JP2006004294A - Individual identifying device and individual identifying method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an individual identifying device and an individual identifying method that can certainly identify an individual even when a high-quality eye image is not acquired. <P>SOLUTION: A comprehensive determining section 150 identifies a subject based on a collation result of a blink pattern obtained from a blink pattern data matching section 136 and a collation result of an iris code obtained from an iris code matching section 146. A user is identified using not only iris unique to each person and a blink pattern that can be known by himself/herself, so that high authentication accuracy can be obtained even when the image quality of eye circumferential image is low. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an individual identification device and an individual identification method.

  2. Description of the Related Art Conventionally, development of an individual identification technique for identifying an individual by generating iris data from an eye image including an iris and comparing the generated iris data with registered data has been underway. For example, the following Patent Documents 1 and 2 disclose a system for identifying an individual using an iris feature that is unique for each individual.

  The price of devices using this individual identification technology has also been reduced. Recently, for the purpose of improving security, individual devices have been identified not only for stationary devices but also for various portable information devices such as mobile phones. Functions are being installed. Portable information devices having an iris authentication function are disclosed in Patent Documents 3 to 4 below.

Japanese Patent Publication No. 5-84166 JP-T 8-504979 JP 2001-333171 A JP 2002-330318 A JP 2003-108983 A

  However, when the conventional iris authentication technology is applied to a portable information device, since the device is lightweight and compact, a so-called camera shake occurs at the time of photographing, and there is a possibility that an eye image with good image quality cannot be obtained. In addition, when the shooting location is outdoors, the external light is superimposed on the eye image, and the image data may be further deteriorated. Similarly, a stationary device equipped with an iris authentication function may deteriorate the eye image if the user (subject) is not stationary when taking the eye image. As described above, when the image quality of the eye image is low, the feature for specifying each individual cannot be accurately extracted from the eye image data, and as a result, the identification accuracy of the individual decreases.

  The present invention has been made in view of such problems, and its object is to provide a new and improved individual that can reliably identify an individual even when a high-quality eye image cannot be obtained. To provide an identification device and an individual identification method.

  In order to solve the above-described problem, according to a first aspect of the present invention, an image input unit that acquires an image including an eye of an individual to be tested and a peripheral portion thereof, a blink pattern is extracted from the image, and the extracted blink The blink pattern matching unit that calculates the degree of coincidence between the pattern and the pre-registered blink pattern, and the iris feature is extracted from the image, and the degree of coincidence between the extracted iris feature and the pre-registered iris feature is calculated. An iris matching unit for calculating, and at least an individual identification unit for identifying the test individual based on the matching degree of the blink pattern calculated by the blink pattern matching unit and the matching degree of the iris feature calculated by the iris matching unit An individual identification device is provided. According to this configuration, the test individual is identified based on the iris and the blink pattern. Therefore, even if the image quality is low, high identification accuracy can be obtained.

  The blink pattern is preferably defined by the opening / closing time of the test subject's eyelids within the predetermined time, the number of times the test individual's eyelids are opened / closed, or the combination of the test subject's eyelid opening / closing time and the number of times of opening / closing. It is possible to create a blink pattern with little overlap for each test individual.

  Further, by providing a blink timing notification means for notifying the test individual of the timing of blinking, the test individual can perform the blinking operation accurately.

In order to solve the above-described problem, according to a second aspect of the present invention, an image input unit that acquires an image including an eye of an individual to be tested and its peripheral part, and one or more characteristic parts are extracted from the image. A feature part matching unit for calculating the degree of coincidence between one or more extracted feature parts and one or more feature parts registered in advance, and extracting iris features from the image and extracting the iris An iris matching unit that calculates the degree of coincidence between the feature of the image and a pre-registered iris feature, and at least the degree of coincidence of one or more feature parts calculated by the feature part matching part and the iris matching part An individual identification unit for identifying a test individual based on the degree of coincidence of iris features;
There is provided an individual identification device characterized by comprising: According to this configuration, the test individual is identified based on the feature parts located around the iris and the eyes. Therefore, even if the image quality is low, high identification accuracy can be obtained. And it is preferable to employ | adopt eyes, a mole, eyebrows etc. as a characteristic part.

