JPH09128483A - Character recognizing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method flexibly recognizing characters on an optimum condition by means of a single character recognition device with respect to even various characters formed in a card. SOLUTION: A character is recognized by controlling a carrying means 2 carrying the card to a prescribed reading position, an illuminating means 3 illuminating the character and a picture processing means 5 picture-processing picture data obtained by a character read by a reading means 4 by a recognizing method suited to the character on the card, based on initializing data including information on the division of a character forming method, a character attribute and character arrangement which are fixed in a series of card group to be recognizing objects. The character forming division is the division of printing character such as an embossed character, a marking character, an OCR character, the character attribute is a lettering, the size and color of the character, etc., and the arrangement of the character is the start position of the character, character pitch, intervals between characters, the total number of characters, the ending position of the character, etc.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a magnetic card and an IC.
The present invention relates to a method for reading characters formed on various cards such as cards by printing or embossing, and more particularly to a method for recognizing characters with different forming methods by an optimum recognition method.

[0002]

2. Description of the Related Art In recent years, various cards have been used in large amounts for various purposes, and characters such as ID numbers and names are formed on these cards by printing, embossing, engraving or the like. Such characters have a meaning for confirmation by the card user, but may be mechanically read by a character recognition device. For example, when manufacturing a new magnetic card or IC card, first form an ID number etc. with embossed characters,
Then, the ID number of the embossed character is optically read, and the data corresponding to the read ID number is written on the card. This is because it is possible to reliably prevent the mismatch between the embossed character and the data to be recorded. Further, even at the use stage of the card, the card reader double-reads the ID number from the magnetic data and the embossed character to deal with alteration of the magnetic data and the embossed character.

[0003]

By the way, in the card issuing device, etc., basically the same issuing device can be used for the same kind of card such as a magnetic card and an IC card, but an ID number or The method of forming letters such as names differs depending on the issuing company that decides the design of the card, and is not constant. For this reason, the character recognition device installed in the issuing device needs to change the character recognition device for each character such as embossed characters, engraved characters, printed characters such as OCR, etc. and different forming methods. One card issuing device equipped with a recognition device could not handle all cards of the same type. Further, for example, even if it is a simple printed character, there are various fonts, sizes, colors, arrangements of characters, etc., and no one character recognition device can deal with them. Therefore, the present invention solves the above-mentioned problems and provides a character recognition method capable of flexibly recognizing various characters formed on a card with a single character recognition device under optimal conditions. With the goal.

[0004]

In the character recognition method of the present invention, a character formed on a card is optically read and the character is recognized by image processing. Of the character forming method, the attribute of the character, and the initial setting data including the information about the arrangement of the character, the conveying means for conveying the card to a predetermined reading position, the illuminating means for illuminating the character of the card, and the reading. Character recognition is performed by controlling the recognition method according to the character of the card to be read by the image processing means for image-processing the image data obtained by reading the character with the means.

Further, in the above-mentioned character recognition method, the information regarding the classification according to the character formation method is at least information for distinguishing embossed characters, engraved characters, OCR characters, and print characters other than OCR. is there. Further, in each of the above character recognition methods, it is also a character recognition method in which the information regarding the attribute of a character has at least one of a font, a character size, and a character color. Further, in each of the above character recognition methods, the information regarding the arrangement of characters is also a character recognition method in which at least information regarding the character start position on the card, the character pitch or character spacing, the total number of characters or the character end position, is included. .

[0006]

BEST MODE FOR CARRYING OUT THE INVENTION The character recognition method of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a schematic block diagram of a character recognition device for performing the character recognition method of the present invention.
FIG. 1 is a schematic device configuration diagram of the character recognition device. In FIG. 1, the character recognition device 10 is composed of a conveying unit 2, an illuminating unit 3, a reading unit 4, an image processing unit 5, and an arithmetic control unit 6. In FIG. 2, the character recognition device includes a character recognition device main body 11 and a carrier device 1 connected to the character recognition device main body 11 for carrying the card 40 to an appropriate position.
2. Illumination devices 13a to 13c, CCD camera 14 and the like. The host computer 20 provides the character recognition device body 11 with initial setting data common to a series of cards to be recognized. In the figure, the character recognition device used for the card issuing device is assumed, and the host computer 2
0 receives a card ID number as character information from the character recognition device and sends write data corresponding to the received card ID number to the card reader / writer 30 to instruct writing to the card.

