KR20030010977A - Method for recognizing Organism and Apparatus therefor - Google Patents

Abstract

PURPOSE: A method for a living body recognition and a security device thereof are provided to discriminate a person from other person by photographing a capillary vessel pattern located between the last knuckle of a finger and a boundary of a fingernail and comparing the capillary vessel pattern with a capillary vessel pattern stored in a database. CONSTITUTION: A database(50) stores capillary vessel pattern data from the last knuckle of a finger to a fingernail boundary portion of each person and information with respect to a specific point of the pattern. An image obtain unit(10) photographs a pattern of a capillary vessel located in a subcutaneous skin from the last knuckle of a finger to a fingernail boundary portion of each person. A pre-processing unit(20) removes a scarfskin image obtained through the image obtain unit(10) and a noise of a light dispersed on the scarfskin, performs a binary coded of a valid blood vessel image, and processes a capillary vessel image. A characteristic extracting unit(30) divides the blood vessel image detected through the pre-processing unit(20) into many blocks, performs the secondary binary coded process, and extracts each characteristic point and characteristic value information of the blood vessel using the capillary vessel data. A recognition control unit(40) receives data of the capillary vessel image and the characteristic value, compares the data with pattern data of a capillary vessel stored in the database(50) for searching matched data. An output unit(60) outputs a corresponding operation signal to a driving circuit in accordance with a recognition result of the recognition control unit(40).

Description

Biometrics and its security devices {Method for recognizing Organism and Apparatus therefor}

The present invention relates to a biometrics system, and in particular, to capture the capillaries located between the last point of the finger and the fingernail borders and to compare each other with the capillary pattern data stored in the database in advance to identify the identity of each individual A method and a security device thereof.

Recently, researches on biometric systems that identify individuals by measuring biological characteristics of the body have been actively conducted, and in particular, the field of vascular recognition used by analyzing the pattern of veins located on the back of the hand or by analyzing the vascular pattern of the retina is precisely determined. High discernment is in the limelight.

However, in the case of the dorsal vein pattern recognizer, the entire size of the back of the hand has to be taken. The size of the device increases. There was a drawback caused.

For this reason, while using capillaries with high personal identification, the user's rejection is minimized, and a low-cost and compact personal identification system has been developed.

Accordingly, an object of the present invention is a biometric method and a security device for capturing capillary patterns positioned between the last node of a finger and a fingernail boundary and comparing each other with capillary patterns recorded in a database to identify whether each individual is identical. To provide.

In addition, another object of the present invention is to provide a biometric method and a security device capable of producing a small and low cost identification device by capturing a capillary pattern of a small area located between the last node of the finger and the nail.

1 is a block diagram showing a biometric security device according to the present invention,

2A and 2B are views for explaining capillary imaging according to FIG. 1,

3 and 4 are vascular diagrams for explaining the feature points of the capillaries of the fingers and their feature values,

5 and 6 are flowcharts for explaining the operation of the biometric security device according to an embodiment of the present invention,

FIG. 7 is a diagram illustrating a reference point setting of an effective portion during capillary imaging.

8 is a diagram illustrating an image segmentation and binarization process of a preprocessor.

* Explanation of symbols for the main parts of the drawings

10: image acquisition means 11: solid state image pickup device

15: light generating means 20: pretreatment unit

30: feature extraction unit 40: recognition control unit

50: database 60: output unit

Technical means of the present invention for achieving the above object, in the system for detecting and recognizing the living body of each individual: capillary pattern data from the last node of each individual finger to the fingernail boundary and information about the characteristic points of the pattern Listed database; Image acquisition means for photographing and capturing a pattern of capillaries located subcutaneously in the skin from the last node of each person's finger to the boundary of the nail; A pre-processing unit for removing the epidermal image and the light scattered from the epidermis obtained through the image acquisition means, binarizing the effective blood vessel image and thinning the capillary image; A feature extractor for dividing the blood vessel image detected by the preprocessor into a plurality of blocks, performing a second binarization process, and extracting each feature point and feature value information of the vessel using thinning data of the capillaries; A recognition control unit receiving image data and feature values of capillaries extracted through the feature extracting unit and comparing the capillary pattern data in the database with each other to determine whether there is a matching data; And an output unit for outputting a corresponding operation signal to a driving circuit side according to the recognition result of the recognition control unit.

