JP2005063173A - Biological authentication apparatus and passage controller - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a biological authentication apparatus which can be always stably be operated without lowering authentication precision even when the biological information of a person to be newly registered is similar to that of an existing user. <P>SOLUTION: In the biological authentication apparatus using biometrics such as face image or fingerprint image, in the case of newly registering dictionary information, all dictionary information, which are already registered, are collated. When there are the dictionary information similar to the newly registered dictionary information at the predetermined range or more, similar factors are extracted and displayed visually. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention provides, for example, a biometric authentication device that authenticates a person using biometric information such as a person's face image, fingerprint image, iris information, palm image, finger image, and voice information in an entrance / exit management system and the like, and Further, the present invention relates to a traffic control device that controls the pass of a passerby (for example, entry / exit to an important facility) using this biometric authentication device.

  In general, this type of biometric authentication apparatus acquires biometric information of a person (authenticated person) to be authenticated using a camera, a sensor, and the like, and the person-specific feature information obtained from the acquired biometric information is dictionaryd. Information is stored (registered) in advance in the dictionary storage means.

  At the time of collation, the biometric information of the person to be authenticated is acquired again using a camera or a sensor, and the person-specific feature information obtained from the acquired biometric information is stored in the dictionary storage means in advance. The similarity of both information is calculated | required by collating with information, and it is determined by the calculated similarity whether the said passer-by is the person concerned. For example, whether or not the input biometric information is similar to dictionary information is evaluated. If the similarity is high, the person is determined to be a registrant (person), and if the similarity is low, the person is not registered. It is determined that the person is not the person.

  However, in this type of biometric authentication device, depending on the verification method of biometric information, for example, due to factors such as wearing the same glasses, twins, growing the same shape of wrinkles, and the same bangs (hairstyle) However, there may be dictionary information that resembles a certain range or more, so that there is a problem that authentication accuracy is significantly reduced. Conventionally, in order to avoid such existence, it has been handled by complicating the operation of all registrants.

  Therefore, the present invention provides a biometric authentication device and a traffic control device that can always operate stably without degrading the authentication accuracy even when the biometric information of a newly registered person is similar to that of an existing registrant. The purpose is to provide.

  The biometric authentication apparatus according to the present invention includes a dictionary storage unit that stores in advance biometric information of a person to be authenticated as dictionary information, a biometric information acquisition unit that acquires the biometric information of the person to be authenticated as an authentication target, The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and whether the person to be authenticated is the person in question based on the obtained similarity When registering new dictionary information with collation means for determining whether or not, the biometric information acquired by the biometric information acquisition means is collated with all the dictionary information stored in the dictionary storage means, respectively. Extraction means for extracting dictionary information that is more than the range and similar factors, display means for visually displaying the extraction result of the extraction means, and registration based on the display contents of the display means When an instruction is input, and a dictionary registration means allowed to stored in the dictionary storage unit the biometric information of the person to be authenticated acquired by the biometric information obtaining means as the dictionary information.

  Further, the traffic control device of the present invention includes a dictionary storage unit that stores biometric information of a passerby to be authenticated as dictionary information in advance, a biometric information acquisition unit that acquires biometric information of a passerby to be authenticated, The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and whether the passerby is the person based on the obtained similarity Collation means for determining whether or not, a passage control means for controlling the passage of the passer-by based on a result of determination by the collation means, and biological information acquired by the biological information acquisition means when registering new dictionary information And extracting means for extracting dictionary information that is more than a certain range and similar factors thereof by comparing each dictionary information with all dictionary information stored in the dictionary storage means. Display means for visually displaying the extraction result of the extraction means, and when a registration command is input based on the display contents of the display means, the dictionary using the passerby's biometric information acquired by the biometric information acquisition means as dictionary information Dictionary registration means for storing in the storage means.

  According to the present invention, when newly registering dictionary information, it collates with all dictionary information already registered before that, and if there is dictionary information that is more than a certain range, the similarity factor is extracted. By displaying it visually, the optimum operation method can be presented to the new registrant based on the similar factors, so that the authentication accuracy can be improved even when the biometric information of the newly registered person is similar to that of the existing registrant. It is possible to provide a biometric authentication device and a traffic control device that can always be stably operated without being lowered.