The conveying means 2 comprises a drive mechanism for appropriately conveying the card to a predetermined character reading position, and when the reading area does not cover the entire surface of the card, it is divided into, for example, three parts in the card feeding direction and intermittently conveyed. However, the character to be recognized can be placed within the reading range of the reading means.

The illuminating means 3 is configured to illuminate the card surface from at least two directions so that the characters can be read by "specular reflection reading" or "diffuse reflection reading" described later. For example, FIG. The lighting device illustrated in
The illuminating devices 13a to 13c can illuminate the card surface from three directions. Such illuminating means is, for example, an LED (Light Emitting Diode) which is a point light source.
e) is a two-dimensionally arranged planar light source, and LED light source rows corresponding to a desired illumination direction are selectively turned on, or linear light sources are one-dimensionally arranged so that each linear light source is It can be realized by selectively lighting. The reading unit 4 is a CCD camera or the like, and by providing a plurality of units, it is possible to read the characters on the card from different directions. For example, reading the characters directly above or obliquely to the card surface, or reading the front and back sides of the card. You can If the field of view of one camera cannot cover the entire card surface, multiple cameras will share it. If the shared area is in the card transport direction, the card may be transported and the card surface within the field of view may be moved for reading.

Here, in FIG. 7, the illumination devices 13a to 13a are provided.
Details of the positional relationship between c and the CCD camera 14 are shown. The light beam emitted from the illumination device 13a is specularly reflected by the area Aa on the card 40 and enters the camera 14. Similarly, the light ray from the illuminating device 13b is specularly reflected by the area Ab, and the light ray from the illuminating device 13c is specularly reflected by the area Ac and enters the camera 14. Where, for example,
If only the lighting device 13b is turned on, the characters in the area Ab are read by the camera by the light rays that are specularly reflected, and the area Aa
The characters in the area Ac are read by the camera by the diffuse reflection light from the light beam spreading from the lighting device 13b. Further, the characters in the region Ab are the lighting device 13a or 13
c, or the lighting devices 13a and 13c are turned on, and the reading is performed by the diffuse reflection light by the light rays spreading from these lighting devices. In this manner, which of the plurality of lighting devices is to be turned on is determined depending on the position of the character to be read on the card and the reading method by the mirror surface or diffusion, and optimal character recognition is performed.

The image processing means 5 appropriately performs image processing such as image enhancement, contour line extraction, and smoothing on the image data read by the reading means to extract binarized character image data. The arithmetic control unit 6 specifies the character by comparing the character image data with reference character data which is a reference stored in advance and obtains the character data.

Next, the character recognition method of the present invention will be described by taking the above character recognition device as an example. First, FIG. 3 is an example showing the arrangement, size, etc. of the character C formed on the card 40.
Depending on the method of forming the card, the characters may be embossed by embossing, engraved by engraving, or O by printing.
It is formed as a CR character or a normal print character. Then, the initial setting data used in the present invention has information about the division according to the method of forming these characters. that is,
The embossed characters have irregularities, the engraved characters are smaller than the embossed characters but have irregularities in the outline of the characters, and the printed characters are flat, and due to this surface irregularity characteristic, the characters are illuminated when the characters are read by the camera. Because it changes the method.

The embossed character is specularly reflected (specularly reflected) at the character portion to be read, and only the illumination device at the position where the line-reflected light is exactly incident on the camera is turned on. As a result, FIG.
As shown in FIG. 5, the convex portion formed by the embossed character C1 parallel to the card surface receives specular reflection light on the camera, and the portion not parallel to the card surface does not enter specular reflection light on the camera. Then, character recognition is performed from the image data obtained by reading the brightness. This is called "specular reflection reading". In the engraved character, as shown in FIG. 6, there is a step (slope) in the contour portion of the engraved character C2, and the step becomes a dark portion, and the contour of the character is obtained by specular reflection reading which is read from the regular reflection direction of the illumination light like the embossed character. The image data in which is emphasized is obtained and characters are recognized. Alternatively, the illumination is performed by diffuse reflection illumination in which the light is reflected from the card surface at an angle that does not enter the camera and character recognition is performed by obtaining image data based on the density of the character. Note that the latter diffuse reflection illumination is referred to as "diffuse reflection reading". In case of engraved character, which one to use,
For example, specular reflection reading is performed when the contrast between characters and the surroundings is low. In the case of printed characters such as OCR, diffuse reflection reading is performed. Since there is no unevenness in the printed characters, image data having a contrast according to the character shape cannot be obtained in the reading stage in specular reflection reading. In addition,
In the description of the present invention, as shown in FIGS. 2 and 7, the illumination device to be turned on is replaced with the camera that looks at the card surface from a certain direction, and specular reflection reading and diffuse reflection reading are controlled. It is also possible to control the specular reflection reading and the diffuse reflection reading by using different cameras (for example, the camera seen from the normal direction is the diffuse reflection reading, and the camera seen from the oblique direction is the specular reflection reading). However, it is possible to make the device compact by controlling only one camera on the lighting device side.