  Further, by providing an image inclination correcting means for correcting the inclination of the image, it is possible to accurately extract the iris, the blinking pattern, and the characteristic part from the image. As a result, the identification accuracy is improved.

  The individual identification device is preferably incorporated in a portable information device such as a cellular phone.

  In order to solve the above-described problem, according to a third aspect of the present invention, an image input step for acquiring an image including the eye of the subject individual and its peripheral region, a blink pattern extracted from the image, and the extracted blink The blink pattern matching process for calculating the degree of coincidence between the pattern and the pre-registered blink pattern, the feature of the iris extracted from the image, and the degree of coincidence between the extracted iris feature and the pre-registered iris feature An iris matching step for calculating, and at least an individual identification step for identifying a test individual based on the degree of coincidence of the blink pattern calculated in the blink pattern matching step and the degree of coincidence of the iris feature calculated in the iris matching step An individual identification method is provided. According to this method, the test individual is identified based on the iris and the blink pattern. Therefore, even if the image quality is low, high identification accuracy can be obtained.

  In order to solve the above-described problems, according to a fourth aspect of the present invention, an image input step for acquiring an image including the eye of the subject individual and its surrounding region, and extracting one or more characteristic regions from the image A feature part matching step for calculating the degree of coincidence between the extracted one or more feature parts and one or more feature parts registered in advance, and extracting iris features from the image and extracting the iris An iris matching step for calculating the degree of coincidence between the feature of the image and the pre-registered iris feature, and at least the degree of coincidence of one or more feature parts calculated in the feature part matching step and the iris matching step And an individual identification step of identifying a test individual based on the degree of coincidence of the iris features. According to this method, the test individual is identified on the basis of the iris and the characteristic part located around the eye. Therefore, even if the image quality is low, high identification accuracy can be obtained.

  According to the present invention, an individual can be identified with high accuracy.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the present specification and drawings, components having substantially the same functional configuration are denoted by the same reference numerals, and redundant description is omitted.

<First Embodiment>
FIG. 1 shows the configuration of a user identification device 100 (individual identification device) according to the first embodiment of the present invention. For example, the user identification device 100 is used in a portable information device such as the mobile phone 900, PDA (Personal Digital Assistant), or notebook personal computer shown in FIG. 2, or a stationary information device such as a desktop personal computer. It is incorporated and used. Hereinafter, this embodiment will be described in the case where the user identification device 100 is incorporated in the mobile phone 900.

  The user identification device 100 identifies each user based on the eyes of the user as a subject (test individual) and the surrounding image (hereinafter referred to as “eye peripheral image”). As shown in FIG. 2, the mobile phone 900 in which the device 100 is incorporated preferably includes a camera 902, an illumination lamp 904, and a timing lamp 906 (timing notification means) in addition to the basic functions as a mobile phone. The function of the timing lamp 906 will be described later.

  As shown in FIG. 3, the illumination light 910 emitted from the illumination lamp 904 included in the mobile phone 900 reaches the user's eyeball 1 and its periphery, and the reflected light 912 enters the camera 902. Accordingly, the user identification device 100 incorporated in the mobile phone 900 acquires the user's eye peripheral image 10 (see FIG. 4). In the present embodiment, the eye peripheral image 10 is the shape and color (saturation, hue, brightness) of the eye and the peripheral part of the eye (for example, eyelid 12, pupil 14, iris 16, sclera 18, eyelashes 20, eyebrows 22). ).

  Next, the configuration of the user identification device 100 will be described with reference to FIG.

  The light emitting circuit 110 emits light that blinks at predetermined intervals in synchronization with the clock signal output from the time control circuit 112. When the user identification device 100 is incorporated in the mobile phone 900 shown in FIG. 2, the light emitting circuit 110 functions as a drive circuit for a timing lamp 906 that emits timing light. That is, the timing light output from the timing lamp 906 blinks at a predetermined interval, and this blinking interval is controlled by the time control circuit 112. Note that the timing light may be emitted from the illumination lamp 904. In this case, the illumination lamp 904 functions as an output source of both the illumination light 910 and the timing light, and the timing lamp 906 is omitted.

