JP4900701B2 - Authentication system - Google Patents

Description

  The present invention relates to an authentication system for performing identity verification using finger vein biometric information, and more particularly to an authentication algorithm for performing high-speed authentication.

  Recently, in biometric identification devices, not only biometric authentication using human fingerprints and retinas, but also vein authentication devices that perform identity verification using finger vein pattern images have become widespread. As a vein pattern imaging device of a vein authentication device, there is Patent Document 1 relating to a biometric identification device that acquires a finger vein pattern in a non-contact manner as follows.

FIG. 2 is a mechanical configuration diagram of Patent Document 1 relating to a personal authentication device that acquires a finger blood vessel pattern in a non-contact manner.
A finger rest on which the finger 201 to be identified is placed; an infrared light source 202, 203 that makes light incident on the finger that captures an image of the finger 201 placed on the finger rest; and the finger rest placed on the finger rest. Provided with a camera 204 that captures an image of the finger 201, infrared incident light is irradiated from both sides of the finger, the infrared light transmitted through the finger 201 is saturated in the finger, and the saturated light (image) is transmitted to the camera 204. Take a picture. The imaging region of the camera 204 covers the first joint 302 to the second joint 303 of the finger 201 as shown in FIG.

Japanese Patent Laid-Open No. 2005-253989

By the way, the finger authentication algorithm used in Patent Document 1 is an algorithm that uses a pattern matching process to verify whether an image that has been captured and registered in advance and an image of a finger to be authenticated match each other. ing.
In the case of finger veins, the image to be authenticated itself does not have a large image area that requires authentication, so even if processing such as pattern matching is performed alternately left and right and up and down, the images match and do not match at a relatively high speed. It is possible to confirm whether or not authentication is performed.
However, when the finger vein image becomes large or the image distortion is large, there is a problem that the range of movement to the left, right, up and down becomes wide and takes more time.
Also, if the transmitted infrared image becomes clearer, or if a large number of finger authentication processes are attempted at the same time, the pattern matching process takes time, and the processing speed decreases or the load is reduced. This may cause a problem that it takes time until authentication.

  The present invention has been made in view of the above problems, and provides an authentication system that shortens the authentication time, improves the authentication accuracy rate, increases the identity authentication rate, and reduces the acceptance rate of others. There is.

In order to achieve the above object, an authentication system according to the present invention is an authentication system for performing identity verification using biological information related to a vein of a human finger,
An imaging unit that acquires image information including the biological information, a biological image information generation unit that generates a polygonal line vectorized image and a ternary image by generating a polygonal line vector and ternarizing the acquired image information, and Authentication processing means for performing identity verification based on the broken line vectorized image and the ternary image ,
The authentication processing means captures the number of fold lines and the number of branch points in the polygonal line vectorized image newly captured by the imaging means and newly generated by the biological image generation means, and the individual folding line of the confirmation target stored in the storage means in advance. It is determined whether or not the number of broken lines and the number of branch points in the vectorized image match. If they match, the first is obtained by removing the gray portion from the ternary image newly generated by the biological image generating means. Is generated, and it is determined whether or not the positional relationship between black and white in the second binarized image of the individual to be confirmed stored in advance in the storage means matches. Generates a first synthesized image obtained by synthesizing the first binarized image with a polygonal line vector image newly generated by the biological image generating means, and also generates the first binarized image with respect to the second binarized image. Memory A second combined image is generated by combining the broken line vectorized image read from the stage, and pattern matching processing is performed on the generated first and second combined images so that the number of broken lines and the number of branch points in both images match. And a means for making the identity verification successful when the positional relationship between black and white matches .

  According to the authentication system according to the present invention, a polygonal line vectorized image and a ternary image are generated from the captured biological image information, authentication determination is performed for each image unit, and the authentication target image is recombined. By performing the determination process based on the synthesis result, it is possible to speed up the authentication, and it is possible to improve the authentication accuracy by performing a plurality of authentication determinations.

