JP2014071784A - Device and method for reading driver's license - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device and a method for reading a driver's license in which, even if a driver's license image is inputted in a state where the upper and lower, the right and the left, and the front and back surfaces are inverted, respectively, it is precisely determined in a short time whether or not the driver's license image is an erect image.SOLUTION: A device 1 for reading a driver's license includes: am image input part into which a driver's license image is inputted in multi-gradation; and an image direction determination part which determines whether or not the driver's license image inputted into the image input part is an erect image. When a ruled line L in the longitudinal direction is detected in a lower left side region P of the driver's license image, it is determined that the inputted driver's license image is the erect image.

Description

  The present invention relates to a driving license reading apparatus and a driving license reading method for optically reading a driving license image and outputting image information.

  Conventionally, a driver's license reading device that reads text, numbers, symbols, etc. printed on a driver's license has been printed on the driver's license after optically reading the driver's license with an image reader. Characters are recognized by reading description contents such as characters, and the information on the recognized characters is output and displayed on an output device.

  Here, in order to accurately read the description of the driver's license, the driver's license image information obtained by reading the driver's license with a scanner or the like must be correct, that is, an upright image. Inverted image (reversed image) information obtained upside down, left and right, and backside image information obtained upside down cannot accurately read the description of the driver's license.

  As described above, there is an apparatus disclosed in Patent Document 1 as an apparatus for determining whether the driver's license image information obtained by reading with a scanner or the like is an upright image, an inverted image (reverse image), or a back image. Patent Document 1 is a driver's license image that can be processed using the gradation value (256 gradations) of a pixel at a specific position of the driver's license image (expiration period area of the driver's license) (areas A and B). Whether or not it is a non-processable driver's license image, specifically, a specific region A and a region B located when it is flipped up and down, left and right or front and back, Based on the difference in pixel distribution between the two, it is determined whether it can be processed as a “driver's license image”, “upside down driver license image”, “inverted driver license image”, or “not a driver license image”. ing.

JP-A-7-65118 (Description [0021] to [0024])

  However, in the technique disclosed in Patent Document 1, the ratio Ka to the total number of pixels in the constant number of gradations range (value between N gradations and M gradations) in the area A and the area B The determination is made based on the magnitude relation between the ratio Kb of the number of pixels in a certain gradation number range to the total number of pixels.

  For this reason, N and M are constants set in advance assuming the blue of the expiration date area of the driver's license. For example, depending on the state in the areas A and B (for example, uneven lighting or dirt on the license) The calculated Ra and Rb may be close to each other. In such a situation, there is a possibility that the determination may be wrong, and the posture cannot be determined correctly.

  The present invention has been made in view of the above points, and its purpose is to provide an upright image with high accuracy in a short time even when a driver's license image is input in an upside down, left and right, upside down state. It is an object of the present invention to provide a driver's license reading device and a driver's license reading method that can determine whether or not the driver is licensed.

  In order to solve the above problems, the present invention provides the following.

  (1) An image input unit that inputs a driver's license image in multiple gradations, and an image direction determination unit that determines whether the driver's license image input by the image input unit is an erect image. A driver's license reading device that reads a driver's license image and outputs image information, wherein the image direction determining unit detects a ruled line in a predetermined area of the driver's license image, A driver's license reading device, wherein the driver's license image is determined to be an erect image.

  According to the present invention, when a ruled line is detected within a predetermined area of the driver's license image, it is determined that the image is an erect image, so that the determination can be made in a short time.

  (2) The driver's license reading apparatus, wherein the image direction determination unit calculates and determines a projection obtained by scanning the number of pixels in the predetermined region in the vertical direction.

  According to the present invention, since the number of pixels in a predetermined area is determined by calculating a projection scanned in the vertical direction, the determination can be made in a shorter time than the projection scanned in the vertical direction / horizontal direction (horizontal / vertical direction). It can be performed.

  (3) The driver's license reading apparatus, wherein the predetermined area is set at a predetermined position based on an outer frame line of the driver's license image.

  According to the present invention, since the predetermined region is set by capturing the geometric features of the driver's license, it is possible to make a highly accurate determination on the driver's license image that is input to the image input unit each time.