  The image input unit 120 has a function of receiving the reflected light from the user's eyeball 1 and its surroundings, that is, the eye periphery image 10, and providing it to the blink detection unit 130 and the iris coding unit 140.

  The blink detection unit 130 detects the user's blink operation (opening / closing operation of the eyelid 12) based on the eye peripheral image 10 (moving image) continuously given from the image input unit 120. This blinking operation can be detected based on, for example, whether or not the eye periphery image 10 includes an image of the pupil 14, the iris 16, or the sclera 18. When the pupil 14 or the iris 16 is adopted as the detection reference for the blink operation, the blink detection unit 130 operates as follows.

  In general, the pupil 14 forms a circular region with low brightness. The blink detection unit 130 determines that the eyelid 12 is in the open state when the circular area with low brightness (that is, the pupil 14) is detected at the center of the eye peripheral image 10, and the circular area is not detected. , 12 is determined to be closed. Then, the blink detection unit 130 detects that the open state and the closed state of the basket 12 are repeated within a predetermined period, and thereby determines that the blink operation is being performed. This determination result is given to the blink pattern data generation unit 132.

  In addition, the blink detection unit 130 can also determine the blink operation by detecting the presence or absence of a specular reflection spot of the illumination light 910 formed on the cornea without using the image of the pupil 14.

  The blink pattern data generation unit 132 generates blink pattern data from the detection result information of the blink operation given from the blink detection unit 130. Here, for example, the blink pattern data generation unit 132 assigns “1” to this state if the cocoon 12 is open in a predetermined time segment, and assigns “0” to this state if the cocoon 12 is closed. As a result, the blink pattern is quantized into a bit string.

  In the blink pattern data registration unit 134, blink pattern data is registered in advance by a user of the mobile phone 900. When a plurality of users use the mobile phone 900, the blink pattern data registration unit 134 registers the blink pattern data of each user. Further, a blink pattern data registration unit 134 may be provided outside the mobile phone 900. For example, when the mobile phone 900 has a function of accessing the Internet, the blink pattern data registration unit 134 can be configured as a database server on the Internet.

  The blink pattern data matching unit 136 acquires blink pattern data (hereinafter referred to as “test blink pattern data”) from the blink pattern data generation unit 132 and is registered in the blink pattern data registration unit 134 and the blink pattern data registration unit 134. Collation with one or more blink pattern data (hereinafter referred to as “registered blink pattern data”) is performed. The blink pattern data matching unit 136 transmits the blink pattern matching result to the overall determination unit 150.

  On the other hand, when the eye coding unit 140 acquires the eye peripheral image 10 from the image input unit 120, the iris coding unit 140 extracts the feature of the iris 16 from the eye peripheral image 10 and codes it. Then, the iris coding unit 140 transmits the iris code to the iris code matching unit 146.

  In the iris code registration unit 144, the user of the mobile phone 900 registers his / her iris code in advance. When a plurality of users use the mobile phone 900, the iris code of each user is registered in the iris code registration unit 144. Further, an iris code registration unit 144 may be provided outside the mobile phone 900. For example, when the mobile phone 900 has a function of accessing the Internet, the iris code registration unit 144 may be configured as a database server on the Internet together with the above-described blink pattern data registration unit 134 or separately. .

  The iris code matching unit 146 obtains an iris code (hereinafter referred to as “test iris code”) from the iris coding unit 140, and the test iris code and one or more irises registered in the iris code registration unit 144. Check against the code (hereinafter referred to as “registered iris code”). Then, the iris code matching unit 146 transmits the result of matching the iris code to the comprehensive determination unit 150.

  Based on the blink pattern matching result obtained from the blink pattern data matching unit 136 and the iris code matching result obtained from the iris code matching unit 146, the comprehensive determination unit 150 determines a subject who is operating the mobile phone 900. Identify. When the comprehensive determination unit 150 determines that the subject is a person who has the right to use the mobile phone 900 (hereinafter referred to as “authorized user”), the comprehensive determination unit 150 sends a control unit (not shown) to the mobile phone 900. The subject is instructed to permit the use of the mobile phone 900.

