Title of the invention: Information processing apparatus, information processing method and storage medium
Technical field
[0001]
The present invention relates to an information processing apparatus, an information processing method and a storage medium.
Background technology
[0002]
Patent Document 1 discloses an information processing device that determines a user by face authentication. The information processing apparatus captures a user's face image and calculates a score obtained by quantifying the degree of matching between the captured face image and the registrant's face image. Then, when there is a registrant whose score is equal to or higher than the threshold value, the information processing apparatus determines that the photographed user is the registrant.
Prior art literature
Patent documents
[0003]
Patent Document 1: Japanese Patent Application Laid-Open No. 2016-218672
Outline of the invention
Problems to be solved by the invention
[0004]
In biometric authentication as described in Patent Document 1, it is required to maintain the collation accuracy so as to sufficiently reduce the occurrence rate of erroneous determination. However, since the state of the biometric authentication system changes depending on the usage status after introduction, it may be difficult to maintain the collation accuracy. Therefore, there is a demand for a method for appropriately managing the collation accuracy.
[0005]
An object of the present invention is to provide an information processing apparatus, an information processing method, and a storage medium capable of appropriately managing collation accuracy.
Means to solve the problem
[0006]
According to one aspect of the present invention, it is obtained by collating a first biometric information obtained from an image including at least a part of a collation target person with a plurality of second biometric information corresponding to a plurality of registrants. In addition, an acquisition unit that acquires a collation score group including a plurality of collation scores indicating the degree of agreement between the first biometric information and the second biometric information, and a ranking among the collation score groups is 1. Provided is an information processing apparatus including a determination unit that makes a determination based on a collation score that is a rank, a collation score that is not the first rank, and an output unit that outputs management information based on the result of the determination. To.
[0007]
According to another aspect of the present invention, by collating the first biometric information obtained from the image including at least a part of the collation target person with the plurality of second biometric information corresponding to the plurality of registrants. The step of acquiring a collation score group including a plurality of collation scores each indicating the degree of agreement between the obtained first biometric information and the second biometric information, and the ranking in the collation score group are An information processing method including a step of performing a determination based on a collation score which is the first place, a collation score whose rank is not the first place, and a step of outputting management information based on the result of the determination is provided. ..
[0008]
According to another aspect of the present invention, the computer collates the first biometric information obtained from the image including at least a part of the collation target person with the plurality of second biometric information corresponding to the plurality of registrants. A step of acquiring a collation score group including a plurality of collation scores each indicating the degree of agreement between the first biometric information and the second biometric information obtained by the above step, and in the collation score group. An information processing method including a step of performing a determination based on a collation score in which the ranking is 1st, a collation score in which the ranking is not 1st, and a step of outputting management information based on the result of the determination. A storage medium is provided in which a program for execution is stored.
Effect of the invention
[0009]
According to the present invention, an information processing apparatus, an information processing method and a storage medium capable of appropriately managing the collation accuracy are provided.
A brief description of the drawing
[0010]
FIG. 1 is a block diagram showing a hardware configuration of a biometric authentication system according to the first embodiment.
FIG. 2 is a functional block diagram of the information processing apparatus according to the first embodiment.
FIG. 3 is a flowchart showing an outline of processing performed in the information processing apparatus according to the first embodiment.
FIG. 4 is a flowchart showing a collation process performed in the information processing apparatus according to the first embodiment.
FIG. 5 is a table showing an example of collation results acquired in the information processing apparatus according to the first embodiment.
FIG. 6 is a flowchart showing a collation state monitoring process performed in the information processing apparatus according to the first embodiment.
FIG. 7 is a table showing an example of the distribution of collation scores acquired in the information processing apparatus according to the first embodiment.
[Fig. 8] This is an example of a graph showing the relationship between the distribution of collation scores and misjudgment.
[Fig. 9] This is an example of a graph showing the relationship between fluctuations in the distribution of collation scores and the probability of occurrence of false positives.
[Fig. 10] Fig. 10 is a table showing an example of the relationship between the cause of deterioration of collation accuracy and the tendency of change in collation score.
FIG. 11 is a flowchart showing a collation state monitoring process performed in the information processing apparatus according to the second embodiment.
FIG. 12 is a functional block diagram of the information processing apparatus according to the third embodiment.
Embodiment for carrying out the invention
[0011]
Hereinafter, an exemplary embodiment of the present invention will be described with reference to the drawings. Similar elements or corresponding elements may be designated by the same reference numerals in the drawings, and the description thereof may be omitted or simplified.
[0012]
[First embodiment]
The biometric authentication system according to this embodiment will be described. The biometric authentication system of the present embodiment is a system for identifying a person by biometric authentication. The biometric authentication system of the present embodiment is used, for example, for entry / exit management in controlled areas, immigration control at airports, borders, etc., login management at terminals, servers, etc., identity verification in electronic payment, etc. could be. However, the biometric authentication system of the present embodiment may be any as long as it identifies a person, and its use is not particularly limited.
[0013]
The biometric authentication method that can be applied to the biometric authentication system of this embodiment can typically be face authentication. In face authentication, the feature amount obtained from the face image of the collation target person is collated with the feature amount obtained from the face images of a plurality of registrants, and the collation target person is the same person as one of the registrants. It can be used to determine whether or not it is. However, the biometric authentication method may be another method that uses an image of at least a part of the collation target person, for example, an image such as a fingerprint, a palm print, a vein, or an iris. In the following description, it is assumed that the biometric authentication system of the present embodiment is a face authentication system provided at a place where a large number of verification target persons sequentially visit, such as an entrance / exit gate.