Embodiments of the present invention will be described below with reference to the drawings.
In the following description, a case where a face image is used as the biometric information of the person to be authenticated will be described as an example, but in addition to the face image, other images such as a fingerprint image, an iris information, a palm image, a finger image, and voice information The same can be done using biological information.

  FIG. 1 schematically shows a configuration of an entrance / exit management system to which a biometric authentication device and a traffic control device according to the present embodiment are applied. This entrance / exit management system, for example, controls a single door and captures a face image (an image including at least a face) of a person to be authenticated (hereinafter also referred to as a person, passerby, or registrant). An image input unit 101 as a biometric information acquisition unit for input, a display unit 102 as a display unit for displaying a face image input to the person to be authenticated, various kinds of guidance, and the like, a personal ID (identification information) for identifying the person to be authenticated ) And an input unit 103 as an input unit for inputting information such as a password, and a biometric feature extraction unit that processes an image input from the image input unit 101 and acquires input feature information (face feature information) unique to the person. Face feature extraction unit 104, dictionary storage unit 105 as dictionary storage means for storing (registering) face feature information of a person registered in advance as dictionary information, and input from input unit 103 From the dictionary registration unit 106 and the registered face feature extraction unit 104 as dictionary registration means for registering (storing) the input feature information obtained from the feature extraction unit 104 in correspondence with the personal ID obtained in the dictionary storage unit 105 as dictionary information. By collating the obtained input feature information with the dictionary information in the dictionary storage unit 105, the similarity between the input feature information and the dictionary information is obtained, and when the obtained similarity is higher than the threshold, the person is If the similarity is determined to be a registrant (person) and the obtained similarity is lower than the threshold value, the collation unit 107 as a collation unit that determines that the person is not a registered person (not the person) The door control unit 109 is a traffic control means for controlling the opening and closing of the entrance door 108 for important facilities.

  In this example, the image input unit 101 captures and inputs a face image (an image including at least a face) of the person to be authenticated. For example, a television camera using an image sensor such as a CCD sensor. It is composed of

  The display unit 102 is composed of, for example, a liquid crystal display panel, and displays the face image of the person to be authenticated, various guidance, messages, and the like input by the image input unit 101. Similarly, a voice message or the like may be output by connecting a device capable of outputting voice.

  The input unit 103 is for inputting the personal ID and password of the person to be authenticated at the entrance of the visited site, for example. Specifically, for example, a numeric keypad, a touch panel, an IC card, a magnetic card, or a wireless card is used. Can be realized by a card reader or the like.

  The face feature extraction unit 104 detects a person's face in the image input from the image input unit 101 and extracts (acquires) input feature information. As a face detection method, for example, literature (Kazuhiro Fukui, Osamu Yamaguchi: “Face feature point extraction by combination of shape extraction and pattern matching”, IEICE Transactions (D), vol. J80-D-II, No. 8, pp 2170-2177 (1997)) can be used to realize highly accurate face detection. In this case, a location having the highest correlation value is determined as a face region by obtaining a correlation value while moving a template prepared in advance in the image.

  The dictionary storage unit 105 registers biometric information of a person permitted to be used in the entire system in advance as dictionary information. As the dictionary information to be registered, the input face image, extracted feature amount, subspace, eigenvector for constructing the subspace, correlation matrix, status information at the time of registration, date / time, place, etc., password, Personal information such as ID numbers can be listed and stored as necessary. In this embodiment, the face cut-out image extracted by the face feature extraction unit 104 is unique by KL expansion. It is assumed that vector information obtained by converting into space is stored, and this information is handled as dictionary information.

  The dictionary registration unit 106 registers (stores) the input feature information obtained from the feature extraction unit 104 in the dictionary storage unit 105 as dictionary information in association with the personal ID input from the input unit 103.