In addition, the technical method of the present invention for achieving the above object, in the method for detecting and recognizing the living body of the individual: a first step of capturing a pattern image of capillaries located between the last point of the finger and the boundary of the nail ; A second step of acquiring capillary images by removing epidermal data and optical noise of a finger from the photographed images; Dividing the capillary image into a plurality of matrix blocks and performing binarization for each block; A fourth step of extracting feature points of the capillaries by analyzing the capillary images and extracting feature value information about the feature points; And a fifth step of comparing the capillary pattern data and the feature values recorded in the database with each other by using the acquired capillary image data and the feature values of the feature points to recognize whether they are the same.

In addition, another technical method of the present invention for achieving the above object is a method for detecting and recognizing a living body of an individual: inserting a specific finger into the finger fixture; Photographing a pattern image of capillaries located between the last node of the finger and the boundary point of the nail; Setting a reference point by searching for a boundary point of the nail using the captured image; Removing skin data and optical noise of a finger from an image located between the searched reference point and the last node of the finger; Thinning the capillary image after binarizing the effective capillary image after removing the noise; Dividing the extracted capillary image into a plurality of matrix blocks and performing secondary binarization of the divided capillary image for each block; Analyzing the thinned capillary image to extract branch points, break points, branches, transformation points, and feature points of the vessels, and extract detailed feature value information about the feature points; Firstly detecting a plurality of comparison data within a tolerance range by directly comparing the capillary pattern data recorded in the database using the obtained second binarization data; Comparing the feature points and the feature value information of the blood vessels extracted with the feature values of the plurality of comparative data detected primarily to recognize whether they are the same; And outputting a system operation control signal according to the recognition result.

Hereinafter, the present invention will be described in more detail with reference to the accompanying drawings.

1 is a block diagram showing a biometric security device according to the present invention, Figure 2a is a view showing the image acquisition means of Figure 1, Figure 2b is a view showing the image acquisition portion of the finger, the image acquisition means (10) ), A preprocessor 20, a feature extractor 30, a recognition controller 40, a database 50, and an output unit 60.

The database 50 includes capillary pattern data from the last segment of each individual finger to the boundary of the nail, and information on the characteristic points and characteristic values of the pattern, and the image acquisition means 10 includes the finger of each individual. It is configured to acquire by capturing a pattern of capillaries located subcutaneously in the skin from the last node of the nail to the boundary of the nail, the pre-processing unit 20 is scattered from the epidermal image and the epidermis obtained through the image acquisition means 10 After removing the noise of light, the effective blood vessel image is binarized and the capillary image is thinned. The feature extractor 30 divides the blood vessel image detected by the preprocessor 20 into a plurality of blocks. After performing the second binarization process and using the thinning data of the capillaries, it is configured to extract each feature point and feature value information of the blood vessels.

In addition, the recognition controller 40 receives the capillary image data and the feature values extracted through the feature extractor 30, respectively, and compares the pattern data of the capillary blood vessels in the database 50 to see if there is a matched data. And the output unit 60 is configured to output a corresponding operation control signal to the corresponding driving circuit side according to the recognition result of the recognition control unit 40.

On the other hand, the image acquisition means 10 is composed of a solid-state imaging device 11 for capturing capillaries and light generating means 15 for emitting light with a finger, the light generating means 15 is an infrared light source It is preferable to use any one of a laser light source or a halogen light source, and the solid state imaging device 11 directly acquires an image in which the emitted light source is reflected from the skin of the finger and subcutaneously, or the optical fiber or the opaque tube (not shown). It can be in close contact with the through to obtain an image transmitted through the skin.

In addition, the output unit 60 can be applied in various ways according to the field, such as door access control, computer login control, network access control or various encryption devices, such as door access or computer booting and network according to personal identification Make the connection available.

2A and 2B are diagrams for explaining capillary imaging, the image acquisition means 10 of the tubular finger inlet (not shown) of the finger to obtain an accurate blood vessel image of the user's finger (1) A finger fastener 9 for designating the degree of insertion is provided.