  (4) A driver's license reading method for reading a driver's license image and outputting image information, the step of inputting the driver's license image in multiple gradations, and scanning a predetermined area of the driver's license image And a step of detecting a ruled line and a step of determining that the input driver's license image is an upright image when a ruled line is detected within a predetermined area of the driver's license image. How to read a license.

  According to the present invention, when a ruled line is detected within a predetermined area of the driver's license image, it is determined that the image is an erect image, so that the determination can be made in a short time.

  The driver's license reading apparatus and the driver's license reading method according to the present invention detect the presence or absence of ruled lines in a predetermined area of the driver's license image based on geometric features such as the ruled line structure of the driver's license. It can be accurately determined in a short time whether the driver's license image is an upright image.

The block diagram which shows the electric constitution of the driver's license reading apparatus which concerns on embodiment of this invention. The figure which shows a driver's license image (upright image). The figure which shows a driver's license image (inverted image). The figure which shows a driver's license image (back image). The figure which shows the projection which scanned the pixel number in an area | region to the vertical direction. The figure which shows the driver's license image for demonstrating the driver's license reading method which concerns on embodiment of this invention. The flowchart which shows the driver's license reading method which concerns on embodiment of this invention. The flowchart which shows the driver's license reading method which concerns on embodiment of this invention. The flowchart which shows the driver's license reading method which concerns on embodiment of this invention. The figure which shows the driver's license image for demonstrating the other driver's license reading method which concerns on embodiment of this invention.

  The best mode for carrying out the present invention will be described below with reference to the drawings.

[Driver license reader]
FIG. 1 is a block diagram showing an electrical configuration of a driver's license reading apparatus 1 according to an embodiment of the present invention.

  The driver's license reading device 1 is an optical character reading device that optically reads the description content such as characters printed on the driver's license 2 with an image scanner, recognizes the characters, and outputs the recognized characters. (Optical Character Reader) is used.

  The driver's license 2 is 54 mm in length, 85.6 mm in width, 0.76 mm in thickness and has an IC chip built-in, and is the same size as a general card conforming to JIS.

  A driver's license reading device 1 shown in FIG. 1 is a device that can optically read a character string displayed on the driver's license 2, and is a close contact type (one-dimensional) in order to perform preprocessing of character string reading. ) The image sensor 11 and the data processor 13 are included.

  As shown in FIG. 1, the contact-type (one-dimensional) image sensor 11 transports the driver's license 2 in a predetermined transport direction (in the direction of the arrow from left to right in FIG. 1) by a transport unit (not shown). Is provided. The image sensor 11 is a line sensor, and is arranged in a line in a direction substantially perpendicular to the conveyance direction (vertical direction of the driver's license 2), and is longer than the length of the upper and lower ends of the driver's license 2. Yes. In addition, the conveyance mechanism 10 is formed with a conveyance guide 10A, and one end of the driver's license 2 is conveyed in contact with the conveyance guide 10A. Note that the imaging element 11 is not limited to one dimension, and may be a multidimensional imaging element. Furthermore, a CCD, a CMOS sensor, or the like may be used as the imaging element. Further, the present invention is not limited to the transport mechanism 10 that transports the driver's license 2, and may be a subject support mechanism that supports and captures the driver's license 2.

  The data processing unit 13 includes an image memory 13a, an image direction determining unit 13b, an image recognizing unit 13c, and an output unit 13d, each of which is an electric circuit such as a CPU, MPU, RAM, ROM, or the like. Depending on the element, it can be realized in hardware.

  The image memory 13a stores image data and image data obtained by processing such image data with respect to a driver's license image obtained by imaging. Specifically, each pixel is represented by multi-valued image data of 256 gradations having an 8-bit information amount, and a luminance value as a pixel value is represented as an integer value from 0 to 255. The image memory 13a is configured to store image data such as RAM, SDRAM, DDR, SDRAM, RDRAM.

  The image direction determination unit 13b determines whether the driver's license image input to the image memory 13a is an upright image. If the image is not an upright image, the image direction determination unit 13b further determines whether the image is an inverted image or a back image. Specifically, a projection obtained by scanning the number of pixels in a predetermined area P of the driver's license image in the vertical direction is calculated, and the determination is made based on whether or not the vertical ruled line L in the area P is detected. 2 detects a vertical ruled line L in the region P by the image direction determination unit 13b, and FIGS. 3 and 4 are not upright images because the vertical ruled line in the region P is not detected. Or back side image).