  The collation unit 107 obtains the similarity between the input feature information and the dictionary information by collating the input feature information obtained from the face feature extraction unit 104 with the dictionary information in the dictionary storage unit 105, and the obtained similarity Is higher than the threshold, the person is determined to be a registrant (person), and when the obtained similarity is lower than the threshold, the person is determined to be an unregistered person (not the person).

  For example, the collation unit 107 outputs a “door open” signal to the door control unit 109 when the obtained similarity is greater than or equal to a threshold, and when the obtained similarity is smaller than the threshold, the collation unit 107 outputs a “door open” signal to the door control unit 109. The door close signal is output. When the door control unit 109 receives the “door open” signal from the verification unit 107, the door control unit 109 controls the entrance door 108 to be in an open state, allows the passer-by to enter, and receives the “door closed” signal. The entrance / exit door 108 is kept closed, and the passerby is refused to enter.

Next, the operation in such a configuration will be described.
First, a first example will be described with reference to the flowchart shown in FIG.
When newly registering dictionary information, when the administrator shifts the system to the administrator mode using the input unit 103 and then selects the search mode (step S1), the search mode is set (step S2).

  Next, the registrant's face image is input by the image input unit 101 (step S3). When the face image is input, the collation unit 107 collates the input feature information extracted by the face feature extraction unit 104 with all the registered dictionary information in the dictionary storage unit 105 (step S4).

  As a result of this collation, if there is dictionary information that is more than a certain range similar to the input feature information, the dictionary information and its similar factors (for example, twins that are wearing the same glasses and having the same shape of pupae are grown. The bangs (hairstyle) are the same) (step S5). At this time, if there is a plurality of dictionary information that is similar to the input feature information by a certain range or more, a predetermined number is extracted from the top (step S6). Here, the processing of steps S5 and S6 corresponds to the extraction means in the present invention.

  Next, the collation unit 107 displays the extraction result (dictionary information similar to a certain range and its similarity factor) on the display unit 102 (step S7). Here, the administrator visually determines the display method of the display unit 102, determines the corresponding method as in the following examples (1) to (4) based on the displayed similar factor as a hint, Guidance is given during operation (step S8).

  (1) Since face decorations such as glasses are similar, if it is determined that the face is more than a certain range as a result of collation, the newly registered registrant will take off the glasses and collate Instruct you to.

  (2) In the case of a family such as a brother, if the entrance and exit of a condominium is managed and there is no problem even if misidentifying people who are similar to a certain range as a result of the comparison, Explain why it is used.

  (3) In the case of another person's similarity, for example, explain that the newly registered registrant will be operated with a higher collation threshold value, and have them understand the operation method.

  (4) In the case of other people's resemblance, for example, explain that the operation of entering the password after verification is applied by raising the verification threshold of the newly registered registrant and having them understand the operation method .

  The above description has been given for a case where new dictionary information is registered. However, when it is found that a person who is prone to collation NG or a manager confirms the traffic history, misrecognition has occurred between registrants. Is also applicable.

Further, an example of determining an operation method for a registrant who newly registers dictionary information based on the result of the collation will be specifically described.
For example, when newly registering the dictionary information of the registrant B, it is assumed that it is found that the dictionary information of the registrant A is similar to a certain range or more after collating with the existing dictionary information. In this case, the registrant B and the registrant A belong to the same affiliation, and it is determined that there is no problem even when erroneous authentication occurs, and the registrant B is operated as it is.

  Further, it is assumed that the registrant C's dictionary information is newly registered, and that it is found that the dictionary information of the registrant B is similar to a certain range or more after collation with the existing dictionary information. In this case, when the registrant C and the registrant B want to discriminate without misrecognition, the collation threshold of the registrant C is increased (stricter).

  Furthermore, if the dictionary information (first candidate) indicating the maximum similarity is the registrant B as a result of the collation, and the second candidate is the registrant C, the collation OK determination result When the second candidate is the registrant A, the determination is made based on the conventional similarity.

  Next, after the administrator selects a registration mode using the input unit 103 and inputs a registration command, the dictionary registration unit 106 stores the biometric information of the registrant (the face image input in step S3) in the dictionary. Information is stored (registered) in the dictionary storage unit 105 (step S9).