  The operation of the user identification device 100 according to the present embodiment configured as described above will be described.

  The user identification device 100 operates by switching at least between the user registration mode and the user identification mode. In the user registration mode, the user identification device 100 stores the blink pattern (blink pattern data) set by each user in the blink pattern data registration unit 134 in response to a request from the user of the mobile phone 900. The user's iris (iris code) is stored in the iris code registration unit 144. In the user identification mode, the user identification device 100 identifies each user by using the blink pattern data registered in the blink pattern data registration unit 134 and the iris code registered in the iris code registration unit 144 in the user registration mode. Identify. Then, the use of the mobile phone 900 (for example, the use of voice communication, data communication, etc.) is permitted only to users who have been authenticated as authorized users from this identification result. Hereinafter, each of the user registration mode and the user identification mode will be described in detail.

[User registration mode]
The light emitting circuit 110 included in the user identification device 100 emits timing light (illumination light 910) that blinks at a predetermined interval in synchronization with the clock signal output from the time control circuit 112. As shown in FIG. 3, the timing light reaches the eyeball 1 of the user as a registrant and its peripheral portion, and the reflected light 912 enters the camera 902. The illumination light 910 also serves as a timing light that informs the registrant of the blink timing.

  The blink pattern is defined by the number of blinks in a predetermined period, the length of each blink (the length of time when the eyes are closed and the time when the eyes are opened), or a combination thereof. At this time, since the illumination light 910 flickers at a constant cycle, the registrant can blink according to a pattern determined by the registrant.

  In this manner, the eye peripheral image 10 (see FIG. 4) including the registrant's blink pattern is input to the image input unit 120 of the user identification device 100. Then, the eye peripheral image 10 is sent to the blink detection unit 130 and the iris coding unit 140 via the image input unit 120.

  The blink detection unit 130 detects the blink operation of the registrant based on the eye peripheral image 10 (moving image) continuously given from the image input unit 120, and transmits the detection result information to the blink pattern data generation unit 132. .

  The blink pattern data generation unit 132 quantizes the blink operation detection result information given from the blink detection unit 130 to generate blink pattern data. The blink pattern data generated in this way is registered in the blink pattern data registration unit 134 as the registered blink pattern data of the registrant. Similarly, registered blink pattern data of a plurality of other registrants may be registered in the blink pattern data registration unit 134.

  On the other hand, the eye peripheral image 10 transmitted from the image input unit 120 to the iris coding unit 140 is iris coded here. The iris data is registered in the iris data registration unit 144 as the registered iris data of the subject. Similarly, registered iris data of a plurality of other subjects may be registered in the iris data registration unit 144.

  As described above, when the registered blink pattern data of one or a plurality of users is registered in the blink pattern data registration unit 134 and the registered iris data is registered in the iris data registration unit 144, each registrant uses the authorized use. Become a person. Then, the user identification device 100 ends the user registration mode.

  A user ID may be assigned to each registrant in advance. This user ID is preferably unique for each user (registrant), but may be the name of each user or a password determined by each user. In the user registration mode, it is preferable that the blink pattern data and the iris code are registered in the blink pattern data registration unit 134 and the iris code registration unit 144, respectively, in association with the user ID.

[User Identification Mode]
The user identification device 100 according to the present embodiment transitions to the user identification mode after the above-described user registration mode. In the user identification mode of the user identification device 100, two sub-modes are prepared: a first user identification mode and a second user identification mode.

  When set to the first user identification mode, the user identification device 100 generates test blink pattern data and a test iris code from the eye peripheral image 10 acquired from the user as the subject, and these and the blink pattern data registration unit The subject is identified by comparing and collating one or more registered blink pattern data stored in 134 and one or more registered iris codes stored in the iris code registration unit 144.