[0014]
In the present specification, the image of the feature portion such as the face image of the collation target person or the feature amount obtained from the image may be more generally referred to as the first biometric information. Further, an image of a feature portion such as a registrant's face image or a feature amount obtained from the feature portion may be more generally referred to as second biometric information.
[0015]
FIG. 1 is a block diagram showing a hardware configuration of a biometric authentication system according to this embodiment. The biometric authentication system includes an information processing device 1, an image pickup device 2, and an administrator terminal 3. The information processing device 1, the image pickup device 2, and the administrator terminal 3 are connected in a communicable manner via a network by wire or wirelessly. The information processing apparatus 1 is a computer that performs collation processing and collation status monitoring processing in a biometric authentication system. The image pickup device 2 is a digital camera that captures a face image of a collation target person. The administrator terminal 3 is a computer used by an administrator who manages the biometric authentication system.
[0016]
The overall configuration of this biometric authentication system is an example, and for example, the information processing device 1 and the image pickup device 2 may be configured as an integrated device. In addition, the biometric authentication system is another such as an entrance / exit gate that opens and closes based on the authentication result, a display device that notifies the verification target person of the authentication result, and a card reader that reads the identification card held by the verification target person. The device may be further included.
[0017]
The information processing device 1 includes a CPU (Central Processing Unit) 101, a RAM (Random Access Memory) 102, a ROM (Read Only Memory) 103, and an HDD (Hard Disk Drive) 104. Further, the information processing device 1 includes a communication I / F (Interface) 105, an input device 106, and an output device 107. Each part of the information processing apparatus 1 is connected to each other via a bus, wiring, a driving device, etc. (not shown).
[0018]
In FIG. 1, each part constituting the information processing device 1 is shown as an integrated device, but some of these functions may be provided by an external device. For example, the input device 106 and the output device 107 may be external devices different from the parts constituting the functions of the computer including the CPU 101 and the like.
[0019]
The CPU 101 is a processor that performs predetermined calculations according to programs stored in the ROM 103, HDD 104, etc., and also has a function of controlling each part of the information processing apparatus 1. The RAM 102 is composed of a volatile storage medium and provides a temporary memory area necessary for the operation of the CPU 101. The ROM 103 is composed of a non-volatile storage medium and stores necessary information such as a program used for the operation of the information processing apparatus 1. The HDD 104 is a storage device composed of a non-volatile storage medium, which temporarily stores data transmitted to and received from the image pickup device 2, stores an operation program of the information processing device 1, and the like.
[0020]
Communication I / F105 is a communication interface based on standards such as Ethernet (registered trademark) and Wi-Fi (registered trademark). The communication I / F 105 is a module for communicating with other devices such as the image pickup device 2 and the administrator terminal 3.
[0021]
The input device 106 is a keyboard, a pointing device, or the like, and is used for the user to operate the information processing device 1. Examples of pointing devices include mice, trackballs, touch panels, pen tablets and the like.
[0022]
The output device 107 is, for example, a display device. The display device is a liquid crystal display, an OLED (Organic Light Emitting Diode) display, etc., and is used for displaying information, a GUI (Graphical User Interface) for operation input, and the like. The input device 106 and the output device 107 may be integrally formed as a touch panel.
[0023]
Note that the hardware configuration of the information processing device 1 is an example, and devices other than these may be added, and some devices may not be provided. Further, some devices may be replaced with other devices having similar functions. Further, a part of the functions of the present embodiment may be provided by another device via a network, or the functions of the present embodiment may be distributed and realized by a plurality of devices. For example, the HDD 104 may be replaced with an SSD (Solid State Drive) using a semiconductor memory. Further, the HDD 104 may be replaced with a cloud storage. In this way, the hardware configuration of the information processing apparatus 1 can be changed as appropriate.
[0024]
FIG. 2 is a functional block diagram of the information processing apparatus 1 according to the present embodiment. The information processing apparatus 1 includes an image pickup control unit 111, a collation unit 112, a score acquisition unit 113, a determination unit 114, an output unit 115, and a storage unit 116. The score acquisition unit 113 may be more generally referred to as an acquisition unit.
[0025]
The CPU 101 performs a predetermined arithmetic process by loading the program stored in the ROM 103, the HDD 104, etc. into the RAM 102 and executing the program. Further, the CPU 101 controls each part of the information processing apparatus 1 such as the communication I / F 105 and the image pickup apparatus 2 based on the program. As a result, the CPU 101 realizes the functions of the image pickup control unit 111, the collation unit 112, the score acquisition unit 113, the determination unit 114, the output unit 115, and the storage unit 116. The specific contents of the processing performed by each functional block will be described later.
[0026]
Note that some or all of the functions of the functional blocks described in the information processing device 1 in FIG. 2 may be provided in an external device of the information processing device 1. That is, each of the above-mentioned functions may be realized by the information processing apparatus 1 or may be realized by cooperation with an external device.
[0027]
FIG. 3 shows information related to this embodiment.It is a flowchart which shows the outline of the processing performed in the processing apparatus 1. In this example, the biometric authentication system performs collation every time a collation target person visits, and monitors the collation state at predetermined time intervals. The process of FIG. 3 is repeatedly executed after the biometric authentication system starts operation until the operation ends.