Next, a second example will be described with reference to the flowchart shown in FIG.
When newly registering dictionary information, when the administrator shifts the system to the administrator mode using the input unit 103 and then selects the search mode (step S11), the search mode is set (step S12).

  Next, the registrant's face image is input by the image input unit 101 (step S13). When the face image is input, the collation unit 107 collates the input feature information extracted by the face feature extraction unit 104 with all the registered dictionary information in the dictionary storage unit 105 (step S14).

  As a result of this collation, if there is dictionary information that is more than a certain range similar to the input feature information, the dictionary information and its similar factors (for example, twins that are wearing the same glasses and having the same shape of pupae are grown. And the same bangs (hairstyle) are extracted) (step S15). At this time, if there is a plurality of dictionary information that is similar to the input feature information by a certain range or more, a predetermined number is extracted from the top (step S16). Here, the processing of steps S15 and S16 corresponds to the extraction means in the present invention.

  Next, the collation unit 107 displays the extraction result (dictionary information similar to a certain range and its similarity factor) on the display unit 102 (step S17). Note that the process of step S17 is not necessarily performed and may be omitted.

  Next, the collation unit 107 determines a response method (collation method) for a registrant who newly registers dictionary information based on the extraction result (step S18). That is, clarify similar factors (same glasses, twins, the same shape of wrinkles, the same bangs (hairstyle)), and other than that part when collating biometric information of the registrant The features in the range are used mainly. Specifically, if the similarity factor is glasses, when collating the registrant, the range other than the glasses portion is weighted and collated. Alternatively, when the entire face is similar (in the case of another person's sky, twins), when collating the registrant, the collation threshold is set high. Here, the process of step S18 corresponds to the collation method determining means in the present invention.

  Here, the administrator ends the search mode with the input unit 103 (step S19), returns to the administrator mode (step S20), and then selects a new dictionary registration mode (step S21), the search mode is set. (Step S22). Then, the dictionary registration unit 106 stores (registers) the biometric information of the registrant (the face image input in step S13) as dictionary information in the dictionary storage unit 105 using the collation method determined in step S18 (step S18). S23).

Next, a third example will be described with reference to the flowchart shown in FIG.
In the normal collation mode, when collation is started (step S31), that is, when the person to be authenticated is positioned in front of the image input unit 101, the image input unit 101 inputs the face image of the person to be authenticated (step S31). S32). When the face image is input, the collation unit 107 collates the input feature information extracted by the face feature extraction unit 104 with all the dictionary information registered in the dictionary storage unit 105 (step S33).

  As a result of this collation, if there is dictionary information that is more than a certain range similar to the input feature information, the dictionary information and its similar factors (for example, twins that are wearing the same glasses and having the same shape of pupae are grown. And the same bangs (hairstyle) are extracted) (step S34). At this time, if there is a plurality of dictionary information that is similar to the input feature information by a certain range or more, a predetermined number is extracted from the higher rank. Here, step S34 processing corresponds to the extracting means in the present invention.

  Next, the collation unit 107 displays the extraction result (dictionary information similar to a certain range and its similarity factor) on the display unit 102 (step S35). The process of step S35 is not necessarily performed and may be omitted.

  Next, the matching unit 107 determines and learns a matching method for the person to be authenticated based on the extraction result (step S36). In other words, when identifying similar factors (same glasses, twins, the same shape of wrinkles, the same bangs (hairstyle)), and verifying the biometric information of that person Learn to use mainly the features in the range other than. Specifically, if the similarity factor is glasses, when collating the person to be authenticated, a range other than the glasses portion is weighted and collated. Alternatively, when the entire face is similar (in the case of another person's sky, twins), when collating the person to be authenticated, collation is performed with a higher collation threshold. Here, the process of step S36 corresponds to the collation method learning means in the present invention.