Hereinafter, an embodiment in which an authentication system according to the present invention is applied to login authentication to an OS when a notebook PC or the like is activated will be described.
FIG. 1 is an explanatory diagram showing the overall configuration of the authentication system according to the first embodiment of the present invention.
The authentication system in this embodiment is a vein authentication device that performs identity verification based on finger information of a registrant who operates the notebook PC 120 when the notebook PC 120 is activated, for example, inside a notebook PC 120 that is a general information processing apparatus or the like. 100.
Here, the finger information is information including image information showing a blood vessel pattern of a finger vein acquired by imaging from a surface such as a ventral side of a finger of a registrant's body.
The notebook PC 120 includes an I / F 121 that transmits and receives signals and information to and from the vein authentication device 100, a CPU 122 that controls operations of each component and performs various operations, and the CPU 122 performs operations such as operations. A memory 123 that temporarily stores information such as results of the operation, an HDD 124 that registers and stores information such as finger information based on the registrant's body, and a registrant who operates the notebook PC 120 performs an input operation such as character information. And a keyboard 125 for performing a selection operation, a liquid crystal display 126 which is a display means for displaying a processing result, information and data of a program executed by the CPU 122, and a speaker 127 which outputs an error sound during operation of the notebook PC 120. And.

The CPU 122 has a function of temporarily storing the image information transmitted from the vein authentication device 100 in the memory 123, extracting, acquiring, and encrypting finger information from the image information, and storing it in the HDD 124.
Further, it has a function of calculating luminance based on the image information and performing processing for adjusting the amount of infrared light emitted from an infrared light source 101 (to be described later) in the vein authentication apparatus 100 according to the luminance.

  The vein authentication apparatus 100 includes a finger rest 103 for placing the finger 105 at a predetermined position, an infrared light source 101 for irradiating the finger 105 with infrared light, and an infrared light source for authenticating the registrant's finger 105. A camera 102 that captures the veins of the finger 105 projected by emitting and irradiating infrared light 101, a photosensor 104 for detecting that the finger 105 is placed on the finger rest 103, and the camera 102 captures an image A / D converter 106 that converts analog data of the processed image into digital data, an image input unit 107 that generates image information of digital information using the digital data converted by A / D converter 106, and CPU 122 And a light emission control unit 108 that controls the operation of the infrared light source 101 in accordance with the light emission control signal transmitted from the I / F 121 via the I / F 121. .

FIG. 6 is an explanatory diagram showing finger information 601 stored in the HDD 124.
The finger information 601 includes finger information obtained from the first finger of the registrant and finger information obtained from a second finger different from the first finger as information such as finger information obtained from the registrant's body. It is registered and memorized.

Hereinafter, the operation of the authentication system in the present embodiment will be described in detail with reference to the drawings.
First, a process of registering finger information including a blood vessel pattern of a registrant's finger vein will be described with reference to a flowchart shown in FIG.
When a registrant who operates the notebook PC 120 operates the keyboard 125 to start the vein pattern registration program stored in the HDD 124, the CPU 122 is necessary for initialization of the entire hardware of the notebook PC 120 and execution of the vein pattern registration program. A process of substituting an initial value for a temporary variable is performed (step 401).
Next, when the transition to the initialization state is completed, the CPU 122 performs a process of determining how many fingers are registered (step 402).
When first registering the first finger, a process of displaying the message “first finger registration operation” on the liquid crystal display 126 is performed, and the registrant's index finger or the like is displayed on the finger rest 103 of the vein authentication device 100. A screen is displayed to prompt the user to place the first finger 105 (step 403). In addition, when the first registration is completed and the second finger is registered, a process of displaying a message “second finger registration operation” on the liquid crystal display 126 is performed, and the finger placement table of the vein authentication device is displayed. A screen for prompting the user to place the second finger of the registrant, such as the middle finger, is displayed at 103 (step 404).