[0028]
In step S10, the information processing apparatus 1 determines whether or not a collation target person exists in the imaging range of the imaging apparatus 2 of the biometric authentication system. When it is determined that the collation target person exists (YES in step S10), the process proceeds to step S20. When it is determined that the collation target person does not exist (NO in step S10), the process proceeds to step S30. It should be noted that this determination can be made based on, for example, an image pickup device 2 acquiring an image in the vicinity and whether or not a person's face is present in the acquired image.
[0029]
In step S20, the information processing apparatus 1 acquires a face image of the collation target person and performs a collation process for collating the feature amount obtained from the face image with the feature amount of the registrant.
[0030]
In step S30, the information processing apparatus 1 determines whether or not a predetermined time has elapsed. If it is determined that the predetermined time has elapsed (YES in step S30), the process proceeds to step S40. If it is determined that the predetermined time has not elapsed (NO in step S30), the process proceeds to step S10. The start time of the predetermined time used for this determination may be, for example, the operation start time of the biometric authentication system or the time when the collation state monitoring process described later is last performed.
[0031]
In step S40, the information processing apparatus 1 performs the collation status monitoring process based on the result of the past collation process. The information processing apparatus 1 transmits management information indicating the collation status to the administrator terminal 3 based on the result of the collation status monitoring process.
[0032]
FIG. 4 is a flowchart showing a collation process performed in the information processing apparatus 1 according to the present embodiment. FIG. 4 shows the collation process of step S20 of FIG. 3 in more detail.
[0033]
In step S201, the image pickup control unit 111 controls the image pickup device 2 to capture and acquire an image including the collation target person. The acquired image is stored in the storage unit 116. In the present embodiment, since the features of the face of the collation target person are extracted as the feature amount, it is sufficient that the image includes at least the face of the collation target person. In the following, this image may be referred to as a face image. When performing an authentication method using a portion other than the face, the image may include at least a portion used for extracting the feature amount.
[0034]
In step S202, the collation unit 112 extracts the feature amount of the collation target person from the feature points such as eyes and nose included in the face image acquired by the image pickup device 2.
[0035]
In step S203, the collation unit 112 collates the feature amount of the collation target person with the feature amount of each of the plurality of registrants stored in advance in the storage unit 116. Here, the plurality of registrants are persons who should be determined in the biometric authentication system of the present embodiment whether or not they are the same person as the collation target person. For example, when the biometric authentication system of the present embodiment is used for determining whether or not the entrance / exit gate can pass, the plurality of registrants may be persons who are permitted to pass through the entrance / exit gate.
[0036]
In this collation, the collation unit 112 calculates a collation score indicating the degree of matching of the feature amounts used in the collation for each of the plurality of registrants. The calculated collation score is stored in the storage unit 116 in association with the registrant. The definition of the collation score value can be appropriately set according to the algorithm used for collation. In the following description, it is assumed that the collation score value is a number between 0 and 1, and the larger the value, the higher the degree of agreement between the feature quantities. A set of collation scores obtained by collating each of a plurality of registrants may be referred to as a collation score group.
[0037]
In step S204, the collation unit 112 calculates the rank of each collation score in the collation score group. It is assumed that the higher the collation score value, that is, the higher the degree of matching between the feature quantities, the higher the ranking. The calculated order is stored in the storage unit 116 in association with the registrant.
[0038]
In step S205, the collation unit 112 determines whether or not there is a collation score registrant in the collation score group whose rank is first and the collation score value is equal to or higher than a predetermined threshold. If such a collation score exists (YES in step S205), the process proceeds to step S206. If such a collation score does not exist (NO in step S205), processing proceeds to step S207. Since there is always a collation score with the first rank in the collation score group, the above-mentioned determination condition may be rephrased as "whether or not the collation score with the first rank is equal to or higher than the threshold value". ..
[0039]
In step S206, the collation unit 112 determines that the collation target person is the same person as the registrant corresponding to the collation score having the first rank. That is, the collation unit 112 determines that the collation target person is one of a plurality of registrants.
[0040]
In step S207, the collation unit 112 determines that the collation target person is not the same person among the plurality of registrants. That is, the collation unit 112 determines that the collation target person is not a registrant.
[0041]
FIG. 5 is a table showing an example of collation results acquired in the information processing apparatus 1 according to the present embodiment and stored in the storage unit 116. The data shown in this table are virtual data created to explain the processing of the present embodiment.
[0042]
In FIG. 5, five "registrants" and "face image data", "feature amount", "collation date and time", "collation score", "rank" and "judgment" stored in association with each registrant are shown. "Results" are shown in tabular form. FIG. 5 shows the collation results of three sets executed at the three collation times (10:21, 10:23, 10:26) on the same day. In the first collation (10:21), the collation score of the registrant AAA is the largest among the five collation score groups and is the first place, so that the registrant AAA is the same person as the collation target person. It has been judged. Since the four registrants other than the registrant AAA cannot be the same person as the collation target person, it is determined that they are other people regardless of the collation score. In this example, it is assumed that the first-ranked collation score exceeds the threshold in all collations.