  As a result of the collation in the collation unit 107, the similarity between the input feature information and the dictionary information is higher than the threshold value, and it is determined that the person to be authenticated is a registrant (person himself / herself) (decision as collation OK) ( In step S37), the collation unit 107 instructs the door control unit 109 to open the door (step S38), and returns to step S31 to repeat the above operation.

  In addition, when the obtained similarity is lower than the threshold and it is determined that the person is not a registered person (not the person) (determined to be collation NG) (step S37), the collation unit 107 notifies the door control unit 109 that the door is closed. Is instructed (step S39). Here, the processing of steps S38 and S39 corresponds to the traffic control means in the present invention.

  If it is determined as collation NG, the person to be authenticated eliminates the similar factor displayed in step S35 (step S40), and then starts collation again. For example, if the glasses were a similar factor, it was displayed after removing the glasses and rematching, or if the hairstyle was a similar factor, changing the hairstyle (raising the bangs) and rematching was displayed. It can be easily verified by devising according to the similarity factor.

  In the above-described embodiment, the case where the present invention is applied to an entrance / exit management system for controlling one door has been described. For example, when a plurality of doors are controlled, the door may be configured as shown in FIG. It is the same as that of an Example. According to FIG. 5, this entrance / exit management system is installed corresponding to one registration unit 100 for registering biometric information (dictionary information) and a plurality of doors, and a plurality of authentication units 200 and 300 for performing biometric authentication. ,..., And a communication line 400 such as a LAN that connects these to each other in a communicable manner.

  The registration unit 100 includes an image input unit 101, an input unit 103, a face feature extraction unit 104, a dictionary storage unit 105, and a dictionary registration unit 106, which are the same as those in the above-described embodiment (FIG. 1).

  The authentication unit 200 includes an image input unit 101a, a display unit 102a, a face feature extraction unit 104a, a collation unit 107a, and a door control unit 109a, which are the same as those in the above embodiment (FIG. 1). The exit door 108a is controlled.

  The authentication unit 300 includes an image input unit 101b, a display unit 102b, a face feature extraction unit 104b, a collation unit 107b, and a door control unit 109b, which are the same as those in the above embodiment (FIG. 1). The exit door 108b is controlled.

  The dictionary information in the dictionary storage unit 105 in the registration unit 100 is transmitted to each of the verification units 107a and 107b of the authentication units 200 and 300 via the communication line 400, and is provided for verification processing. .

1 is a block diagram schematically showing a configuration of an entrance / exit management system to which a biometric authentication device and a traffic control device according to an embodiment of the present invention are applied. The flowchart explaining the 1st example of operation | movement. The flowchart explaining the 2nd example of operation | movement. The flowchart explaining the 3rd example of operation | movement. The block diagram which shows schematically the structure of the entrance / exit management system with which the biometrics apparatus and traffic control apparatus which concern on the other Example of this invention are applied.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 101 ... Image input part (biological information acquisition means), 102 ... Display part (display means), 103 ... Input part (input means), 104 ... Facial feature extraction part (biological feature extraction means), 105 ... Dictionary storage part (dictionary) Storage means), 106 ... dictionary registration section (dictionary registration means), 107 ... collation section (collation means), 108 ... door for entry / exit, 109 ... door control section (traffic control means).

Claims (8)

