JP5951817B1 - Finger vein authentication system - Google Patents

Abstract

A finger vein authentication system for recognizing, collating and discriminating a finger vein pattern of a user at a low capacity / high speed by a method of analyzing a level set curvature. The system captures an image including a light source unit that emits near-infrared light, an optical filter unit that transmits light of a specific wavelength from the light source unit, and a user's finger. It includes a CCD camera unit for converting to an electrical signal, a memory unit for storing a level set curvature program for extracting a finger vein pattern from an image using the level set curvature, and a control unit for driving the program. The level set curvature program removes noise from the image including the finger acquired by the CCD camera unit with respect to the control unit, separates the image of only the finger part by removing the peripheral part from the image, The finger vein pattern is extracted from the image of only the finger portion using the level set curvature, and the finger vein pattern is binarized to instruct the finger vein pattern to be thinned. [Selection] Figure 2

Description

  The present invention relates to a finger vein authentication system, and more specifically, a finger vein for recognizing, collating and discriminating a finger vein pattern of a user at a low capacity / high speed by a method of analyzing a level set curvature (level set curvature). It relates to the authentication system.

  In general, biometrics is a method of recognizing an individual using human physiological or behavioral characteristics. Typical bio-recognition technologies include fingerprints, faces, irises, finger veins, and signature recognition. Among such bio-recognition methods, fingerprint recognition and finger vein recognition have the advantages of high recognition performance, reduced resistance and inconvenience for the user, and fast recognition time.

  The concept that human identity can be determined using human fingerprints has been widely used for a long time to search for criminals, and it is already widely known that the shape of fingerprints varies from person to person. It is a fact. However, when the finger of the fingerprint is injured, a dry finger, or a wet finger, there is a drawback that it is difficult to obtain an epidermal fingerprint pattern with a general fingerprint scanner.

  In order to solve this problem, “Lumidigm” has proposed a method in which visible light is transmitted through a finger to obtain a dermal fingerprint of the finger and used for recognition. A dermal fingerprint has the advantage that it can be obtained under any conditions as long as there are no severe scratches on the finger.

  Further, the research of “Yanagawa” has proved the concept that even a human finger vein can determine whether or not the person is a person. According to the research of “Yanagawa”, if the finger vein patterns of ten fingers are all different for each individual, one finger vein pattern can be used for personal authentication with a degree of freedom similar to an iris pattern with very high reliability. It is said that it is suitable for.

  As described above, research for eliminating the defect that the pattern of the epidermal fingerprint is not successfully acquired by acquiring the dermis fingerprint of the finger and acquiring the finger vein for personal authentication has been underway.

  The present invention has been devised in view of the above-described points, and recognizes and collates a finger vein pattern of a user at a low capacity / high speed by a method of analyzing a level set curvature (level set curvature). It is another object of the present invention to provide a finger vein authentication system that can be discriminated.

  In order to achieve the above-described object, a preferred embodiment of the present invention includes a light source unit configured to emit near-infrared light, and light having a specific wavelength among light incident from the light source unit. An optical filter unit configured to transmit only light, a CCD camera unit configured to capture an image including a user's finger and then convert the captured image into an electrical signal; A memory unit storing a level set curvature program for performing an operation of extracting a finger vein pattern from an image including the finger using a level set curvature, and the level set curvature program. A finger vein authentication system including a driving control unit, wherein the level set curvature program is acquired from the control unit by (a) the CCD camera unit. Removing noise from the image including the finger; and (b) separating an image of only the finger portion by removing a peripheral portion from the image including the finger from which noise has been removed; c) extracting a finger vein pattern from the image of only the finger portion using a level set curvature; (d) binarizing the extracted finger vein pattern; (e) Provided is a finger vein authentication system characterized by instructing to thin a binarized finger vein pattern.

  According to a preferred embodiment, extracting the finger vein pattern using the level set curvature in (c) defines (c1) a pixel intensity for each of all pixels constituting an image of only the finger portion. And (c2) a second operation for generating one or more contour sets (level set) that are closed curves generated by connecting pixels having the same pixel intensity among all the pixels. (C3) a third operation that defines the one or more contour lines as a contour line set; and (c4) each point of the one or more contour lines that has a minimum curvature with respect to the outside of the contour line set. A fourth calculation that generates one or more connected lines, and (c5) a fifth calculation that recognizes and extracts the one or more lines as a finger vein pattern. Done.

  According to a preferred embodiment, the curvature in (c4) is

Where:

Is the curvature, and w is the gradient vector of the contour set.