[0043]
Similarly, in the second collation (10:23), it is determined that the registrant CCC with the first ranking is the same person as the collation target person, and in the third collation (10:26), the ranking is It is determined that the first-ranked registrant DDD is the same person as the collation target person. That is, the example shown in FIG. 5 shows that the collation target at 10:21 is AAA, the collation target at 10:23 is CCC, and the collation target at 10:26 is DDD. ing.
[0044]
FIG. 6 is a flowchart showing a collation state monitoring process performed in the information processing apparatus 1 according to the present embodiment. FIG. 6 shows in more detail the collation state monitoring process in step S40 of FIG.
[0045]
In step S401, the score acquisition unit 113 reads out the collation score group obtained by the past collation and the rank of each collation score from the storage unit 116 and acquires them.
[0046]
In step S402, the determination unit 114 extracts a plurality of collation scores with the first rank and generates a distribution of the collation scores with the first rank. In step S403, the determination unit 114 extracts a plurality of collation scores with the second rank and generates a distribution of the collation scores with the second rank. Hereinafter, a distribution including a plurality of collation scores in the first rank may be referred to as a first distribution, and a distribution including a plurality of collation scores in the second rank may be referred to as a second distribution.
[0047]
In step S404, the determination unit 114 analyzes the collation state in the collation unit 112 based on the first distribution and the second distribution, and determines the deterioration of the collation accuracy. The details of this process will be described later. If it is determined that the collation accuracy has deteriorated (YES in step S405), the process proceeds to step S406. If it is not determined that the collation accuracy has deteriorated (NO in step S405), step S406 is skipped and the collation status monitoring process ends.
[0048]
In step S406, the output unit 115 transmits the management information to the administrator terminal 3. As a result, the administrator can acquire the monitoring result of the collation status in a timely manner. The management information typically includes information indicating that the matching accuracy of the biomatching system has deteriorated. As a specific example of this transmission, a message warning that the collation accuracy of the bio-verification system has deteriorated is sent by e-mail to the administrator's destination so that the administrator can view the management information on the administrator terminal 3. It can be something to do. By transmitting the management information in the form of an e-mail or the like that can be viewed immediately by the administrator, the administrator can quickly take measures against the biometric authentication system.
[0049]
In the process of step S406, the management information transmitted to the administrator terminal 3 may include identification information (registrant number, registrant name, etc.) indicating the registrant who is ranked first. .. As a result, the administrator can take appropriate measures while confirming the validity of the collation with the actually acquired face image or the like. This identification information may be any information that can be associated with the registrant, such as the registrant's number, the registrant's name, the registrant's face image file name, and the registrant's feature amount file name.
[0050]
FIG. 7 is a table showing an example of the first distribution and the second distribution of the collation score acquired in the information processing apparatus 1 according to the present embodiment. The data shown in this table show the data included in the first distribution and the second distribution obtained when the collation score is tentatively extracted from the data of FIG. By extracting only the collation score with the first rank from the large number of collation scores shown in FIG. 5, three collation scores of 0.91, 0.88, and 0.79 are obtained. Similarly, if only the collation score with the second highest rank is extracted, three collation scores of 0.44, 0.36, and 0.49 can be obtained. The table in FIG. 7 shows these arranged in the order of the collation date and time.
[0051]
Comparing the first distribution of the first-ranked collation score and the second distribution of the second-ranked collation score, the value of the first distribution is larger as a whole. In addition, there are variations in the values ​​of each distribution. Here, the first distribution is the distribution of the collation score when it is determined that the collation target person and the registrant are the same person, and the second distribution is the case where the collation target person and the registrant are determined to be other people. It is the distribution of the collation score of. From these facts, it is possible to analyze the collation state by collecting a large amount of data of the first distribution and the second distribution and examining the tendency of the center or the spread of the collation score in the form of a histogram or the like.
[0052]
The method for analyzing the collation state from the first distribution and the second distribution in the determination unit 114 will be described in more detail. FIG. 8 is an example of a graph showing the relationship between the distribution of matching scores in biometric authentication and erroneous determination. In FIG. 8, the “first distribution (1st place)” is the collation when the collation score is the 1st place, that is, it is determined that the collation target person and the registrant are the same person (the collation target person is the registrant himself / herself). This is the distribution obtained when a large number of scores are obtained. "Second distribution (2nd place)" is the collation score when the collation score is 2nd, that is, it is determined that the collation target person and the registrant are not the same person (the collation target person and the registrant are other people). This is the distribution obtained when a large number is acquired.
[0053]
The "threshold" in FIG. 8 is the threshold of the collation score for determining whether the person is the person or another person in the collation. If the collation score is equal to or higher than the threshold, it is determined to be the person himself, and if the collation score is smaller than the threshold, it is determined to be another person.
[0054]
There may be overlapping parts in the two distributions. Therefore, set the threshold appropriately Even if it is set, an erroneous judgment will occur at least for the overlapping part. There are two modes for erroneous judgment. There is a misjudgment that the part of the first distribution where the collation score is smaller than the threshold, that is, the part shown as "personal refusal" in FIG. 8, is not the same person even though he / she is actually the person. It is the range that occurs. In the second distribution, the part where the collation score is equal to or higher than the threshold value, that is, the part shown as "acceptance of another person" in FIG. 8, is erroneously determined to be the same person even though it is actually another person. It is a range.