  Here, the predetermined area P may be determined in advance in the driver's license image, or may be set every time the driver's license image is input. For example, in a driver's license reader having a subject support mechanism that supports the driver's license 2 and picks up an image with an area sensor, the driver's license 2 is set at a predetermined position and picked up. Can be fixedly determined in advance without floating.

The detection of the vertical ruled line L in the area P in the image direction determination unit 13b is performed by calculating a projection obtained by scanning the number of pixels in the area P in the vertical direction, as shown in FIG. That is, if there is a section X _a -X _b where the projection gradually decreases, a substantially constant section X _b -X _c , and a section X _c -X _d where the projection gradually increases, the section X _a -X _d (or X _b- A vertical ruled line L exists in the section ( _Xc ). In FIG. 5, the vertical axis indicates the gradation value, 0 is black, and 255 is white. The horizontal axis indicates pixels.

  In the present embodiment, the vertical ruled lines L in the predetermined region P are used as geometric features such as the ruled line structure of the driver's license, but the present invention is not limited to this. The geometric features are those that protrude in one direction. That is, it is only necessary that when the number of pixels in the region P is projected in a predetermined direction such as a vertical direction or a horizontal direction, the number of pixels protrudes in the scanned direction. Specifically, in addition to the vertical ruled lines L extending in the vertical direction shown in FIG. 2 and the like, horizontal ruled lines extending only in the horizontal direction (direction perpendicular to the vertical direction) may be used. Further, a ruled line may be inclined. Any shape that protrudes in one direction other than the ruled line may be used.

  The image recognition unit 13c reads the driver's license image stored in the image input unit 13a by performing OCR processing. The output unit 13d outputs image information obtained by being read by the image recognition unit 13c. As the reading target, the description contents of the driver's license 2 such as name, date of birth, address, and license number are set as desired.

[How to read a driver's license]
A driving license reading method according to an embodiment of the present invention will be described using the driving license image shown in FIG. 7, 8 and 9 are flowcharts showing the driver's license reading method according to the embodiment of the present invention. FIG. 8 is a subroutine for image direction determination processing, and FIG. 9 is a subroutine for vertical ruled line presence / absence determination processing. is there.

  In the driving license reading method according to the present embodiment, as shown in FIG. 7, first, an image reading process for reading an image of the driving license is executed (step S101). One end of the driver's license 2 is conveyed in contact with the conveyance guide 10A. An image of the conveyed driver's license 2 is taken by the line sensor 11. The driver's license image obtained by imaging is stored in the image memory 13a as 256-level multi-value image data. Furthermore, image data obtained by rotating the captured multi-value image data or the like is also stored in the image memory 13a.

  Next, an image direction determination process is performed (step S102). In the image direction determination process, it is determined whether or not the driver's license image input to the image memory 13a is an erect image. If it is not an erect image, it is further determined whether it is an inverted image or a back image. Details of this processing will be described later with reference to FIG.

  Next, image recognition processing is performed by the image recognition unit 13c (step S103), and characters such as a driver's license number, an address, and a name are recognized using the multivalued image data of the driver's license 2, for example. Display processing for displaying the image information recognized in step S104 on the output unit 13d is performed (step S104). Note that the image recognition process in step S103 and the display process in step S104 are commonly performed processes, and thus detailed description thereof is omitted.

[Image orientation determination processing]
The contents of the image direction determination process will be described with reference to FIG.

  First, vertical ruled line presence / absence determination processing is performed (step S201). The contents of this process will be described later with reference to FIG. 9. To briefly describe, the area P is set in order to detect the vertical ruled line L in the area P. After setting the region P, the presence or absence of the vertical ruled line L in the region P is determined.

  As described above, it is determined whether or not the region setting has been completed, and if the region setting has been completed, the process proceeds to step S202. If the region setting has not been completed, for example, the input is operated. If it is not a license image, the process is terminated.

  Next, it is determined whether or not there is a vertical ruled line L in the region P (step S202). If it is determined that there is a vertical ruled line L, it is determined that the image is a processable image (step S206), and this subroutine is terminated. The upright image shown in FIG. 6A corresponds to this example.