Dictionary storage means for storing the biometric information of the person to be authenticated as dictionary information in advance,
Biometric information acquisition means for acquiring biometric information of the person to be authenticated who is to be authenticated;
The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and the person to be authenticated is the person based on the calculated similarity. Checking means for determining whether or not
When registering new dictionary information, the biometric information acquired by the biometric information acquisition means is respectively compared with all dictionary information stored in the dictionary storage means, so that dictionary information that is similar to a certain range or more Extraction means for extracting the similar factors;
Display means for visually displaying the extraction result of the extraction means;
When a registration command is input based on the display content of the display means, a dictionary registration means for storing the biometric information of the person to be authenticated acquired by the biometric information acquisition means in the dictionary storage means as dictionary information;
A biometric authentication device comprising: Dictionary storage means for storing the biometric information of the person to be authenticated as dictionary information in advance,
Biometric information acquisition means for acquiring biometric information of the person to be authenticated who is to be authenticated;
The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and the person to be authenticated is the person based on the calculated similarity. Checking means for determining whether or not
When registering new dictionary information, the biometric information acquired by the biometric information acquisition means is respectively compared with all dictionary information stored in the dictionary storage means, so that dictionary information that is similar to a certain range or more Extraction means for extracting the similar factors;
Based on the extraction result of the extracting means, a matching method determining means for determining a matching method for a person to be authenticated who newly registers dictionary information;
A dictionary registration unit that stores the biometric information of the person to be authenticated acquired by the biometric information acquisition unit in the dictionary storage unit as dictionary information by the verification method determined by the verification method determination unit;
A biometric authentication device comprising: Dictionary storage means for storing the biometric information of the person to be authenticated as dictionary information in advance,
Biometric information acquisition means for acquiring biometric information of the person to be authenticated who is to be authenticated;
The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and the person to be authenticated is the person based on the calculated similarity. Checking means for determining whether or not
When there is dictionary information that is similar to the biometric information acquired by the biometric information acquisition unit at a certain range or more at the time of collation by the collation means, extraction that extracts dictionary information that is similar to the specific range or more and its similarity factor Means,
Based on the extraction result of the extracting means, a matching method learning means for determining and learning a matching method for the person to be authenticated;
A biometric authentication device comprising:   4. The biometric authentication apparatus according to claim 3, further comprising display means for visually displaying the extraction result of the extraction means. Dictionary storage means for storing biometric information of a passerby to be authenticated in advance as dictionary information;
Biometric information acquisition means for acquiring biometric information of a passerby to be authenticated;
The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and whether the passerby is the person by the calculated similarity Verification means for determining whether or not,
A passage control means for controlling the passage of the passer based on the result of the determination by the matching means;
When registering new dictionary information, the biometric information acquired by the biometric information acquisition means is respectively compared with all dictionary information stored in the dictionary storage means, so that dictionary information that is similar to a certain range or more Extraction means for extracting the similar factors;
Display means for visually displaying the extraction result of the extraction means;
When a registration command is input based on the display content of the display means, dictionary registration means for storing the passerby's biometric information acquired by the biometric information acquisition means as dictionary information in the dictionary storage means;
A traffic control device comprising: Dictionary storage means for storing biometric information of a passerby to be authenticated in advance as dictionary information;
Biometric information acquisition means for acquiring biometric information of a passerby to be authenticated;
The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and whether the passerby is the person by the calculated similarity Verification means for determining whether or not,
A passage control means for controlling the passage of the passer based on the result of the determination by the matching means;
When registering new dictionary information, the biometric information acquired by the biometric information acquisition means is respectively compared with all dictionary information stored in the dictionary storage means, so that dictionary information that is similar to a certain range or more Extraction means for extracting the similar factors;
Based on the extraction result of the extracting means, a matching method determining means for determining a matching method for a passerby who newly registers dictionary information;
A dictionary registration unit that stores the biometric information of the passerby acquired by the biometric information acquisition unit in the dictionary storage unit as dictionary information by the verification method determined by the verification method determination unit;
A traffic control device comprising: Dictionary storage means for storing biometric information of a passerby to be authenticated in advance as dictionary information;
Biometric information acquisition means for acquiring biometric information of a passerby to be authenticated;
The biometric information acquired by the biometric information acquisition means and the dictionary information stored in the dictionary storage means are collated to determine the similarity between the two information, and whether the passerby is the person by the calculated similarity Verification means for determining whether or not,
When there is dictionary information that is similar to the biometric information acquired by the biometric information acquisition unit at a certain range or more at the time of collation by the collation means, extraction that extracts dictionary information that is similar to the specific range or more and its similarity factor Means,
Based on the extraction result of the extracting means, a matching method learning means for determining and learning a matching method for the passer-by,
A passage control means for controlling the passage of the passer based on the result of the determination by the collating means;
A traffic control device comprising:   8. The traffic control apparatus according to claim 7, further comprising display means for visually displaying an extraction result of the extraction means.

Images