic similarity obtained by comparing magnetic recognition data read using the magnetic reading means and MICR dictionary information;
Calculating a tilt angle of the magnetic ink character based on the image data of the magnetic ink character read using the image reading unit;
In the magnetic character recognition step, the MICR dictionary information used when recognizing the magnetic ink characters printed on the original is changed based on the inclination angle of the magnetic ink characters calculated in the calculation step.
A control program for an image reading apparatus. The control program according to claim 9 is stored.
A computer-readable storage medium characterized by the above.