  On the other hand, if it is determined in step S202 that there is no vertical ruled line L, the input driving license image is rotated 180 degrees (step S203). The driver's license image after the image is changed by rotating 180 degrees is stored in the image memory 12. In this example, the inverted image shown in FIG. 6B and the back image shown in FIG. 6C in which the vertical ruled line L does not exist in the region P correspond to this example, and when the inverted image is rotated 180 degrees, it becomes an upright image, and the image memory 13a The image data inside is rewritten from the driver's license image of FIG. 6 (b) to the driver's license image of FIG. 6 (a). Similarly, in the case of the back image shown in FIG. 6C, an image (not shown) obtained by rotating the back image by 180 degrees is rewritten in the image memory 13a.

  Next, a vertical ruled line presence / absence determination process similar to step S201 is performed on the rewritten image (step S204), and the presence or absence of the vertical ruled line L in the region P is determined (step S205). If it is determined that there is a vertical ruled line L, it is determined that the image is a processable image (step S206), and this subroutine is terminated. The inverted image shown in FIG. 6B corresponds to this example.

  On the other hand, if it is determined in step S205 that there is no vertical ruled line L, it is determined that the image cannot be processed (step S207), and this subroutine is terminated. A back side image (an image obtained by rotating the back side image by 180 degrees) shown in FIG. 6C corresponds to this example.

  In step S203, the input driver's license image is rotated 180 degrees. Instead of rotating the image 180 degrees, the area P is moved to a point-symmetrical position, and the area moved in step S204 is moved. It may be determined whether or not there is a vertical ruled line L in P.

[Vertical ruled line presence / absence judgment processing]
The content of the vertical ruled line presence / absence determination process will be described with reference to FIG.

  First, in the vertical ruled line presence / absence determination process, the area P is set to detect the vertical ruled line L in the area P before the process of step S301 shown in FIG. This region setting is the same for the inverted image of the driver's license shown in FIG. 6B and the back image of the driver's license shown in FIG. 6C, and the reading direction with respect to the inputted driver's license image. The left end W of the reference is used as a reference. Further, in the present embodiment, the region O is set as a previous stage for setting the region P. Specifically, according to the erect image of the driver's license 2 shown in FIG. 6A, after detecting the left end W of the driver's license image, a predetermined region O set in advance with the left end W as a reference. A projection obtained by scanning the number of pixels in the vertical direction is calculated, and the left outer frame line K is detected.

  A predetermined region P is set with the left outer frame line K as a reference. The detection of the left outer frame line K is the same as the detection method of the vertical ruled line L described with reference to FIG. 5, and thus detailed description thereof is omitted here.

  Since the left outer frame line K is longer than the vertical ruled line L, the vertical length of the region O is set longer than that of the region P, and the line segment can be accurately detected from the projection shown in FIG. It is preferable that the characters printed on the driver's license 2 do not enter the area O or the area P. For example, various settings can be made such that the width of the left outer frame line K or the vertical ruled line L is used as a reference and n times the width is used as the width of the region. As described above, the region P is set so that a part of the vertical ruled line exists and no vertical ruled line exists at a point-symmetrical position.

  As described above, it is determined whether or not the region setting has been completed, and if the region setting has been completed, the process proceeds to step S202. If the region setting has not been completed, for example, the input is operated. If it is not a license image, the process is terminated. If it is determined in step S201 that the region setting has been completed, an image direction determination process is performed (step S102). Details of this processing will be described later.

  Returning to the flowchart shown in FIG. 9, taking as an example a case where scanning is performed by dividing 10 pixels in the vertical direction with 10 times the width of the vertical ruled line L as the width of the region P, first, the left end side (Xs = X position of vertical ruled line-5) The horizontal position X is set as X = Xs (step S301).

  Next, the vertical position of the region P is fixed as y = Ys to Ye, and the pixel P on X is added to calculate the projection P (x) = Σp (x, y) (step S302).

  Next, it is determined whether or not scanning has been performed up to the right end side of the region P (Xe = vertical ruled line X coordinate + 5) (step S303), and if it is determined that Xe is not satisfied, the x coordinate of one pixel is shifted to the right. Then (step S304), the processing of step S302 and step S303 is repeated.

  If it is determined in step S303 that scanning has been performed up to the right end side of the region P (Xe = X coordinate of the vertical ruled line + 5), the projection in which the number of pixels in the region P is fixed as y = Ys to Ye (see FIG. 5) is added. Then, it is determined whether or not the minimum value Min (P (x)) is equal to or less than a predetermined threshold (step S305).