As described above, by specifying the division according to the character forming method as the initial setting data, the "specular reflection reading" or "diffuse reflection reading" and the reading method by illumination are determined. However, the engraved character also depends on the contrast with the background as described above.

Next, as the initial setting data relating to the character attributes, there are a character typeface, a character size, a character color and the like.
The character typeface means, for example, a typeface such as Mincho typeface or Gothic typeface. The character size also includes information such that the number of points, the number of items, or the vertical and horizontal dimensions are independently designated in units of mm, for example. The color of the characters is especially engraved characters and OCR
It is the color of the characters in the printed characters such as. For example, there are various stamp characters such as black, gold, and blue. The same applies to print characters. The reference character data to be compared and collated differs depending on the typeface and the character size of the character, and a binarization threshold value for binarizing the image read by the camera depending on the character color and (diffuse reflected light for reading the character color is obtained. In order to perform optimal image input and processing, the lighting pattern of the lighting device that is turned on is controlled. There are various known typefaces for printing other than the Mincho typeface and the Gothic typeface, for example. By giving information about these various typefaces to the character recognition device as initial setting data, any typeface can be used. In addition, in the case of the engraved characters of gold or silver, both of the lighting device for specular reflection reading and the lighting device for diffuse reflection reading are turned on. This is because gold and silver colors strongly reflect light, and in the case of specular reflection reading, the contrast of the dark part in the contour of the engraved character will decrease, so the contour portion will be reflected by the light beam of diffuse reflection reading. This is because the light is specularly reflected to form the bright portion. In this case, the influence of the design is more likely to occur than in the specular reflection reading, but since the reflection of the gold / silver color character portion is sufficiently strong, it is easier to recognize.

Next, the initial setting data relating to the arrangement of the characters on the card is the area on the card where the character to be read is present, and further information for specifying the character to be read. As shown in the card 40 of FIG. 3, the start position L of the character C
(X1, y1), character pitch P1, character spacing P2, total number of characters n, character end position L (x2, y1), and the like. If the character size is determined by the information about the character attribute, the size of the character C in the x direction is determined.
Either the character pitch P1 or the character pitch P1 is good. Also, the character end position can be derived from the character start position, the character interval, and the total number of characters, while the total character number can be derived from the character start position, the character interval, and the character end position. Therefore, either one of the total number of characters and the character end position may be used. In this way, when two or more good ones are designated redundantly, the character recognition device may check the consistency to see if the designations of both are consistent. Moreover, since the character size in the y direction is determined from the character size of the character attribute, the height H in the y direction is determined. Also,
In FIG. 3, there is an "x" mark in the third character C (due to the convenience of the drawing description rather than the actual x mark), which indicates that the third character is not a recognition target, and recognition is unnecessary. It may include information that specifically specifies the letter E as the letter. If the characters are arranged upside down, the information of upside down is specified.

With the above-mentioned information regarding the arrangement of the characters, the character recognition device can determine which area of the character on the card should be read for each series of cards.

Then, in order to read a predetermined area of the card with a reading means such as a CCD camera, the card is conveyed to a predetermined reading position by the conveying device so that the target character is contained within the visual field of the camera. To However, it is possible to read a normal horizontal card with one camera as the entire field of view of the camera, but it is better to read one card surface separately for the resolution of the image data. Is advantageous. Therefore, for example, as illustrated in FIG. 4, the vertical field of view of the camera covers about ⅔ of the card, and the horizontal direction covers about ⅓ of the card. Vum, Vur, Vdl, Vdm,
The Vdl is divided into six, and two cameras are vertically arranged in parallel (in the vertical direction) to cover, and in the left and right directions, the card is skipped three times by the conveying means to cover the entire width of the card in the horizontal direction. Control. Of course, a necessary area of each of these visual field areas may be read according to the recognized character area, and the division size and position may be controlled to an optimum value depending on the character formation area. In this way, if the card is held up and down by two cameras, and this is provided on the front and back of the card, two cameras are provided on each of the front and back sides, for a total of four cameras (and a lighting device therefor). ), It is possible to recognize characters by covering the entire front and back of the card.