  On the other hand, when the second user identification mode is set, the user identification device 100 first requests the subject to input a user ID. Using this user ID as a key, one registered blink pattern data is extracted from the blink pattern data registration unit 134 and one registered iris code is extracted from the iris code registration unit 144. Then, the user identification device 100 generates the test blink pattern data and the test iris code from the eye peripheral image 10 acquired from the subject, and one of the registered blink pattern data and iris code extracted from these and the blink pattern data registration unit 134. A subject is identified by comparing and collating one registered iris code extracted from the registration unit 144.

  As described above, in the first user identification mode, all the registered blink pattern data stored in the blink pattern data registration unit 134 are compared with the test blink pattern data and all stored in the iris code registration unit 144. The registered iris code is to be compared with the test iris code. On the other hand, in the second user identification mode, one of a plurality of registered blink pattern data stored in the blink pattern data registration unit 134 is a comparison target of the test blink pattern data and is stored in the iris code registration unit 144. One of the plurality of registered iris codes is to be compared with the test iris code. Therefore, the comparison and verification process can be performed at high speed with fewer resources in the second user identification mode. However, in the case of the second user identification mode, it is necessary to obtain a user ID from the subject. Hereinafter, the operation of the user identification device 100 in the user identification mode will be described while comparing the first user identification mode and the second user identification mode.

  In the case of the second user identification mode, the user identification device 100 first requests the subject to input a user ID. In the first user identification mode, the user ID input request process is omitted.

  Then, the user identification device 100 acquires the eye peripheral image 10 from the user as a subject, and generates blink pattern data (test blink pattern data) and an iris code (test iris code) from the eye peripheral image 10. The operations of the light emitting circuit 110, the time control circuit 112, the image input unit 120, the blink detection unit 130, the blink pattern generation unit 132, and the iris coding unit 140 at this time are the same as those in the above-described user registration mode. It is.

  Subsequently, the blink pattern data matching unit 136 compares the test blink pattern data generated by the blink pattern generation unit 132 with one or more registered blink pattern data stored in the blink pattern data registration unit 134. Do. The iris code matching unit 146 compares the test iris code generated by the iris coding unit 140 with one or more registered iris codes stored in the iris code registration unit 144. The comparison and collation processing in the blink pattern data matching unit 136 and the iris code matching unit 146 and the subsequent determination processing in the comprehensive determination unit 150 are different between the first user identification mode and the second user identification mode.

  In the case of the first user identification mode, the blink pattern data matching unit 136 compares the test blink pattern data with all the registered blink pattern data registered in the blink pattern data registration unit 134, and some similar And the degree of coincidence (hereinafter referred to as “blink pattern data coincidence”) and the degree of coincidence is transmitted to the overall determination unit 150 as a blink pattern data collation result. Further, the iris code matching unit 146 compares the test iris code with all the registered iris codes registered in the iris code registration unit 144, and compares several similar codes with respective matching degrees (hereinafter, “ The degree of coincidence of iris code ”is calculated, and this degree of coincidence is transmitted to the overall determination unit 150 as an iris code collation result. Then, the overall determination unit 150 identifies the subject based on the blink pattern data matching result and the iris code matching result.

  In order to identify the subject, for example, the comprehensive determination unit 150 identifies a registered iris code that is closest to the subject iris code based on the degree of coincidence of the iris code, and acquires a user ID associated with the registered iris code. . If the degree of coincidence of the blink pattern data between the registered blink pattern data associated with the acquired user ID and the test blink pattern data is equal to or higher than a predetermined level (approximate), the comprehensive determination unit 150 Is a user identified by the user ID (that is, a authorized user). On the other hand, if the blink pattern data coincidence is below a predetermined level (if it is not approximated), the overall determination unit 150 registers the registered iris that is second closest to the test iris code based on the iris code coincidence. Identify the code and do the same. By repeating the above processing, the comprehensive determination unit 150 can determine whether or not the subject is a authorized user, and can further identify which of the authorized users the subject is.

  In addition, it is also preferable to employ the following method to identify the subject in the first user identification mode.