[0055]
The area of ​​the part of "rejection of person" and "acceptance of others" corresponds to the probability of occurrence of these erroneous judgments. Since the areas of "rejection of the person" and "acceptance of others" can be changed by changing the setting of the threshold value, it is possible to adjust the probability of occurrence of erroneous determination to some extent. However, when there is an overlap in the distributions of "1st distribution (1st position)" and "2nd distribution (2nd position)", the probability of occurrence of two types of erroneous judgments is a trade-off, and both are simultaneous. It cannot be zero. The susceptibility to erroneous determination depends on the size of the overlapping area of ​​the two distributions. The size of the overlapping area of ​​the two distributions is determined by the center of the collation score of the two distributions, the size of the spread, and the like.
[0056]
FIGS. 9 (a) and 9 (b) are examples of graphs showing the relationship between fluctuations in the distribution of collation scores and the probability of occurrence of erroneous determination. FIG. 9A is an example of the distribution of the collation score in the initial state immediately after the start of operation of the biometric authentication system. In the initial state, the biometric authentication system is preliminarily adjusted so that the overlapping area of ​​the first distribution and the second distribution is reduced in order to sufficiently reduce the occurrence probability of erroneous determination. FIG. 9B is an example of the distribution of the collation score after the distributions of the first distribution and the second distribution have changed after the biometric authentication system is operated. As the biometric authentication system operates, the probability of occurrence of erroneous determination may change due to various factors such as changes in the number of registrants registered in the database and changes in the usage environment.
[0057]
For example, as shown in FIG. 9B, when the collation score of the first distribution becomes smaller and approaches the second distribution, the overlapping area of ​​the two distributions becomes larger, and erroneous judgments of "rejection of the person" and "acceptance of others" are made. The probability of occurrence increases. For this reason, in the operation of the biometric authentication system, it is desirable to monitor changes in the probability of occurrence of erroneous determination due to changes in the state. Since the probability of occurrence of erroneous determination depends on the state of the first distribution and the second distribution, it is effective to acquire the first distribution and the second distribution in performing this monitoring.
[0058]
The analysis of the collation state and the determination of the deterioration of the collation accuracy in step S404 will be described in more detail. The collation state analysis method may be any method as long as the collation state can be grasped from the first distribution and the second distribution. For example, when the first distribution and the second distribution are close to each other, an erroneous judgment is likely to occur. Therefore, the degree of proximity between the first distribution and the second distribution is quantified, and the first distribution and the second distribution are higher than a predetermined threshold. A method of determining that the collation accuracy has deteriorated when the two distributions are close to each other can be adopted. To quantify the degree of proximity, for example, a method that statistically considers the entire distribution such as the mean value, median value, and dispersion of the first distribution and the second distribution may be used, but the first distribution may be used. And the overlapping area of ​​the second distribution may be used. Since the overlapping area is directly linked to the probability of occurrence of erroneous judgment, it is effective in extracting the matching accuracy.
[0059]
Further, the method of analyzing the collation state and determining the deterioration of the collation accuracy is to determine whether or not the collation accuracy is deteriorated based on the time change of at least one of the first distribution and the second distribution. May be good. When the collation accuracy deteriorates, the feature may appear more prominently in the time change than in the absolute value of the first distribution and the second distribution. Therefore, it may be better to determine the deterioration of the collation accuracy based on the time change. As a specific example of the determination criterion, whether or not the amount of change in the distribution between the initial state and the time when the collation state is monitored may be equal to or greater than a predetermined threshold value may be used as the determination criterion. Further, the direction in which the distribution has changed from the initial state may be included in the determination criteria. Regarding the method of setting the judgment criteria, the relationship between the deterioration factor of the collation accuracy and the change in the distribution will be described in more detail.
[0060]
FIG. 10 is a table showing the relationship between the cause of deterioration of collation accuracy and the tendency of change in collation score. With reference to the example shown in FIG. 10, the tendency of the change of the first distribution and the second distribution for each cause of deterioration of the collation accuracy will be described.
[0061]
If the number of registrants stored in the storage unit 116 increases by additionally registering registrants, the probability that a registrant with a high collation score will appear other than the person himself / herself increases. Therefore, in this case, the collation score of the second distribution increases. By using an algorithm that determines that the collation accuracy has deteriorated when the second distribution changes by a predetermined threshold or more in the direction of rising from the initial state, it is possible to determine the deterioration of the collation accuracy due to this deterioration cause. can.
[0062]
In addition, if the distribution of the attributes (age, gender, etc.) of the registrant changes, a high matching score will be obtained even if the collation target person and the registrant are the same person under the collation conditions corresponding to the initially assumed attributes. Is difficult to obtain. Therefore, in this case, the collation score of the first distribution is lowered. For this determination, an algorithm may be used in which it is determined that the collation accuracy is deteriorated when the first distribution changes by a predetermined threshold value or more in the direction of decreasing from the initial state. As a specific example of the change in the attribute distribution, there is a case where the biometric matching system was initially intended for young people, but many elderly people were registered during actual operation.
[0063]
In addition, many low-quality facial images may be registered due to changes in the registration method of facial images. In this case, even if the collation target person and the registrant are the same person, it is difficult to obtain a high collation score. Further, in a low-quality facial image, a high collation score may be obtained accidentally even when the collation target person and the registrant are not the same person. Therefore, in this case, the collation score of the first distribution decreases and the collation score of the second distribution increases. In this determination, the collation accuracy deteriorates when the first distribution changes by a predetermined threshold or more in the direction of decreasing from the initial state and the second distribution changes by a predetermined threshold or more in the direction of increasing from the initial state. An algorithm can be used to determine that. Examples of the low-quality face image include a face image taken by a low-quality image pickup device such as a camera mounted on a mobile phone, a face image in which the face is slanted, and the like. As an example of accidentally obtaining a high collation score, there is a case where both the registered face and the face of the collation target person are slanted.