The finger inserted as described above is photographed by the solid state image pickup device 11, and as shown in FIG. 2B, the part 7 from the last node 3 of the finger 1 to the boundary of the nail 5 is photographed. Done.

3 and 4 classify the plurality of feature points extracted from the feature extraction unit 30 of FIG. 1 and the feature values corresponding to the information of the feature points by shape and thickness, respectively.

That is, the feature points may be classified into FIGS. 3A to 3F according to the shapes and features of blood vessels, which are divided from the main blood vessels, the cutting points at which the blood vessel ends, and the branching points at which the plurality of blood vessels meet or divide. The branches can be classified into shapes such as branching branches, transition points that temporarily change in thickness, convex pears with thickened blood vessels, and concave pears with small blood vessels, and each feature point can be classified into FIGS. 4A to 4C again according to thickness. It can be divided into thin, thick, spread, and the like in order from Figure 4a.

In addition, although not shown in the drawings, the positional coordinates of the feature points and the change in brightness and pulsation of blood vessels according to blood flow can be taken as feature values, and the brightness and pulsation can be used as information for judging that blood vessels are particularly living vessels. Counterfeit can be prevented.

In this way, the blood vessels to be recognized as the region 7 between the finger nail 5 and the last segment 3 of the finger is because the epidermis of the skin in this region is thin, so that capillary images can be easily obtained.

In addition, the capillary tube is similar to the skin and has a high transmittance, and scans a separate light source 15 to increase the visibility of blood vessels, and then acquires an image with the imaging device 11. The imaging device 11 is a CCD device or the like. An optical device such as an imaging device and a lens is used, which is located at the vertical top of the photographing site as shown in FIG.

The captured image is sent to the preprocessing unit 20 to remove the noise of the skin and scattered light from the epithelium to improve the visibility of the capillary image. To this end, the preprocessing unit 20 performs image processing through a process such as smoothing, binarization, thinning, etc., and extracts an effective blood vessel image, using a Fast Fourier Transform (FFT) method or a Gabor in frequency space. Apply filters, etc.

The image processed by the preprocessor 20 is sent to the feature extractor 30 to extract data valid for recognition. The feature extractor 30 divides the image information into blocks of a predetermined size for each image. Binarization is performed for each block.

In addition, according to the shape of the capillaries, as shown in Fig. 4, the cutting point, the branch point, the branch, the conversion point, the shape of the ship, etc. are classified into feature points, and then, various feature value information about the feature points is extracted.

In this case, the feature value corresponding to the information of the feature point is data related to the thickness, contrast and pulsation of the blood vessel along with the position information corresponding to the two-dimensional coordinates of the feature point.

The thickness of the blood vessel is divided into thin, thick or spread according to its thickness as shown in FIG. 4 and used as information on the feature points, and the degree of change of the intensity of each image is measured and the degree of change is examined per unit time. Infer the pulsation. This is used as important information for determining blood vessels of a living person, and the degree of change and the average brightness value are used when recognizing feature values.

The extracted data is sent to the recognition controller 40 and compared with the data stored in advance in the database 50 to determine whether the recognition.

The overall operation of the present invention configured as described above will be described with reference to FIGS. 5 and 6 and the accompanying drawings.

First, a user inserts a specific finger into the finger insertion hole of the security device and pushes it to the finger fastener 9 as shown in FIG.

When the finger is inserted, the light generating means 15 of the image capturing means 10 emits infrared, laser or halogen rays to the finger, and the solid-state image sensor 11 between the last node of the finger and the fingernail boundary point. The image is taken in sequence a plurality of times (S10). Various methods can be applied by recognizing whether a finger is inserted or not, using an optical sensor using a light-emitting and light-receiving diode, configuring a finger stopper 9 as a limit switch, or installing a manual switch that the user presses directly. This makes it easy to see if a finger is inserted.

As described above, in order to normalize the image of the finger photographed as the specific part of the finger is photographed, the preprocessing unit 20 should set a reference point. This unnecessary image is eliminated. That is, as shown in FIG. 7, the center of the nail is taken as a center point, and a straight line is drawn downward while maintaining the same angle and distance from both center lines at the center point, and a horizontal line is drawn at the nail boundary to be perpendicular to both boundary lines, and both sides meet the horizontal line. The site of the boundary line is the reference point (A, B) of normalization. Therefore, the rectangular portion formed between the reference point and the node of the finger becomes an effective portion of normalization, and the image of the effective portion is rotated and enlarged to normalize (S11).