  Next, the CPU 122 sets the vein authentication device 100 in an idling state, performs a light emission amount adjustment process in response to the finger placed on the finger rest 103, appropriately adjusts the light emission amount of the infrared light source 101, and An image showing a clear vein blood vessel pattern is captured (step 405). Here, when an abnormality is detected in the light emission amount adjusting process and a process of displaying a light emission amount abnormality message is performed, the process ends.

  Next, the CPU 122 performs processing for extracting a vein blood vessel pattern (step 406). The image information showing the clear vein blood vessel pattern obtained by the light emission amount adjustment processing is read from the memory 123, and the portion of the image information image where the blood vessel pattern is not shown is deleted to obtain the vein blood vessel pattern, etc. The process which extracts the part which imaged as finger information is performed.

Next, the CPU 122 performs a process for encrypting the finger information (step 407). For example, the extracted finger information is encrypted using a darkening pattern based on a predetermined encryption program.
Next, the CPU 122 performs processing for determining what finger information of the finger is registered (step 408). For example, by determining how many times the process of registering finger information is performed, it is determined how many fingers the extracted finger information is obtained. The finger information is registered and stored in the finger information 601 as a blood vessel pattern of the first finger vein (step 409).

Next, the CPU 122 performs processing for displaying a message for instructing the registrant to remove the finger 105 from the finger rest 103 on the liquid crystal display 126 (step 411). For example, a process of displaying a message “Please remove your finger” on the liquid crystal display 126 is performed.
Next, the CPU 122 performs a process of determining whether or not a switch-off signal is transmitted from the photosensor 104 (step 413). When a switch-off signal is transmitted (YES in step 413), the process after step 402 is performed. Repeat the process. If the switch-off signal has not been transmitted (NO in step 413), the process of step 413 is repeatedly executed until it is transmitted.
Here, in step 408, for example, by determining how many times the process of registering the finger information has been performed, it is determined how many fingers the extracted finger information is acquired, and the second one. In this case, the extracted finger information is registered and stored in the finger information 601 as a blood vessel pattern of the second finger vein (step 410).

  Next, the CPU 122 performs a process of displaying a message notifying the registrant that the registration process is completed on the liquid crystal display 126 (step 412). For example, a process of displaying a “registration complete” message on the liquid crystal display 126 is performed.

Next, the light emission amount adjustment process executed in step 405 described above will be described with reference to the flowchart shown in FIG.
First, the CPU 122 sets the vein authentication device 100 in an idling state and performs a process of transmitting a light emission control signal for emitting light with a predetermined light emission amount as an initial value to the light emission control unit 108 via the I / F 121 (step 501). .
When receiving the light emission control signal, the light emission control unit 108 performs control to emit infrared light from the infrared light source 101 with light emission with a predetermined light emission amount.

Next, the CPU 122 performs a process of determining whether or not a switch-on signal is transmitted from the photosensor 104 (step 502). In accordance with a message such as “first finger registration operation” displayed on the liquid crystal display 126, the photo sensor 104 detects this in response to the registrant placing his / her first finger on the finger rest 103. Then, a process for determining whether or not a switch-on signal is transmitted to the CPU 122 via the I / F 121 is performed.
If the switch-on signal has not been transmitted (NO in step 502), the process of step 502 is repeatedly executed until it is transmitted.

Next, when a switch-on signal is transmitted (YES in step 502), the CPU 122 performs a process of displaying a message notifying that a vein image is being captured (step 503). For example, a process of displaying a message “in vein imaging” on the liquid crystal display 126 is performed.
Next, the CPU 122 performs a process of capturing an image showing a blood vessel pattern of a finger vein (step 504). Processing for imaging the image of a vein displayed by operating the camera 102 via the light emission control unit 108 by irradiating the infrared light emitted from the infrared light source 101 from the finger 105 placed on the finger placing base 103. I do. The analog data of the vein image captured by the camera 102 is converted into digital data by the A / D converter 106, and the image input unit 107 generates digital information image information using the digital data. The data is transmitted to the memory 123 via the I / F 121 and stored.