  The overall determination unit 150 plots the plurality of blink pattern data matching degrees obtained from the blink pattern data matching unit 136 and the plurality of iris code matching degrees obtained from the iris code matching unit 146 in a two-dimensional space. Next, the overall determination unit 150 selects the pattern data coincidence and iris code coincidence closest to a predetermined determination reference area in this two-dimensional space, and the registered blink pattern data and registered iris code related to these coincidences. Is identified. Then, the overall determination unit 150 determines whether or not the subject is a voted user based on the registered blink pattern data and the user ID set in the registered iris code. Identify whether you are a user.

  The above is the comparison / collation processing in the blink pattern data matching unit 136 and the iris code matching unit 146 in the first user identification mode, and the subsequent determination processing in the comprehensive determination unit 150. On the other hand, in the second user identification mode, the user identification device 100 operates as follows.

  As described above, in the second user identification mode, the user identification device 100 acquires a user ID from the subject. The blink pattern data matching unit 136 accesses the blink pattern data registration unit 134 and acquires one registered blink pattern data associated with the user ID of the subject. Similarly, the iris code matching unit 146 accesses the iris code registration unit 144 and acquires one registered iris code associated with the user ID.

  Next, the blink pattern data matching unit 136 compares the test blink pattern data generated by the blink pattern generation unit 132 with one registered blink pattern data acquired from the blink pattern data registration unit 134, and the degree of coincidence thereof. (Blink pattern data matching degree) is transmitted to the overall determination unit 150 as a blink pattern data collation result. Further, the iris code matching unit 146 compares the test iris code generated by the iris coding unit 140 with one registered iris code acquired from the iris code registration unit 144, and the degree of coincidence (the degree of iris code coincidence). ) As an iris code collation result. Then, when the blink pattern data matching degree is equal to or higher than the predetermined level and the iris code matching degree is equal to or higher than the predetermined level, the comprehensive determination unit 150 determines that the subject is a qualified user. Determines that the subject is not a voted user.

  As described above, according to the first embodiment, the user is identified using not only the iris unique to each person but also the blink pattern that only the person can know. Therefore, even if the image quality of the eye periphery image is low, it is possible to obtain higher authentication accuracy than the conventional technique for identifying the user based only on the iris.

  Also, the blink pattern in the present embodiment is an alternative to a password or passphrase. In general, in order to acquire a password or the like, a user's key operation is indispensable, but in order to acquire a subject's blink pattern, the subject only needs to blink toward the camera 902. . Therefore, the user can be accurately identified without impairing convenience for the user.

<Second Embodiment>
FIG. 5 shows the configuration of a user identification device 200 (individual identification device) according to the second embodiment of the present invention.

  The user identification device 200 is different from the user identification device 100 according to the first embodiment in that an angle detection unit 210, an image rotation unit 212 (image inclination correction unit), an eye periphery feature detection unit 214, and an eye periphery A feature data registration unit 216 and an eye peripheral feature data matching unit 218 are added. In FIG. 5, the light emitting circuit 110 and the time control circuit 112 are omitted. Although not shown in FIG. 5, the blink detection unit 130, the blink pattern data generation unit 132, the blink pattern data registration unit 134, which are components of the user identification device 100 according to the first embodiment, and The blink pattern data matching unit 136 may be provided in the user identification device 200 according to the present embodiment.

  The angle detection unit 210 detects a relative difference in the rotation direction between the test iris code and the registered iris code compared in the iris code matching unit 146. Specifically, a shift in the rotation direction of the test iris code with respect to the registered iris code is detected. The code difference detected here corresponds to the difference in rotation angle between the eye peripheral image that is the basis of the test iris code and the eye peripheral image that is the basis of the registered iris code. This difference is caused when the position of the camera 902 is shifted in the rotation angle direction with respect to the eye of the subject. In this case, when the eye periphery image 10 shown in FIG. 4 is to be input to the user identification device 200, the tilted eye periphery image 10a as shown in FIG. 6 is input. Such a phenomenon becomes particularly prominent when the user identification device 200 is incorporated into a small device such as the mobile phone 900. This is because camera shake tends to occur when a small device is held.

  The image rotation unit 212 uses the rotation angle difference detected by the angle detection unit 210 to rotate the eye periphery image input to the image input unit 120 and corrects the tilt. As a result, the tilted eye periphery image 10a as shown in FIG. 6 is converted into a normalized eye periphery image 10b without tilt as shown in FIG.