Specifically, the detection of the vertical ruled line L in the region P in the image direction determination unit 13b is performed by calculating a projection obtained by scanning the number of pixels in the region P in the vertical direction, as shown in FIG. That is, if there is a section X _a -X _b where the projection gradually decreases, a substantially constant section X _b -X _c , and a section X _c -X _d where the projection gradually increases, the section X _a -X _d (or X _b- A vertical ruled line L exists in the section ( _Xc ).

  If it is determined in step S305 that the minimum value is equal to or less than the threshold value, the vertical ruled line L is present (step S306), and if the minimum value is determined to be equal to or greater than the threshold value, the vertical ruled line L is not present (step S307).

  As described above, the area setting process for the area P follows the processing contents of the detection of the left end W → the setting of the area O → the detection of the left outer frame line K → the setting of the area P. In this case, for example, the processing content of detection of the left end W → setting of the region P may be used.

[Other driver's license reading methods]
Another driving license reading method according to the embodiment of the present invention will be described using the driving license image shown in FIG. The driver's license image shown in FIG. 10 is different from the one shown in FIG. 6, the front side has no “Personal Registration” entry column and has two outer frames (FIGS. 10A and 10B), and the back side has an outer frame. (FIG. 10 (c)).

  In another method of reading a driver's license according to the embodiment of the present invention, the driver's license can be read by executing the processing of each step of FIG. Here, in step S102, the regions O and P can be set for an upright image and an inverted image (FIGS. 10A and 10B), but there is no left outer frame line for a back image. Therefore, the area P cannot be set (FIG. 10C). For this reason, it is determined that the region setting has not been completed, for example, because the back side image cannot be processed or the input image is not a driver's license image, and the processing is terminated.

  Also, in another method of reading a driver's license according to the embodiment of the present invention, the image direction can be determined by executing the processing of each step of FIG. Since the image orientation determination process shown in FIG. 8 is an upright image and an inverted image for which region setting has been completed, it is determined NO in step S205 and is an unprocessable image in step S207. There is nothing. In addition, in the other driver license reading method according to the embodiment of the present invention, the presence or absence of the vertical ruled line L can be determined by executing the processing of each step of FIG.

[Effect of this embodiment]
The driver's license reading device and the driver's license reading method according to the embodiment of the present invention detect whether the driver's license image is correct by detecting the presence or absence of a vertical ruled line in the region P on the lower left side of the driver's license image. Whether the image is a standing image can be accurately determined in a short time. Further, since the setting of the region P of the driver's license image is performed from the pixel values in the partial region on the left side of the input driver's license image, the image processing time is shortened compared to processing the entire driver's license image. It is possible to determine whether the image is an erect image with high accuracy. If the image is not an erect image, the inverted image can be made an erect image by rotating the image 180 degrees, and the pixel value is calculated again in the region P that has already been set.

  The present invention detects the presence or absence of a vertical ruled line in the lower left region of the driver's license image based on geometric features such as the ruled line structure of the driver's license, so that the driver's license image is an upright image. It is useful as a driver's license reading apparatus and a driver's license reading method that can accurately determine whether or not

DESCRIPTION OF SYMBOLS 1 Driver's license reader 2 Driver's license 11 Image sensor 13 Data processing part

Claims (4)

An image input unit for inputting a driver's license image in multiple gradations;
An image direction determination unit that determines whether the driver's license image input by the image input unit is an erect image;
A driver's license reading device that reads a driver's license image and outputs image information,
The driver's license reading apparatus, wherein the image direction determining unit determines that the input driver's license image is an erect image when a ruled line is detected in a predetermined area of the driver's license image.   2. The driving license reading apparatus according to claim 1, wherein the image direction determination unit calculates and determines a projection obtained by scanning the number of pixels in the predetermined area in the vertical direction.   3. The driving license reading apparatus according to claim 1, wherein the predetermined area is set at a predetermined position based on an outer frame line of the driving license image. A driver's license reading method for reading a driver's license image and outputting image information,
Inputting the driver's license image in multiple gradations;
Scanning a predetermined area of the driver's license image to detect ruled lines;
A step of determining that the input driver's license image is an erect image when a ruled line is detected within a predetermined area of the driver's license image;
A method for reading a driver's license, comprising:

Images