Next, the CPU 122 performs a process of calculating the luminance of the light emission amount of the image information (step 505). Image information is read from the memory 123, and processing for calculating the luminance of the central portion 304 of the imaging region 302 is performed from the image information as shown in FIG.
Next, the CPU 122 performs a process of determining whether or not the luminance of the light emission amount of the image information is within a preset target range (step 506). The upper limit value and lower limit value of the target range of luminance of image information set and stored in advance in the HDD 124 are read, and it is determined whether or not the luminance of the calculated light emission amount is included in the range of these upper limit value and lower limit value. Perform the process.
Here, the upper limit value and the lower limit value of the target range of the luminance of the image information are the upper limit of the luminance when the image is clearly projected so that the vein blood vessel pattern displayed in the central portion 304 of the imaging region 303 can be grasped. Value and lower limit are set. That is, at the luminance within the target range of the upper limit value and the lower limit value, the shape and pattern of the vein blood vessel pattern displayed in the central portion 304 is clearly displayed, and the image information of the vein blood vessel pattern is used. Verification processing and the like can be executed.

Next, when the luminance of the light emission amount of the image information is within a preset target range (YES in Step 506), the CPU 122 sends a light emission control signal for stopping the light emission of the infrared light source 101 to the I / F 121. The process of transmitting to the light emission control unit 108 is performed (step 512). Also, the light emission control unit 108 performs control to stop the light emission of the infrared light source 101 when receiving this light emission control signal. Then, the process ends.
Here, when the luminance of the light emission amount of the image information is not within the preset target range (NO in step 506), the CPU 122 performs a process of determining whether or not the target range is exceeded (step 507). ). Processing is performed to determine whether the calculated luminance of the emitted light amount exceeds the upper limit value of the target range of the luminance of the image information or is lower than the lower limit value of the target range.

  Next, when the luminance of the calculated light emission amount exceeds the upper limit value of the target range (YES in step 507), the CPU 122 performs a process of reducing the light emission amount of the infrared light source 101 (step 508). Processing for transmitting a light emission control signal for light emission with a small light emission amount reduced by one step from the initial light emission amount to the light emission control unit 108 via the I / F 121 as the infrared light emission amount emitted from the infrared light source 101. Do.

In addition, when the luminance of the calculated light emission amount is lower than the lower limit value of the target range (NO in step 507), the CPU 122 performs a process of increasing the light emission amount of the infrared light source 101 (step 509). Processing for transmitting a light emission control signal for light emission with a large light emission amount increased by one step from the initial light emission amount to the light emission control unit 108 via the I / F 121 as an infrared light emission amount emitted from the infrared light source 101. Do.
When the light emission control unit 108 receives this light emission control signal, it determines whether or not the light emission amount of the infrared light source 101 specified by the light emission control signal is within the range of the infrared light emission amount that can be emitted by the infrared light source 101. If it is determined (step 510) and the light emission is within the range that can be emitted by the infrared light source 101 (YES in step 510), the infrared light source 101 is controlled to emit infrared light with the light emission amount specified by the light emission control signal. Then, the processing after step 504 is repeatedly executed.

Next, when it is not within the range where the infrared light source 101 can emit light (NO in step 510), the CPU 122 performs a process of displaying a message notifying that the light emission amount is abnormal (step 511).
The light emission control unit 108 determines that the light emission amount of the infrared light source 101 specified by the light emission control signal is not within the range of the infrared light emission amount that can be emitted by the infrared light source 101, and accordingly, the light emission amount is abnormal. In response to the notification information to be notified being transmitted to the CPU 122 via the I / F 121, the CPU 122 performs a process of displaying a message “abnormal emission amount” on the liquid crystal display 126, for example.

  Next, the CPU 122 performs a process of transmitting a light emission control signal for stopping the light emission of the infrared light source 101 to the light emission control unit 108 via the I / F 121 (step 513). When receiving the light emission control signal, the light emission control unit 108 performs control to stop the light emission of the infrared light source 101. Then, the process ends.