According to a preferred embodiment, the optical filter unit is a near infrared band pass filter (NIR Pass Filter), of hemoglobin contained in the finger vein of the user, oxyhemoglobin (HbO ₂₎ and deoxyhemoglobin (Hb) Only light having a wavelength of 630 to 1,000 nm suitable for sensing is transmitted.

  According to a preferred embodiment, the finger vein authentication system according to the present invention further includes a key input unit configured of a plurality of keys including numbers, letters, and symbols, and outputting a key signal corresponding to the key selected by the user. It is.

The perspective view which shows the external appearance of the finger vein authentication system which concerns on one Embodiment of this invention. FIG. 2 is a diagram for explaining an internal structure of an image processing apparatus in the finger vein authentication system of FIG. The block diagram for demonstrating the internal structure of the said finger vein authentication system of FIG. 1a. 2 is an algorithm flowchart showing a process in which the finger vein authentication system of FIG. 1a extracts a vein pattern according to a command of a level set curvature program. The figure which compared and showed the image regarding the finger from which noise was removed, and the image before noise removal according to the level set curvature program. The figure which shows the screen of the result of having isolate | separated the image of only a finger part from the whole image according to a level set curvature program. The figure which shows a mode that the image of the finger | toe was converted into the form of the contour line according to the level set curvature program. 6 is a diagram illustrating a method in which a level set curvature program discriminates a finger vein pattern using the contour lines of FIG. 6a. Level set curvature Drawing showing values normalized by 8-bit gray level.
The figure which compared and showed the vein image extracted by the level set curvature program, and the vein image by a conventional method.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. First, reference numerals are assigned to constituent elements in each drawing, and it is noted that the same constituent elements are given the same reference numerals as much as possible even if they are shown in other drawings. Should. Further, in the description of the present invention, when it is determined that a specific description of a related known configuration or function obscures the gist of the present invention, a detailed description thereof is omitted.

  Throughout the specification, when any part “includes” what component, unless otherwise specified, other components are not excluded, but other components are excluded. It means that it can be further included. Further, terms such as “... part”, “... device”, “module” described in the specification mean a unit for processing at least one function or operation, and this means hardware, software, or hardware. And a combination of software.

  FIG. 1 a is a perspective view showing an appearance of a finger vein authentication system 1 according to the present invention, and FIG. 1 b is a diagram for explaining an internal structure of a finger vein image processing apparatus 10 in the finger vein authentication system 1. FIG. 2 is a block diagram for explaining the internal configuration of the finger vein authentication system 1.

  The finger vein authentication system 1 according to the present invention obtains an image obtained by irradiating a user's finger with near-infrared light as a digital image using a CCD camera, and then obtains the characteristics of the finger vein pattern from the obtained image. It is a system for identifying individuals by extracting them using a level set curvature program and comparing the extracted pattern features with previously registered user information.

  For example, the finger vein authentication system 1 is installed on the wall near the doorway, and the user inserts (attaches) the finger 20 into the finger resting portion S of the finger vein image processing apparatus 10 for entrance / exit authentication. The entry / exit authentication can be completed easily with only the operation. At this time, if the vein pattern of the user and the vein pattern registered in the system 1 do not match, the entry / exit authentication is rejected.

  On the other hand, in the following description, the system 1 will be described on the assumption that the user completes authentication only by vein pattern recognition, but in other embodiments, before and after the user authenticates the vein pattern, Security can be enhanced by a method of inputting a password by the key input unit 210. For example, after the user inputs the password to the key input unit 210, the user can also be authenticated by a method of further confirming the finger vein pattern.

  Referring to FIGS. 1a, 1b and 2 together, the finger vein authentication system 1 according to the present invention is roughly divided into a finger vein image processing apparatus 10 for acquiring a finger image and a finger vein pattern from the acquired image. It comprises an authentication processing device 20 that extracts and performs user authentication processing. In addition, the finger vein image processing apparatus 10 includes a light source unit 110 configured to emit near-infrared light, and an optical configured to transmit only light having a specific wavelength among incident light. The filter unit 120 and a CCD camera unit 130 configured to perform an operation of converting the captured image into an electrical signal after capturing an image of a finger. The authentication processing device 20 includes a number of keys such as numbers, characters, symbols, and the like. The key input unit 210 outputs a key signal corresponding to the key selected by the user, and various inputs related to user authentication. The display unit 220 includes a display unit 220 that displays a screen and a result screen, a memory unit 230 that stores a level set curvature program, and a control unit 240 that drives the level set curvature program.