  The eye periphery feature detection unit 214 receives the normalized eye periphery image 10b from the image rotation unit 212, extracts the position, shape, and other features of the region around the eye (feature part) from the eye periphery image 10b. Each is defined by a descriptor. For example, the inclination of the eye is defined by an angle α formed by a straight line connecting both ends of the eye (the eye 30a and the corner of the eye 30b) and a horizontal line, and the mole 32 around the eye has two-dimensional coordinates x1, The curvature of the eyebrows 22 is defined by the angles β1, β2, β3,..., βn formed by a plurality of line segments virtually drawn at the center of the eyebrows 22. In addition, the shape of the ridge 12 (single fold, double fold, etc.), wrinkles around the eyes, etc. may be adopted as the user identification element.

  When the user identification device 200 is set to the user registration mode, the eye periphery feature detection unit 214 transmits each descriptor to the eye periphery feature data registration unit 216 as eye periphery feature data. The eye peripheral feature data registration unit 216 stores the eye peripheral feature data in association with each user ID. On the other hand, when the user identification device 200 is set to the user identification mode, the eye peripheral feature detection unit 214 transmits each descriptor to the eye peripheral feature data matching unit 218 as the test eye peripheral feature data. The eye peripheral feature data registration unit 216 may be provided outside the mobile phone 900 in the same manner as the blink pattern data registration unit 134 and the iris code registration unit 144 described above. For example, when the mobile phone 900 has a function of accessing the Internet, the eye peripheral feature data registration unit 216 can be configured as a database server on the Internet.

  The eye periphery feature data matching unit 218 extracts predetermined eye periphery feature data (hereinafter referred to as “registered eye periphery feature data”) from the eye periphery feature data registration unit 216 and sends it from the eye periphery feature detection unit 214. Compare and compare with the characteristic data around the subject eye, and calculate the degree of coincidence between them. The data comparison / collation operation of the eye peripheral feature data matching unit 218 is the same as that of the blink pattern data matching unit 136 and the iris code matching unit 146. Then, the eye peripheral feature data matching unit 218 transmits the result of matching the eye peripheral feature data to the comprehensive determination unit 150. The overall determination unit 150 determines whether or not the subject is a qualified user based on the eye periphery feature data matching result received from the eye periphery feature data matching unit 218 and the iris code matching result received from the iris code matching unit 146. Judgment is made, and further, the authorized user is identified.

  As described above, according to the second embodiment, a user is identified using not only the iris unique to each person but also the positions, shapes, and other characteristics of a plurality of parts around the eye. Therefore, even if the image quality of the eye periphery image is low, it is possible to obtain higher authentication accuracy than the conventional technique for identifying the user based only on the iris.

  In the second embodiment, since the inclination of the input image around the eye is corrected, means for fixing the positional relationship between the camera 902 and the eye of the subject (for example, a holder for fixing the mobile phone 900) is provided. Without using it, each subject can be identified accurately and quickly.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, this invention is not limited to the example which concerns. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are of course within the technical scope of the present invention. Understood.

  The present invention is applicable to a personal authentication system that authenticates users of various information devices and various services.

It is a block diagram which shows the structure of the user identification apparatus concerning the 1st Embodiment of this invention. FIG. 2 is an external view of a mobile phone in which the user identification device of FIG. 1 is incorporated. It is explanatory drawing which shows the positional relationship of the mobile telephone of FIG. 2, and a test subject's eyes. It is explanatory drawing of the eye periphery image input into the user identification device of FIG. It is a block diagram which shows the structure of the user identification device concerning the 2nd Embodiment of this invention. It is explanatory drawing of the eye periphery image input into the user identification device of FIG. It is explanatory drawing of the site | part of the eye periphery extracted by the user identification apparatus of FIG.