Next, a process of performing personal authentication using finger information when the notebook PC 120 is activated will be described with reference to a flowchart shown in FIG.
When a registrant who operates the notebook PC 120 is activated by turning on the notebook PC 120, the CPU 122 sets temporary variables necessary for initialization of the entire hardware of the notebook PC 120 and execution of the vein pattern matching program. A process for substituting an initial value is performed (step 701).

Next, when the transition to the initialization state is completed, the CPU 122 performs a process of determining how many fingers are to be compared (step 702). When collating the first finger for the first time, a process of displaying the message “first finger collating operation” on the liquid crystal display 126 is performed, and the registrant's index finger, for example, is displayed on the finger rest 103 of the vein authentication device 100. A screen is displayed to prompt the user to place the first finger 105 such as (step 703).
When the first verification is completed and the second finger is verified, a message of “second finger verification operation” is displayed on the liquid crystal display 126, and the finger placement of the vein authentication device is performed. A screen is displayed to urge the registrant to place a second finger such as the middle finger of the registrant (step 704).

  Next, the CPU 122 sets the vein authentication device 100 in an idling state, performs light emission amount adjustment processing similar to the above-described step 505 in response to the finger placed on the finger rest 103, and the light emission amount of the infrared light source 101 Is appropriately adjusted, and an image showing a blood vessel pattern of a clear vein of a finger is captured (step 705). Here, when abnormality is detected in the light emission amount adjustment processing and processing for displaying a light emission amount abnormality message is performed, the processing from step 702 onward is repeatedly executed.

Next, the CPU 122 performs processing for extracting a vein blood vessel pattern (step 706). The image information showing the clear vein blood vessel pattern obtained by the light emission amount adjustment processing is read from the memory 123, and the portion of the image information image where the blood vessel pattern is not shown is deleted to obtain the vein blood vessel pattern, etc. The process which extracts the part which imaged as finger information is performed.
Next, the CPU 122 performs processing for determining what finger information of the finger is to be collated (step 707). For example, by determining how many times the process of collating finger information is performed, it is determined how many fingers the extracted finger information is obtained.

Next, when the extracted finger information is the finger information acquired from the first finger, the CPU 122 performs processing for reading the finger information of the first finger registered in the finger information 601 (step 708). . The HDD 124 searches the finger information of the first finger registered in the finger information 601 by searching the HDD 124, and performs a process of decompressing (decrypting) the encryption of the read finger information of the first finger.
Further, here, when the extracted finger information is finger information acquired from the second finger, the CPU 122 performs a process of reading the finger information of the second finger registered in the finger information 601 (step 709). ). The HDD 124 searches the finger information of the second finger registered in the finger information 601 by searching the HDD 124, and performs a process of decompressing the read finger information of the second finger.

Next, the CPU 122 performs a process of comparing / collating whether or not the extracted finger information matches the registered finger information (step 710). For example, the extracted finger information and the read finger information are subjected to a process of comparing and collating whether or not the shape or pattern of the vein blood vessel pattern included in the finger information matches.
Next, when the extracted finger information matches the registered finger information (YES in step 710), the CPU 122 displays a message for instructing the registrant to remove the finger 105 on the liquid crystal display 126. (Step 711). For example, a process of displaying a message “Please remove your finger” on the liquid crystal display 126 is performed.
Next, the CPU 122 performs processing for determining whether or not a switch-off signal is transmitted from the photosensor 104 (step 713). If the switch-off signal is not transmitted (NO in step 713), the CPU 122 transmits the switch-off signal. Step 713 is repeatedly executed until

Next, when a switch-off signal is transmitted (YES in step 713), the CPU 122 performs a process of determining whether or not the second finger information collation is completed (step 715). ).
In step 710, a process is performed to determine whether or not it is determined that the extracted finger information matches the registered finger information for both the first and second fingers. If it is not determined that both the first and second fingers match (NO in step 715), the processes in and after step 702 are repeatedly executed.