  Specifically, the finger vein image processing apparatus 10 has a structure in which a finger is inserted, and an incised cylindrical finger rest S is formed. The main body surface on the upper part of the finger rest S The light source unit 110 is installed, and the optical filter unit 120 and the CCD camera unit 130 are installed on the surface of the lower body of the finger resting unit S.

  For example, when the user places a finger on the finger rest S for user authentication, near-infrared light emitted from the light source 110 passes through the finger on the finger rest S. Then, after passing through the optical filter unit 120, an image of a finger is picked up by a method of reaching the CCD camera unit 130. Thereafter, the finger image captured by the CCD camera unit 130 is stored in the memory unit 230 in accordance with the control of the control unit 240 of the authentication processing device 20. Thereafter, the control unit 240 determines whether or not the finger vein pattern in the finger image matches the previously registered finger vein pattern of the user in accordance with an instruction of the level set curvature program 231. Perform processing operations.

  The light source unit 110 includes, for example, one or a plurality of LEDs or laser diodes having a structure that emits near infrared light having a wavelength of 630 to 1,000 nm. According to a preferred embodiment, an optical waveguide (Light Wave Guide) made of PDMS (Polydimethylsiloxane) material is attached to the rear surface of the light source unit 110, so that light can be incident on the finger without diffusing. .

The optical filter unit 120 includes, for example, a near infrared band pass filter (NIR Pass Filter). The optical filter unit 120 performs a function that allows only the finger vein image to be captured even when the user's finger is exposed to visible light. According to a preferred embodiment, the optical filter unit is configured to transmit only light having a wavelength of 630 to 1,000 nm suitable for sensing oxyhemoglobin (HbO ₂ ) and deoxyhemoglobin (Hb) of hemoglobin.

  The CCD camera unit 130 performs an operation of converting an image relating to the finger F captured by near-infrared light emitted from the light source unit 110 into an electrical signal. The electrical signal image regarding the finger F acquired by the CCD camera unit 130 is stored in the memory unit 230 under the control of the control unit 240.

  FIG. 3 is an algorithm flowchart showing a process in which the control unit 240 extracts a vein pattern from the image relating to the finger F received from the CCD camera unit 130 in accordance with the command of the level set curvature program 231.

  Referring to FIG. 3, the control unit 240 of the authentication processing device 20 drives a “level set curvature program 231” stored in the memory unit 230, and (a) is acquired from the CCD camera unit 130. Processing operation for removing noise from the image relating to the finger (S301), and (b) processing operation for removing the peripheral portion from the image relating to the finger from which noise has been removed, and separating the image of only the finger portion (S302). (C) a processing operation (S303) for extracting a finger vein pattern from an image of only the finger portion using a level set curvature (S303), and (d) the extracted finger vein pattern A binarized processing operation (S304), and (d) a process of thinning the binarized finger vein pattern. The physical operation (S305) is performed.

  Thereafter, the control unit 240 can determine whether the thinned finger vein pattern matches the finger vein pattern registered for the user according to the command of the level set curvature program 231. Needless to say.

  Such a level set curvature program 231 can be an external storage medium such as a flexible disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, magnetic tape, nonvolatile memory card, ROM, etc. It can also be used and supplied to the finger vein authentication system 1.

Hereinafter, the finger vein extraction algorithm provided by the level set curvature program 231 will be described in detail with reference to FIGS.
[Noise removal processing (S301)]
The level set curvature program 231 applied to the finger vein authentication system 1 according to the present invention performs an operation of removing noise from an image related to a finger using, for example, a Gaussian filtering method.

  For example, the level set curvature program 231 calculates a standard deviation with respect to a pixel value for each of the divided blocks, and a calculation process for dividing an image relating to a finger into blocks of a predetermined size, that is, a size of M lines × N pixels. A process, an arithmetic process for selecting the smallest standard deviation and calculating a minimum standard deviation (σmin) quantized to an integer value, and the minimum standard deviation (σmin) within a predetermined range Block with standard deviation (

) Is selected.
At this time, block (

) Can be derived using the following mathematical formula 1.
[Mathematical Formula 1]

here,

Indicates the selected block,

Is a function that obtains the maximum integer value that is smaller than or equal to a given value as a FLOOR function.

  The level set curvature program 231 then selects the selected block (

), A filter coefficient for performing Gaussian filtering is calculated. At this time, the filter coefficient can be calculated from the quantized minimum standard deviation (σmin) by the following mathematical formula 2.
[Mathematical formula 2]

Next, the level set curvature program 231 performs filtering on the selected block (σmin) with the calculated filter coefficient.
Next, the level set curvature program 231 determines the standard deviation of the difference between the selected block value and the Gaussian filtered value of the selected block as the noise measurement value (

), Which is expressed by the following mathematical formula 3.
[Formula 3]

By such a process, the level set curvature program 231 can calculate noise that is unrelated to the finger portion from the image related to the finger, and can remove it.