Explanation of symbols

10 Eye Peripheral Image 12 １４ 14 Pupil 16 Iris 18 Sclera 20 Eyelash 22 Eyebrow 100 User Identification Device 110 Light Emitting Circuit 112 Time Control Circuit 120 Image Input Unit 130 Blink Detection Unit 132 Blink Pattern Data Generation Unit 134 Blink Pattern Data Registration Unit 136 Blink pattern data matching unit 140 Iris coding unit 144 Iris code registration unit 146 Iris code matching unit 150 Total determination unit 200 User identification device 210 Angle detection unit 212 Image rotation unit 214 Eye peripheral feature detection unit 216 Eye peripheral feature data registration unit 218 Eye peripheral feature data matching unit 900 Mobile phone 902 Camera 904 Illumination lamp 906 Timing lamp

Claims (13)

An image input unit for acquiring an image including the eye of the test individual and its peripheral part;
A blink pattern matching unit that extracts a blink pattern from the image and calculates a degree of coincidence between the extracted blink pattern and a previously registered blink pattern;
An iris matching unit that extracts iris features from the image and calculates a degree of coincidence between the extracted iris features and pre-registered iris features;
At least an individual identification unit for identifying the test individual based on the degree of coincidence of the blink pattern calculated by the blink pattern matching unit and the degree of coincidence of the iris feature calculated by the iris matching unit;
An individual identification device characterized by comprising: The blink pattern is defined by the opening / closing time of the test subject's eyelids within a predetermined time, the number of times the test individual's eyelids are opened / closed, or a combination of the test subject's eyelid opening / closing time and the number of times of opening / closing. The individual identification device according to claim 1. The individual identification apparatus according to claim 1, further comprising a blink timing notifying unit that notifies the subject of the blink timing. An image input unit for acquiring an image including the eye of the test individual and its peripheral part;
A feature part matching unit that extracts one or more feature parts from the image and calculates a degree of coincidence between the extracted one or more feature parts and one or more feature parts registered in advance; ,
An iris matching unit that extracts iris features from the image and calculates a degree of coincidence between the extracted iris features and pre-registered iris features;
An individual identifying unit for identifying the test individual based on at least the degree of coincidence of one or more characteristic parts calculated by the characteristic part matching unit and the degree of coincidence of iris features calculated by the iris matching unit;
An individual identification device characterized by comprising: The individual identification device according to claim 4, wherein one of the characteristic parts is one of eyes, moles, and eyebrows. The individual identification apparatus according to claim 1, further comprising an image inclination correction unit that corrects the inclination of the image. The individual identification device according to claim 1, wherein the individual identification device is incorporated in a portable information device. An image input step of acquiring an image including the eye of the subject individual and the surrounding area;
A blink pattern matching step of extracting a blink pattern from the image and calculating a degree of coincidence between the extracted blink pattern and a previously registered blink pattern;
An iris matching step of extracting iris features from the image and calculating a degree of coincidence between the extracted iris features and pre-registered iris features;
At least an individual identification step for identifying the test individual based on the degree of coincidence of the blink pattern calculated in the blink pattern matching step and the degree of coincidence of the iris feature calculated in the iris matching step;
An individual identification method characterized by including: The blink pattern is defined by the opening / closing time of the test subject's eyelids within a predetermined time, the number of times the test individual's eyelids are opened / closed, or a combination of the test subject's eyelid opening / closing time and the number of times of opening / closing The individual identification method according to claim 8. The individual identification method according to claim 8 or 9, wherein the test individual is notified of the timing of blinking. An image input step of acquiring an image including the eye of the subject individual and the surrounding area;
A feature part matching step of extracting one or more feature parts from the image and calculating a degree of coincidence between the extracted one or more feature parts and one or more feature parts registered in advance; ,
An iris matching step for extracting iris features from the image and calculating a degree of coincidence between the extracted iris features and pre-registered iris features;
An individual identifying step for identifying the subject individual based on at least the degree of coincidence of one or more feature portions calculated in the feature portion matching step and the degree of coincidence of iris features calculated in the iris matching step;
An individual identification method characterized by including: The individual identification method according to claim 11, wherein one of the characteristic parts is one of eyes, moles, and eyebrows. The individual identification method according to claim 8, further comprising an image correction step of correcting the inclination of the image.

Images