[0064]
Also, if the face of the registrant changes over time after a long time has passed since registration, it becomes difficult to obtain a high matching score even if the collation target person and the registrant are the same person. Therefore, in this case, the collation score of the first distribution is lowered. For this determination, an algorithm may be used in which it is determined that the collation accuracy is deteriorated when the first distribution changes by a predetermined threshold value or more in the direction of decreasing from the initial state.
[0065]
Further, when the lighting environment of the place where the image pickup device 2 is installed changes, the brightness of the face image deviates from the expected brightness, so that it becomes difficult to obtain a high matching score in both the person matching and the other person matching. .. Therefore, in this case, the matching scores of the first distribution and the second distribution are lowered. For this determination, an algorithm may be used in which it is determined that the collation accuracy has deteriorated when both the first distribution and the second distribution change from the initial state to a predetermined threshold value or more.
[0066]
As described above, the first distribution and the second distribution change in various modes depending on the factors that deteriorate the collation accuracy. Therefore, it is possible to appropriately determine the deterioration of the collation accuracy by estimating the deterioration factors that may occur depending on the specifications of the biological collation system, the installation environment, etc., and then using the above-mentioned algorithms alone or in combination as appropriate.
[0067]
In the example described in the present embodiment, the collation score included in the second distribution is only the one having the second rank, but if it is other than the first rank, the one having the third rank or lower may be extracted. .. Since the collation score of the 3rd place or lower can be treated as the collation score when collating another person in the same way as the 2nd place, more data can be obtained by including not only the 2nd place but also the 3rd place or less in the 2nd distribution. Can be extracted.
[0068]
As described above, the information processing apparatus 1 of the present embodiment determines the deterioration of the collation accuracy based on the collation score in which the rank is the first in the collation score group and the collation score in which the rank is not the first. can do. Then, the information processing apparatus 1 can notify the administrator of the collation status by outputting the management information according to the determination result. As a result, the administrator can grasp the collation state in a timely manner, and when the collation accuracy is deteriorated, the administrator can promptly deal with the biometric authentication system. Further, since the information processing apparatus 1 of the present embodiment can monitor the collation state while operating the biological collation system as usual, it is not necessary to separately inspect the collation accuracy. As described above, according to the present embodiment, the information processing apparatus 1 capable of appropriately managing the collation state is provided.
[0069]
[Second embodiment]
In the biometric authentication system of this embodiment, the content of the collation status monitoring process is different from that of the first embodiment, but the other parts are the same as those of the first embodiment. Hereinafter, the differences from the first embodiment will be mainly described, and the description of the common parts will be omitted or simplified.
[0070]
FIG. 11 is a flowchart showing a collation state monitoring process performed in the information processing apparatus 1 according to the present embodiment. FIG. 11 shows in more detail the collation state monitoring process in step S40 of FIG. Since the processes from step S401 to step S405 are the same as those in the first embodiment, the description thereof will be omitted.
[0071]
In step S407, the determination unit 114 determines the deterioration factor of the collation accuracy based on the first distribution and the second distribution. In order to determine this deterioration factor, for example, the correlation between the deterioration cause and the change in distribution as shown in FIG. 10 is investigated in advance, and the deterioration factor is determined from the tendency of the change in the first distribution and the second distribution. Can be used to determine. From this survey result, it is possible to determine the deterioration factor of the collation accuracy by constructing a rule-based determination algorithm that outputs the deterioration factor by inputting the change direction and the change amount of the first distribution and the second distribution. Further, the deterioration factor of the collation accuracy may be determined by using the trained model generated by the machine learning using the survey result as the teacher data. If it is difficult to narrow down the deterioration factors to one, a plurality of deterioration factor candidates may be determined and presented.
[0072]
In step S406, the output unit 115 transmits the management information to the administrator terminal 3. The management information includes information on the deterioration of the collation accuracy of the biological collation system and the deterioration factor of the collation accuracy. A specific example of this transmission method may be to send a warning message indicating that the matching accuracy of the biomatching system has deteriorated and a message indicating a candidate for a probable deterioration factor to the administrator's destination by e-mail. .. As a result, the administrator can quickly take measures for the biometric authentication system while utilizing the determined deterioration factor.
[0073]
According to the present embodiment, in addition to the effects described in the first embodiment, the deterioration factor can be acquired by the determination unit 114, so that the collation accuracy can be managed more appropriately. Device 1 is provided.
[0074]
The system described in the above-described embodiment can also be configured as in the following third embodiment.
[0075]
[Third embodiment]
FIG. 12 is a functional block diagram of the information processing apparatus 4 according to the third embodiment. The information processing apparatus 4 includes an acquisition unit 401, a determination unit 402, and an output unit 403. The acquisition unit 401 is the first obtained by collating the first biometric information obtained from the image including at least a part of the collation target person with the plurality of second biometric information corresponding to the plurality of registrants. Multiple that each indicates the degree of agreement between the biometric information and the second biometric information Get the collation score group including the collation score of. The determination unit 402 makes a determination based on the collation score in which the rank is the first in the collation score group and the collation score in which the rank is not the first. The output unit 403 outputs management information based on the result of the determination.