FIG. 4 shows an image relating to a finger from which noise has been removed according to the level set curvature program 231 compared with an image before noise removal.
[Process of Separating Image of Finger Part Only (S302)]
The level set curvature program 231 applied to the finger vein authentication system 1 according to the present invention performs an operation for separating an image of only a finger portion from an image related to a finger.

The image picked up by the CCD camera unit 130 includes not only an image of the finger part but also an image of surrounding objects.
According to a preferred embodiment, the level set curvature program 231 performs an operation of separating an image of only a finger portion from an entire image using an edge detection function using a divergence of gradient. be able to.

For example, the following mathematical formula 4 can be used as an edge detection function using gradient divergence.
[Formula 4]

FIG. 5 shows a screen obtained as a result of separating the image of only the finger portion from the entire image in accordance with the level set curvature program 231.

[Process for Extracting Finger Vein Pattern Using Level Set Curvature (S303)]
The level set curvature program 231 applied to the finger vein authentication system 1 according to the present invention performs an operation of extracting a finger vein pattern from an image of a finger part using a level set curvature.

  Here, the level set refers to a set of pixels having the same intensity when each pixel value of the image is regarded as an intensity function.

Therefore, if the intensity function is a continuous function, the points forming one level set form one contour line.
Specifically, the level set curvature program 231 includes (c1) a first calculation that defines the pixel intensity for each of all the pixels constituting the finger image, and (c2) the same among all the pixels. A second operation for generating one or more contour sets that are closed curves generated by connecting pixels having pixel intensities to each other; and (c3) a third operation that defines the one or more contour lines as a contour line set. And (c4) a fourth calculation for generating one or more lines connecting the points of the one or more contour lines, the curvature of which is minimized with reference to the outside of the contour line set, and (c5) A process of extracting a finger vein pattern from the finger image by performing a fifth operation of recognizing and extracting the one or more lines as a finger vein pattern I do.

  FIG. 6a shows how the image of a finger is converted into a contour line set form according to the level set curvature program 231, and FIG. 6b shows the level set curvature program 231 using the curvature of the contour set in FIG. 6a. A method for discriminating a vein pattern is shown.

  First, referring to FIG. 6a, an operation for defining a pixel intensity for each of all pixels constituting an image of only a finger portion in a finger image, and pixels having the same pixel intensity among all the pixels. A set of contour lines is created by an operation for generating one or more contour lines (level set) that are closed curves generated by connecting the two.

  Next, referring to FIG. 6b, the level set curvature program 231 generates one or more lines connecting the points of the one or more contour lines that have the minimum curvature with respect to the outside of the contour line set. The finger vein pattern is extracted from the image of the finger by the calculation to perform and the calculation to recognize and extract the one or more lines as the finger vein pattern.

From the viewpoint of geography, the line connecting the points having the smallest curvature with reference to the outside of the contour line set is similar to the method of finding the valleys in the contour lines.
Specifically, the level set curvature program 231 connects the points where the curvature becomes maximum when calculated based on the outside of the contour set in each contour line included in the contour set included in the finger image. The line becomes a ridge line, the line connecting the smallest points becomes a valley, and the vein blood vessel distribution in the finger image corresponds to the valley pattern on the contour line set. It is based on the principle of becoming.

  The point at which the curvature is minimum is given by the following mathematical formula 5, and at each contour line:

Is found by finding the smallest point.
[Mathematical formula 5]

here,

Is the curvature, w is the gradient vector, and the part corresponding to the finger vein is clearly low

Has a value.

As described above, the level set curvature is obtained using the divergence of the normalized gradient, and at this time, the gradient vector is gathered in the ridge portion in the above-described contour line, and in the valley portion. Intuitively confirm that the gradient vector diverges.
FIG.

The values are normalized and shown as 8-bit gray levels, and it can be confirmed that the portion corresponding to the central line of the vein has a clearly small value.

[Process for leveling curvature of level set (S304) and thinning process (S305)]
The level set curvature program 231 applied to the finger vein authentication system 1 according to the present invention includes an operation (S304) for binarizing a finger vein pattern extracted using the level set curvature (S304). An operation (305) for thinning the evolved image is performed.

  Here, the binarization process performed by the level set curvature program 231 is for omitting the storage space of the extracted vein pattern in consideration of the capacity of the memory unit 230, and is a matching process that is a subsequent process. This is to save time.