Subsequently, the preprocessing unit 20 obtains a valid capillary image through a smoothing process of removing noise corresponding to an epidermal image or scattered light at an effective site, binarizes the capillary image with digital data, and then captures the capillary image. It becomes thinner (S12-S14).

The thinned capillary image data as described above is output to the feature extractor 30, the feature extractor 30 is provided with such data image processing process (S15, S16) using the pattern matching of blood vessels, Image processing processes S17 and S18 using the feature points are performed separately.

In the image processing process using the pattern matching, as shown in FIG. 8, the capillary image obtained through the preprocessor 20 is divided into a plurality of matrix blocks (S15), and the divided capillary image is blocked. Binarization is performed for each output to the recognition controller 40 (S16).

On the other hand, the image processing process using the feature point is to analyze the binarized and thinned capillary image through the pre-processing unit 20, as shown in Figs. The feature points of the back are extracted (S17), and the feature values of position coordinates, thicknesses, brightness, and pulsation information of adjacent blood vessels of the feature points are measured and output to the recognition controller 40 (S18). When measuring the brightness of the blood vessels, the blood flow flowing through the blood vessels is constantly changed according to time, so that the average brightness value is obtained by analyzing a plurality of consecutive blood vessel images, and the change in brightness is based on the pulsation. Can be easily inferred.

Subsequently, the recognition controller 40 directly compares and compares various capillary pattern data stored in the database 50 by using the vascular pattern-related binarization data acquired through the feature extractor 30 to obtain a plurality of tolerances within a tolerance range. The comparison data is first detected (S19), and then the feature points and the feature values extracted above are compared with the feature points and feature values of the plurality of comparison data detected primarily to determine and recognize the identity (S20). , S21).

In addition, the process of comparing the characteristic point information and the characteristic value information of the blood vessel in the recognition control unit 40 can be broken down as shown in Figure 6, the capillary feature points obtained first and the plurality of capillary feature points contained in the database 50 Compare the coordinate matching of (S20-1), compare the coordinates and compare the average thickness of the blood vessels adjacent to the feature point (S20-2), a plurality of capillaries contained in the feature point of the capillary and the database 50 obtained By comparing the average brightness value of the vascular feature points to determine whether the match (S20-3), and finally selected two other feature points adjacent to each of the capillary feature points obtained and the plurality of capillary feature points contained in the database 50 Thereafter, the feature points and the feature values may be compared with each other by comparing the circle diameters generated based on the three selected feature points (S20-4).

As such, the reason for comparing the characteristic point information and the characteristic value information of blood vessels through a multi-step process is to increase the accuracy of comparison, and it is preferable to reduce or add more steps as necessary.

Subsequently, the recognition controller 40 outputs an operation control signal to the door access control unit, the computer boot controller, the network connection controller, or the like, which is the output unit 60 according to the recognition result (S22).

While specific embodiments of the present invention have been described and illustrated above, it will be apparent that the present invention may be embodied in various modifications by those skilled in the art. Such modified embodiments should not be individually understood from the technical spirit or the prospect of the present invention, but should fall within the claims appended to the present invention.

Therefore, in the present invention, by capturing the capillary pattern located between the last node of the finger and the boundary point of the fingernail and comparing the capillary pattern recorded in the database with each other to identify whether each individual is the same, as well as further improving individual identification ability. Since the imaging apparatus can be configured in a small size, it is inexpensive, and the size of the apparatus can be easily applied to a processing device such as a door, a set-top box, or a personal computer, and can be linked with an existing security apparatus. In addition, by detecting the brightness change and pulsation of the blood vessel caused by the blood flow through the blood vessel to use as recognition information, it is possible to maximize the accuracy and security of the recognition by eliminating the possibility of forgery.