Next, when it is determined that both the first finger and the second finger match (YES in step 715), the CPU 122 displays a message notifying the registrant that the authentication is completed on the liquid crystal display 126. A display process is performed (step 716). For example, a process of displaying a message “authenticated” on the liquid crystal display 126 is performed. Then, the process ends.
If the extracted finger information and the registered finger information do not match at step 709 (NO at step 710), the CPU 122 displays a message notifying the registrant of the mismatch on the liquid crystal display 126. (Step 712). For example, a process of displaying a “mismatch” message on the liquid crystal display 126 is performed.

  Next, the CPU 122 performs processing for determining whether or not a switch-off signal is transmitted from the photosensor 104 (step 714). If the switch-off signal is not transmitted (NO in step 714), the CPU 122 transmits the switch-off signal. Step 714 is repeatedly executed until In addition, when a switch-off signal is transmitted (YES in step 714), the CPU 122 repeatedly executes the processing after step 702.

Next, with reference to FIGS. 8 and 9, an authentication algorithm based on the polygonal line vectorization processing of the finger vein image in FIG. 10 will be described.
FIG. 8 is a diagram showing the overall flow of the polygonalization process.
A finger image 801 is photographed using a finger vein authentication device for a finger to be authenticated, and processing for extracting vein image data 802 is performed.
The image data 802 simply shows the finger vein imaging result.
FIG. 9 is a diagram showing an example in which a curve is bent.
The curved line 902 and the curved line 904 are subjected to broken line processing using line segment approximation to generate bent lines 903 and 904. Since line segment approximation is a well-known technique, description thereof is omitted here.
Reference numeral 803 in FIG. 8 indicates a result obtained by performing a polygonal process on the vein image data 802 that is a curve.
A tracking start point is designated for a blood vessel that has been subjected to a broken line process, an automatic line segment tracking process is performed, and the position and number of broken lines and branches are confirmed. When performing automatic tracking of line segments, it is only necessary to extract a branching point that is divided into multiple line segments, but when determining the position of a broken line, make sure that each line segment has a corner point. The number of broken line portions can be grasped by extracting the vertexes of the broken line.

FIG. 10 is a flowchart for explaining an authentication algorithm using a broken line vectorization process of a finger vein image.
The vein blood vessel portion is extracted from the imaging data obtained by infrared transmitted light (step 1001). The child extraction method is repeated until a tracking start point is established and the automatic tracking of the line segment is completed.
Next, the extracted venous blood vessel data is subjected to a polygonalization process to convert the curve into a polygonal line vector (step 1002).
Next, a broken line part and a branching point of the broken blood vessel image are extracted (step 1003). For the extraction of the broken line part and the branching point, a tracking start point is designated for the blood vessel subjected to the bending process, Minute tracking is performed, and the position and number of broken lines and branches are confirmed. When performing automatic tracking of line segments, it is only necessary to extract a branching point that is divided into multiple line segments, but when determining the position of a broken line, make sure that each line segment has a corner point. To extract the vertices of the broken line.

Separately, the processing from step 1001 to step 1003 is also performed on an image that is a registered basic pattern for authentication, and the number of broken line portions and branch points is extracted.
Next, the basic pattern of authentication is compared with the number of broken lines and branch points of the vein pattern to be authenticated (step 1004).

  When the number of broken line portions and the number of branch points coincide (step 1005), the process proceeds to the next process (step 1006). If the number of broken lines and the number of branch points do not match (step 1007), the authentication process is terminated as an authentication failure (step 1008).

Next, an authentication algorithm based on the ternary processing of the finger vein image in FIG. 12 will be described with reference to FIG.
FIG. 11 is a diagram in which ternary processing is performed on captured vein image data.
The captured finger vein image data is a grayscale image at first glance, and it is difficult to determine the position of the vein from this image. Therefore, a process for obtaining a ternary image from the grayscale image is performed.
The three-valued image here is an image photographed with infrared transmitted light, since the unevenness of the edge is difficult to understand, and it is difficult to clarify only with the shades of black and white, so three color values of black, white, and gray are used. Divide into areas that you have.
The ternary image data 1101 is a diagram that has been subjected to ternary processing. A figure obtained by dividing the color value areas into black, white, and gray by ternarization and then deleting the gray part is ternary image data 1102.
The reason for deleting the gray portion is that the region segmentation is not clear and the edge portion of the blood vessel is difficult to understand.