  The thinning performed by the level set curvature program 231 is to convert a thick vein display into a thin vein display so that a larger amount of vein information can be stored without distortion.

  FIG. 8 shows a final vein image obtained by extracting a vein pattern from a finger image according to the level set curvature program 231, binarizing and thinning the extracted pattern, and existing line tracking (line tracking). FIG. 6 is a view showing a comparison between an image of a first vein according to a method and an image of a second vein according to an existing local threshold method. FIG.

  As can be seen from FIG. 8, the form of the vein image obtained in accordance with the level set curvature program 231 according to the present invention is much clearer than the existing vein image, thereby reducing errors on the boundary surface. Is clear.

  According to the present invention, the level set curvature program 231 applied to the finger vein authentication system 1 can also be recorded on a computer-readable recording medium. Examples of such computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy (registered trademark) disk, optical data storage device, etc., and carrier wave (for example, via the Internet). In the form of transmission). In addition, the computer-readable recording medium is distributed to computer systems connected via a network, and codes that can be read by the computer in a distributed manner are stored and executed. In addition, a functional program, code, and code segment for implementing the present embodiment should be easily inferred by a programmer in the technical field to which the present embodiment belongs.

  Note that, as described above, it has been described that all the components constituting the embodiment of the present invention are combined and operated, but the present invention is not necessarily limited to such an embodiment. That is, as long as it is within the scope of the object of the present invention, all of the components can be selectively combined and operated. In addition, all the constituent elements can be embodied as one independent hardware, but a part or all of each constituent element is selectively combined, and a part combined with one or a plurality of hardware. Alternatively, the present invention can be embodied as a computer program having program modules that perform all functions.

  The above description is merely illustrative of the technical idea of the present invention. If the person has ordinary knowledge in the technical field to which the present invention belongs, the essential characteristics of the present invention are described. Various modifications and variations are possible without departing from the scope. Accordingly, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention, but to illustrate, and the scope of the technical idea of the present invention is limited by these examples. It is not a thing. The protection scope of the present invention should be construed in accordance with the following claims, and all technical ideas within the equivalent scope should be construed as being included in the scope of the right of the present invention.

1: finger vein authentication system, 10: finger vein image processing device, 20: authentication processing device, 110: light source unit, 120: light source filter unit, 130: CCD camera unit, 210: key input unit, 220: display unit, 230 : Memory unit, 231: Level set curvature program, 240: Control unit

Claims (4)

A light source unit configured to emit near-infrared light;
An optical filter unit configured to transmit only light of a specific wavelength among the light incident from the light source unit;
A CCD camera unit configured to capture an image including a user's finger and then convert the captured image into an electrical signal;
A memory unit for storing a level set curvature program for performing an operation of extracting a finger vein pattern from an image including the finger using a level set curvature;
A control unit for driving the level set curvature program;
A finger vein authentication system including:
The level set curvature program is for the control unit.
(A) removing noise from the image including the finger acquired by the CCD camera unit;
(B) separating an image of only the finger portion by removing a peripheral portion from an image including the finger from which noise has been removed;
(C) extracting a finger vein pattern from the image of only the finger part using a level set curvature (level set curvature);
(D) binarizing the extracted finger vein pattern;
(E) Instructing the binarized finger vein pattern to be thinned,
Extracting the finger vein pattern using the level set curvature in (c),
(C1) a first calculation that defines a pixel intensity for each of all the pixels constituting the image of only the finger portion;
(C2) a second operation for generating one or more contour lines that are closed curves generated by connecting pixels having the same pixel intensity among all the pixels;
(C3) a third operation defining the one or more contour lines as a contour line set;
(C4) a fourth operation for generating one or more lines connecting respective points of the one or more contour lines, the curvature of which is minimized with respect to the outside of the contour line set;
(C5) a fifth calculation for recognizing and extracting the one or more lines as a finger vein pattern;
A finger vein authentication system characterized by being performed by: The curvature in (c4) is
Where:
The finger vein authentication system according to claim 1, wherein is the curvature, and w is a gradient vector of the contour set. The optical filter unit is a near-infrared band pass filter (NIR Pass Filter) and is suitable for sensing oxyhemoglobin (HbO ₂ ) and deoxyhemoglobin (Hb) among hemoglobins contained in the finger vein of the user. The finger vein authentication system according to claim 1, wherein the finger vein authentication system is configured to transmit only light having a wavelength of ˜1,000 nm.   2. The finger vein authentication system according to claim 1, further comprising a key input unit configured by a plurality of keys including numbers, letters, and symbols, and outputting a key signal corresponding to the key selected by the user.