The image data read by the reading means is subjected to image processing such as appropriate filtering by the image processing means, and character image data is extracted. The image processing method also depends on the character formation classification. Since they are different, the image processing means is also made to perform the image processing by selecting the image processing method designated in advance by the information regarding the character classification given in the initial setting data. If it is an embossed character or an engraved character, an image is input by specular reflection reading, edge enhancement processing, etc. are performed, and then binarization is performed. When a printed character such as OCR is read by diffuse reflection, an image is input by diffused light, smoothing, noise removal, etc. are performed, and then binarization is performed to obtain character image data.

In the character image data extracted as described above, the arithmetic control means selects the reference character data, which is a reference for comparison registered in advance, from the character attributes designated by the initial setting data. Are compared and collated, the characters are specified, and finally the characters are recognized. If the type of the read target character is a specific one such as numbers, letters, kana, kana, kanji, etc., it is advisable to include the character type in the initial setting data. Thereby, the number of reference character data to be compared and collated can be limited in advance, and the character recognition rate and the character recognition speed are improved.

Further, in the above description, the arithmetic control means, when given the initial setting data from the outside, also appropriately controls the conveying means, the reading means and the image processing means based on this.

The character recognition method of the present invention is not limited to the above description without departing from the spirit of the invention, and may be determined in advance according to the character formation classification, the character attribute, and the character arrangement. The character recognition method is taught to the character recognition device by designating initial setting data, and the character recognition is performed by an optimum method. Therefore, for example, when the character recognition device is used as a part of the card issuing device, the carrying means can be omitted on the character recognition device side as the carrying means included in the card issuing device. Further, the arithmetic control means may be divided into one for comparing and identifying the characters and one for controlling the other means.

[0022]

According to the character recognition method of the present invention, in character recognition for a series of cards, even when different types of characters such as different formation methods are read, one character recognition device can flexibly and Character recognition can be performed under optimal recognition conditions.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram showing an example of a character recognition device for performing a character recognition method of the present invention.

FIG. 2 is a schematic device configuration diagram showing an example of the character recognition device of FIG.

FIG. 3 is an explanatory view of arrangement of characters formed on a card.

FIG. 4 is an explanatory diagram of a visual field area of a camera that reads a card.

FIG. 5 is an explanatory diagram of specular reflection reading for reading embossed characters.

FIG. 6 is an explanatory diagram of specular reflection reading for reading engraved characters.

FIG. 7 is an explanatory diagram of specular reflection reading and diffuse reflection reading by controlling an illumination method.

[Explanation of symbols]

 2 conveying means 3 illuminating means 4 reading means 5 image processing means 6 arithmetic control means 10 character recognizing device 11 character recognizing device main body 12 conveying device 13a, 13b, 13c illuminating device 14 CCD camera 20 host computer 30 card reader / writer 40 card C character Aa, Ab, Ac Area on the card (with characters) C1 Embossed character C2 Engraved character E Unrecognized character H Height L (x1, y1) Character start position L (x2, y2) Character end position P1 Character pitch P2 character Interval Vul camera view area (upper left area) Vum camera view area (upper center area) Vur camera view area (upper right area) Vdl camera view area (lower left area) Vdm camera view area (lower center area) Vdr camera view area (Lower right area)

Claims (4)

[Claims] 1. A character recognition method for optically recognizing a character formed on a card and recognizing the character by image processing. In the character recognition method, a group of a series of cards to be recognized, a classification by the character formation method, and a character attribute. , Image data obtained by reading characters by a reading unit, a conveying unit that conveys the card to a predetermined reading position, an illuminating unit that illuminates the characters on the card, based on initial setting data including information regarding the arrangement of the characters. A character recognition method characterized in that character recognition is performed by controlling an image processing means for processing and a recognition method suitable for a character on a card to be read. 2. The character recognition method according to claim 1, wherein the information on the classification according to the character forming method includes at least information for distinguishing embossed characters, engraved characters, OCR characters, and print characters other than OCR. . 3. The character recognition method according to claim 1, wherein the information on the attribute of the character has at least one of a font, a character size, and a character color. 4. The information regarding the arrangement of characters includes at least information regarding a character start position on a card, a character pitch or a character interval, a total number of characters or a character end position. Character recognition method described.