Claims (14)

In a system that detects and recognizes a living body of an individual: A database containing capillary pattern data from the last node of each individual finger to the nail boundary and information on the feature points of the pattern; Image acquisition means for photographing and capturing a pattern of capillaries located subcutaneously in the skin from the last node of each person's finger to the boundary of the nail; A pre-processing unit for removing the epidermal image and the light scattered from the epidermis obtained through the image acquisition means, binarizing the effective blood vessel image and thinning the capillary image; A feature extractor for dividing the blood vessel image detected by the preprocessor into a plurality of blocks, performing a second binarization process, and extracting each feature point and feature value information of the vessel using thinning data of the capillaries; A recognition control unit receiving image data and feature values of capillaries extracted through the feature extracting unit and comparing the capillary pattern data in the database with each other to determine whether there is a matching data; And And an output unit for outputting a corresponding operation signal to a driving circuit side according to the recognition result of the recognition control unit. The method according to claim 1, The image acquisition means is a biometric security device, characterized in that for using the finger fixture to obtain an accurate blood vessel image of the finger. The method according to claim 1, The image acquisition means includes a solid-state imaging device for capturing capillaries and light generating means for emitting light to the finger. The method according to claim 3, The light generating means, biometric security device, characterized in that using any one of the infrared light source, laser light source or halogen light source. The method according to claim 3, The solid-state image pickup device, the biometric security device, characterized in that directly obtained by the emitted light source is reflected from the skin and subcutaneous finger or the optical fiber or opaque tube in close contact with the skin to obtain an image transmitted through the skin . The method according to claim 1, The output unit is a biometric security device, characterized in that any one of the door access control, computer login control, network access control. In the method of detecting and recognizing an individual's living body: A first step of capturing a pattern image of capillaries located between the last node of the finger and the boundary of the nail; A second step of acquiring capillary images by removing epidermal data and optical noise of a finger from the photographed images; Dividing the capillary image into a plurality of matrix blocks and performing binarization for each block; A fourth step of extracting feature points of the capillaries by analyzing the capillary images and extracting feature value information about the feature points; And And a fifth step of comparing the capillary pattern data and the feature values recorded in the database with the capillary image data and the feature values of the feature points to recognize whether they are identical. Way. The method according to claim 7, The second step, Setting a reference point by searching for a boundary point of the nail using the image photographed in the first step; Removing skin data and optical noise of a finger from an image located between the searched reference point and the last node of the finger; And thinning the capillary image after binarizing the effective capillary image after removing the noise. The method according to claim 7, The fourth step, Analyzing the thinned capillary image and extracting feature points corresponding to branching points, cutting points, branches, transformation points, and vessels of the blood vessels; And measuring position coordinates, thicknesses, and brightness of blood vessels with respect to the feature points, respectively. The method according to claim 9, In the step of measuring the brightness of the blood vessels, biometric recognition method further comprising the step of inferring the pulsation according to the change of the brightness while taking the average brightness value by analyzing a plurality of consecutive blood vessel images. The method according to claim 7, The fifth step, Comparing coordinate agreements of the acquired capillary features with the plurality of capillary features in the database; And comparing the coordinates and comparing average thicknesses of blood vessels adjacent to feature points. The method according to claim 7, And the fifth step includes comparing the obtained capillary feature points with average brightness values of the plurality of capillary feature points included in the database to determine whether they match. The method according to claim 7, The fifth step may include selecting two feature points adjacent to each of the acquired capillary feature points and a plurality of capillary feature points recorded in a database; And comparing the diameters of the circles generated based on the three feature points. In the method of detecting and recognizing an individual's living body: Inserting a specific finger into the finger fixture; Photographing a pattern image of capillaries located between the last node of the finger and the boundary point of the nail; Setting a reference point by searching for a boundary point of the nail using the captured image; Removing skin data and optical noise of a finger from an image located between the searched reference point and the last node of the finger; Thinning the capillary image after binarizing the effective capillary image after removing the noise; Dividing the extracted capillary image into a plurality of matrix blocks and performing secondary binarization of the divided capillary image for each block; Analyzing the thinned capillary image to extract branch points, break points, branches, transformation points, and features of the vessels, and extract detailed feature value information about the feature points; Firstly detecting a plurality of comparison data within a tolerance range by directly comparing the capillary pattern data recorded in the database using the obtained second binarization data; Comparing the feature points and the feature value information of the blood vessels extracted with the feature values of the plurality of comparative data detected primarily to recognize whether they are the same; And And outputting a system operation control signal according to the recognition result.