[0076]
According to the present embodiment, the information processing apparatus 4 capable of appropriately managing the collation accuracy is provided.
[0077]
[Modified embodiment]
The present invention is not limited to the above-described embodiment, and can be appropriately modified without departing from the spirit of the present invention. For example, an example in which a partial configuration of any one embodiment is added to another embodiment or an example in which a partial configuration of another embodiment is replaced with another embodiment is also an embodiment of the present invention.
[0078]
A processing method in which a program for operating the configuration of the embodiment is recorded in a storage medium so as to realize the functions of the above-described embodiment, the program recorded in the storage medium is read out as a code, and the program is executed in a computer is also described in each embodiment. Included in the category. That is, a computer-readable storage medium is also included in the scope of each embodiment. Further, not only the storage medium in which the above-mentioned program is recorded but also the program itself is included in each embodiment. Further, one or more components included in the above-described embodiment are circuits such as ASIC (Application Specific Integrated Circuit) and FPGA (Field Programmable Gate Array) configured to realize the functions of each component. There may be.
[0079]
As the storage medium, for example, a floppy (registered trademark) disk, a hard disk, an optical disk, a photomagnetic disk, a CD (Compact Disk) -ROM, a magnetic tape, a non-volatile memory card, or a ROM can be used. In addition, the program recorded on the storage medium is not limited to the one that executes the processing by itself, but the one that operates on the OS (Operating System) and executes the processing in cooperation with other software and the function of the expansion board. Is also included in the category of each embodiment.
[0080] [0080]
The service realized by the functions of each of the above-described embodiments can also be provided to the user in the form of SaaS (Software as a Service).
[0081]
It should be noted that the above-mentioned embodiments are merely examples of embodiment in carrying out the present invention, and the technical scope of the present invention should not be construed in a limited manner by these. That is, the present invention can be implemented in various forms without departing from its technical idea or its main features.
[0082]
A part or all of the above-described embodiment may be described as in the following appendix, but is not limited to the following.
[0083]
(Appendix 1)
The first biometric information obtained by collating a first biometric information obtained from an image including at least a part of a collation target person with a plurality of second biometric information corresponding to a plurality of registrants, and the above-mentioned first biometric information. An acquisition unit that acquires a collation score group including a plurality of collation scores, each of which indicates the degree of agreement with the second biometric information.
A determination unit that makes a determination based on a collation score that ranks first in the collation score group and a collation score that does not rank first.
Based on the result of the above judgment, the output unit that outputs management information and
Information processing device equipped with.
[0084]
(Appendix 2)
The determination unit makes the determination based on the first distribution including a plurality of collation scores in which the rank is first and the second distribution including a plurality of collation scores in which the rank is not first.
The information processing device described in Appendix 1.
[0085]
(Appendix 3)
The determination unit makes the determination based on the time change of at least one of the first distribution and the second distribution.
The information processing device described in Appendix 2.
[0086]
(Appendix 4)
The output unit outputs the management information when it is determined that the first distribution has changed by a predetermined threshold value or more in a direction in which the first distribution decreases from the initial state.
The information processing device described in Appendix 3.
[0087]
(Appendix 5)
The output unit outputs the management information when it is determined that the second distribution has changed by a predetermined threshold or more in the direction of rising from the initial state.
The information processing device described in Appendix 3.
[0088]
(Appendix 6)
When it is determined that the output unit changes by a predetermined threshold or more in the direction in which the first distribution decreases from the initial state, and changes by a predetermined threshold or more in the direction in which the second distribution rises from the initial state. Output the management information,
The information processing device described in Appendix 3.
[0089]
(Appendix 7)
The output unit outputs the management information when it is determined that both the first distribution and the second distribution have changed by a predetermined threshold value or more in a direction of decreasing from the initial state.
The information processing device described in Appendix 3.
[0090]
(Appendix 8)
The determination unit makes the determination based on the degree of proximity between the first distribution and the second distribution.
The information processing device described in Appendix 2.
[0091]
(Appendix 9)
The determination unit makes the determination based on the overlapping area between the first distribution and the second distribution.
The information processing device described in Appendix 8.
[0092]
(Appendix 10)
The management information includes identification information indicating a registrant whose rank is first.
The information processing device according to any one of Supplementary note 1 to 9.
[0093]
(Appendix 11)
The management information includes information indicating that the collation accuracy has deteriorated.
The information processing device according to any one of Supplementary note 1 to 10.
[0094]
(Appendix 12)
The management information includes information on factors that deteriorate the collation accuracy.
The information processing device according to Appendix 11.
[0095]
(Appendix 13)
The determination unit further determines the deterioration factor based on the collation score in which the rank is the first and the collation score in which the rank is not the first.
The information processing device described in Appendix 12.
[0096]
(Appendix 14)
The output unit outputs the management information by sending a message warning the administrator terminal that the collation accuracy has deteriorated.
The information processing device according to any one of Supplementary note 1 to 13.
[0097]
(Appendix 15)
The first biometric information is a feature amount extracted from the face image of the collation target person or the face image of the collation target person.
The second biometric information is a feature amount extracted from the registrant's face image or the registrant's face image.
The information processing device according to any one of Supplementary note 1 to 14.