FIG. 12 is a flowchart for explaining an authentication algorithm using a finger vein image ternary process.
A part other than the blood vessel portion of the vein is extracted from the photographing data by the infrared transmitted light (step 1201).
As an extraction method here, when extracting a blood vessel image in FIG. 10, chroma key processing may be used in which only the extracted portion is inverted, or a color is added to the blood vessel portion and the edge portion is excluded and then excluded. It doesn't matter.
A ternary process is performed on the extracted image (step 1202).
There are already known technologies such as a method based on density histogram, a method that uses the local property of an image, a method that uses this property, and a method that divides an image into several parts and sets each region. Use.

In this case, a certain threshold

It is better to use the minimum error method that minimizes the error rate when the target is mistaken for the background, but any of the P-tile method, mode method, and discriminant analysis method in the density histogram method May be used.
The gray area is deleted from the image divided into three values of black, white, and gray by the ternary processing (step 1203). When this deletion is performed, the vein image data may be divided into a plurality of regions (matrix-like 8 * 8, 16 * 16).
Also, at the stage of dividing into three values of black, white, and gray, the pixels existing within a certain threshold are divided into area units around the pixel having the highest color value, and each of black, white, and gray It may be divided into areas.

  Next, the processing from Step 1701 to Step 1703 is also performed on the separately registered image that is the basic pattern of authentication, and image processing having an area divided into black and white is performed. .

Next, the basic pattern of authentication is compared with the number and position of the black and white areas of the vein pattern to be authenticated (step 1204).
As a result of the comparison, if the number and the positional relationship match (step 1205), the process proceeds to the next process (step 1206).
If even the number and the positional relationship do not match (step 1207), the authentication process is terminated as an authentication failure (step 1208).

Next, the authentication algorithms for performing authentication by receiving the results of successful authentication in FIGS. 10 and 16 and recombining the image data after the polygonalization processing and the image data after the ternary processing will be described with reference to FIGS. 15 will be used to explain the flowchart of FIG.
Image data 1301 in FIG. 13 is data obtained by extracting a broken line portion and a branch point using the broken line processing shown in FIG.
Image data 1302 is image data that has been subjected to the ternarization processing shown in FIG.

FIG. 14 is a diagram showing the result of recombining data 1301 and data 1302.
FIG. 15 is a flowchart for explaining an algorithm used for authentication by recombining image data converted into a polygonal line vector and a binarized (actually ternary) image.
First, the image data 1301 and the image data 1302 shown in FIG. 13 are recombined (step 1501). When re-synthesizing, the position that is the reference for each synthesis is set to the data of the polygonalization process and the ternarization process with the coordinates of the tracking start point specified when tracking the line segment of the blood vessel. A reference point for re-synthesis using the set coordinates.
Next, it is calculated whether the broken line portion and branch point of the broken line data calculated in FIG. 10 are in the black or white region of the binarized image calculated in FIG. 12 (step 1502).
Here, the calculation is made based on whether or not there is a broken line portion or a branch point in the binary region of black and white. Also, coordinates are set in the binary area of black and white, and coordinates are also set in advance at the positions of the broken line part and branch point, and the broken line part and branch point are set to the coordinates in the area (four coordinate points). It is also possible to use a method for confirming whether or not there is a coordinate (coordinate of one point).

Next, the processing from step 1501 to step 1503 is also performed for the image that is separately registered and used as the basic pattern for authentication, and the broken line portion and branch point of the broken line data in FIG. 10 are calculated in FIG. It is calculated which area is black or white of the digitized image.
Next, the basic pattern of authentication is compared with whether or not the broken line portion to be authenticated and the branch point are within the binarized black and white areas (step 1504).
If the broken line portion and the branch point coincide with each other in the region (step 1505), the authentication process is terminated as authentication is successful, and authentication is permitted (step 1506). If the broken line part and the branch point do not match in the area (step 1507), the authentication process is terminated as an authentication failure (step 1508).
However, the processing shown in FIG. 15 is not performed, and the authentication processing shown in FIG. 10 and FIG.