[0098]
(Appendix 16)
The first biometric information obtained by collating a first biometric information obtained from an image including at least a part of a collation target person with a plurality of second biometric information corresponding to a plurality of registrants, and the above-mentioned first biometric information. A step of acquiring a collation score group including a plurality of collation scores each indicating the degree of agreement with the second biometric information, and
A step of making a judgment based on a collation score in which the ranking is the first in the collation score group and a collation score in which the ranking is not the first.
Based on the result of the above judgment, the step to output management information and
Information processing method with.
[0099]
(Appendix 17)
On the computer
The first biometric information obtained by collating a first biometric information obtained from an image including at least a part of a collation target person with a plurality of second biometric information corresponding to a plurality of registrants, and the above-mentioned first biometric information. A step of acquiring a collation score group including a plurality of collation scores each indicating the degree of agreement with the second biometric information, and
A step of making a judgment based on a collation score in which the ranking is the first in the collation score group and a collation score in which the ranking is not the first.
Based on the result of the above judgment, the step to output management information and
A storage medium in which a program for executing an information processing method is stored.
Description of the sign
[0100]
1, 4, information processing device
2 Imaging device
3 Administrator terminal
101 CPU
102 RAM
103 ROM
104 HDD
105 Communication I / F
106 Input device
107 Output device
111 Imaging control unit
112, collation department
113 Score acquisition department
114, 402 Judgment unit
115, 403 output section
116 Storage unit
401 Acquisition department
The scope of the claims
[Claim 1]
The first biometric information obtained by collating a first biometric information obtained from an image including at least a part of a collation target person with a plurality of second biometric information corresponding to a plurality of registrants, and the above-mentioned first biometric information. An acquisition unit that acquires a collation score group including a plurality of collation scores, each of which indicates the degree of agreement with the second biometric information.
A determination unit that makes a determination based on a collation score that ranks first in the collation score group and a collation score that does not rank first.
Based on the result of the above judgment, the output unit that outputs management information and
Information processing device equipped with.
[Claim 2]
The determination unit makes the determination based on the first distribution including a plurality of collation scores in which the rank is first and the second distribution including a plurality of collation scores in which the rank is not first.
The information processing device according to claim 1.
[Claim 3]
The determination unit makes the determination based on the time change of at least one of the first distribution and the second distribution.
The information processing device according to claim 2.
[Claim 4]
The output unit outputs the management information when it is determined that the first distribution has changed by a predetermined threshold value or more in a direction in which the first distribution decreases from the initial state.
The information processing device according to claim 3.
[Claim 5]
The output unit outputs the management information when it is determined that the second distribution has changed by a predetermined threshold or more in the direction of rising from the initial state.
The information processing device according to claim 3.
[Claim 6]
When it is determined that the output unit changes by a predetermined threshold or more in the direction in which the first distribution decreases from the initial state, and changes by a predetermined threshold or more in the direction in which the second distribution rises from the initial state. Output the management information,
The information processing device according to claim 3.
[Claim 7]
The output unit outputs the management information when it is determined that both the first distribution and the second distribution have changed by a predetermined threshold value or more in a direction of decreasing from the initial state.
The information processing device according to claim 3.
[Claim 8]
The determination unit makes the determination based on the degree of proximity between the first distribution and the second distribution.
The information processing device according to claim 2.
[Claim 9]
The determination unit makes the determination based on the overlapping area between the first distribution and the second distribution.
The information processing device according to claim 8.
[Claim 10]
The management information includes identification information indicating a registrant whose rank is first.
The information processing device according to any one of claims 1 to 9.
[Claim 11]
The management information includes information indicating that the collation accuracy has deteriorated.
The information processing device according to any one of claims 1 to 10.
[Claim 12]
The management information includes information on factors that deteriorate the collation accuracy.
The information processing device according to claim 11.
[Claim 13]
The determination unit further determines the deterioration factor based on the collation score in which the rank is the first and the collation score in which the rank is not the first.
The information processing device according to claim 12.
[Claim 14]
The output unit outputs the management information by sending a message warning the administrator terminal that the collation accuracy has deteriorated.
The information processing device according to any one of claims 1 to 13.
[Claim 15]
The first biometric information is a feature amount extracted from the face image of the collation target person or the face image of the collation target person.
The second biometric information is a feature amount extracted from the registrant's face image or the registrant's face image.
The information processing device according to any one of claims 1 to 14.
[Claim 16]
The first biometric information obtained by collating a first biometric information obtained from an image including at least a part of a collation target person with a plurality of second biometric information corresponding to a plurality of registrants, and the above-mentioned first biometric information. Second living body A step to obtain a collation score group containing multiple collation scores, each of which indicates the degree of agreement with the information, and
A step of making a judgment based on a collation score in which the ranking is the first in the collation score group and a collation score in which the ranking is not the first.
Based on the result of the above judgment, the step to output management information and
Information processing method with.
[Claim 17]
On the computer
The first biometric information obtained by collating the first biometric information obtained from an image including at least a part of the collation target person with a plurality of second biometric information corresponding to a plurality of registrants, and the above. A step of acquiring a collation score group including a plurality of collation scores each indicating the degree of agreement with the second biometric information, and
A step of making a judgment based on a collation score in which the ranking is the first in the collation score group and a collation score in which the ranking is not the first.
Based on the result of the above judgment, the step to output management information and
A storage medium in which a program for executing an information processing method is stored.