As is clear from the above description, in this embodiment, a polygonal line vectorized image and a ternary image of a biological image are generated, and when the identity verification is performed, the confirmation target stored in the storage unit in advance is stored. When the number of fold lines and the number of branch points in a personal fold line vectorized image are compared with the number of fold lines and the number of branch points in a newly photographed individual fold line vectorized image, if they do not match, authentication fails and they match. When the black and white positions in the ternary image coincide with each other, the comparison process is performed using an image obtained by combining the polygonal line vectorized image and the ternary image. Whether or not the confirmation target is another person can be determined only by the comparison result of the polygonal line vectorized image, and the authentication result can be output at high speed.
Further, when the subject of confirmation is the person himself / herself, the determination is made based on an image obtained by synthesizing the polygonal line vectorized image and the ternary image.

It is a system configuration figure showing an embodiment of an authentication system concerning the present invention. It is explanatory drawing which shows the structure of a vein authentication apparatus. It is a figure which shows the example of the image of the finger | toe image | photographed with the vein authentication apparatus. It is a flowchart which shows the process which registers finger information in the authentication system of embodiment. It is a flowchart which shows the light emission amount adjustment process in the authentication system of embodiment. It is explanatory drawing which shows the example of the finger information registered into the memory | storage means in embodiment. It is a flowchart which shows the process which performs identity verification by biometric information in embodiment. It is explanatory drawing of a broken line process in embodiment. It is explanatory drawing which shows the example of curve-folding of a curve. It is a flowchart which shows the comparison process of the polygonalization vector image of a finger vein blood vessel image. It is explanatory drawing which shows the ternarization process of a finger vein blood vessel image. It is a flowchart which shows the comparison process of the ternary image of a finger vein blood vessel image. It is a figure which shows the example of the polygonalization vector image and binarization image of a synthetic | combination object. It is a figure which shows the example of a synthesis | combination of a polygonalization vector image and a binarization image. It is a flowchart which shows the comparison process by the synthesized image of a polygonalization vector image and a binarized image.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Vein authentication apparatus 101 Infrared light source 102 Camera 103 Finger rest 104 Photo sensor 105 Finger 106 A / D converter 107 Image input part 108 Light emission control part 120 Notebook PC
121 I / F
122 CPU
123 Memory 124 HDD
125 Keyboard 126 Liquid crystal display 127 Speaker

Claims (1)

An authentication system that performs identity verification using biometric information related to a human finger vein,
Imaging means for acquiring image information including the biological information;
Biometric image information generating means for generating a polygonal line vectorized image and a ternary image by performing polygonal line vectorization and ternarization on the acquired image information;
An authentication processing means for performing identity verification based on the generated polygonal line vectorized image and the ternary image ;
The authentication processing means captures the number of fold lines and the number of branch points in the polygonal line vectorized image newly captured by the imaging means and newly generated by the biological image generation means, and the individual folding line of the confirmation target stored in the storage means in advance. It is determined whether or not the number of broken lines and the number of branch points in the vectorized image match. If they match, the first is obtained by removing the gray portion from the ternary image newly generated by the biological image generating means. Is generated, and it is determined whether or not the positional relationship between black and white in the second binarized image of the individual to be confirmed stored in advance in the storage means matches. Generates a first synthesized image obtained by synthesizing the first binarized image with a polygonal line vector image newly generated by the biological image generating means, and also generates the first binarized image with respect to the second binarized image. Memory A second combined image is generated by combining the broken line vectorized image read from the stage, and pattern matching processing is performed on the generated first and second combined images so that the number of broken lines and the number of branch points in both images match. And an authentication system , further comprising means for successfully verifying the identity when the positional relationship between black and white matches .