WO2022070349A1 - 閾値算出システム、閾値算出方法、及びコンピュータプログラム - Google Patents
閾値算出システム、閾値算出方法、及びコンピュータプログラム Download PDFInfo
- Publication number
- WO2022070349A1 WO2022070349A1 PCT/JP2020/037286 JP2020037286W WO2022070349A1 WO 2022070349 A1 WO2022070349 A1 WO 2022070349A1 JP 2020037286 W JP2020037286 W JP 2020037286W WO 2022070349 A1 WO2022070349 A1 WO 2022070349A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- threshold value
- collation
- information
- value calculation
- attribute information
- Prior art date
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 202
- 238000004590 computer program Methods 0.000 title claims description 19
- 238000000605 extraction Methods 0.000 claims abstract description 29
- 238000000034 method Methods 0.000 claims description 27
- 230000008569 process Effects 0.000 claims description 24
- 230000007613 environmental effect Effects 0.000 claims description 14
- 239000000284 extract Substances 0.000 claims description 10
- 238000005070 sampling Methods 0.000 description 15
- 230000008859 change Effects 0.000 description 13
- 238000010586 diagram Methods 0.000 description 10
- 230000000694 effects Effects 0.000 description 8
- 230000004048 modification Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
- G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/1912—Selecting the most significant subset of features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/50—Maintenance of biometric data or enrolment thereof
Definitions
- This disclosure relates to a threshold calculation system for calculating a threshold for biometric authentication, a threshold calculation method, and a technical field of a computer program.
- Patent Document 1 discloses that a threshold value for each registered data is generated based on a collation degree distribution for each registered data.
- Patent Document 2 discloses that a threshold value is directly determined or statistical processing is performed based on the similarity distribution between individuals and the similarity distribution between individuals.
- Patent Document 3 discloses that a threshold value is calculated using the mean value and standard deviation of the distribution of similarity.
- One aspect of the threshold calculation system of this disclosure is a first acquisition means for acquiring collation information used for collation of a living body, a second acquisition means for acquiring attribute information indicating the attributes of the living body or the collation information, and the above-mentioned.
- a storage means for storing the collation information and the attribute information for each living body, a sample extraction means for extracting a plurality of the collation information as sample data from the storage means based on a predetermined condition regarding the attribute information, and the sample.
- a population estimation means for estimating a population from data and a threshold calculation means for calculating a threshold for the collation information based on the estimated population distribution are provided.
- One aspect of the threshold calculation method of this disclosure is to acquire collation information used for collation of a living body, acquire attribute information indicating the attribute of the living body or the collation information, and obtain the collation information and the attribute information for each living body.
- a plurality of the matching information is extracted as sample data from the storage means based on a predetermined condition regarding the attribute information, a population is estimated from the sample data, and the estimated population is estimated.
- the threshold value for the collation information is calculated based on the distribution of.
- One aspect of the computer program of this disclosure is to acquire collation information used for collation of a living body, acquire attribute information indicating the attribute of the living body or the collation information, and obtain the collation information and the attribute information for each living body.
- a plurality of the matching information is extracted as sample data from the storage means based on a predetermined condition regarding the attribute information, a population is estimated from the sample data, and the estimated population is used.
- the computer is operated to calculate the threshold value for the collation information based on the distribution of.
- FIG. 1 is a block diagram showing a hardware configuration of the threshold value calculation system according to the first embodiment.
- the threshold calculation system 10 includes a processor 11, a RAM (Random Access Memory) 12, a ROM (Read Only Memory) 13, and a storage device 14.
- the threshold value calculation system 10 may further include an input device 15 and an output device 16.
- the processor 11, the RAM 12, the ROM 13, the storage device 14, the input device 15, and the output device 16 are connected via the data bus 17.
- Processor 11 reads a computer program.
- the processor 11 is configured to read a computer program stored in at least one of the RAM 12, the ROM 13, and the storage device 14.
- the processor 11 may read a computer program stored in a computer-readable recording medium by using a recording medium reading device (not shown).
- the processor 11 may acquire (that is, may read) a computer program from a device (not shown) located outside the threshold calculation system 10 via a network interface.
- the processor 11 controls the RAM 12, the storage device 14, the input device 15, and the output device 16 by executing the read computer program.
- a functional block for calculating a threshold value related to biometric authentication is realized in the processor 11.
- processor 11 a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an FPGA (field-programmable gate array), a DSP (Demand-Side Platform), and an ASIC (Application) are used. Alternatively, a plurality of them may be used in parallel.
- CPU Central Processing Unit
- GPU Graphics Processing Unit
- FPGA field-programmable gate array
- DSP Demand-Side Platform
- ASIC Application Specific integrated circuit
- the RAM 12 temporarily stores the computer program executed by the processor 11.
- the RAM 12 temporarily stores data temporarily used by the processor 11 while the processor 11 is executing a computer program.
- the RAM 12 may be, for example, a D-RAM (Dynamic RAM).
- the ROM 13 stores a computer program executed by the processor 11.
- the ROM 13 may also store fixed data.
- the ROM 13 may be, for example, a P-ROM (Programmable ROM).
- the storage device 14 stores data stored for a long period of time by the threshold value calculation system 10.
- the storage device 14 may operate as a temporary storage device of the processor 11.
- the storage device 14 may include, for example, at least one of a hard disk device, a magneto-optical disk device, an SSD (Solid State Drive), and a disk array device.
- the input device 15 is a device that receives an input instruction from the user of the threshold value calculation system 10.
- the input device 15 may include, for example, at least one of a keyboard, a mouse and a touch panel.
- the output device 16 is a device that outputs information about the threshold value calculation system 10 to the outside.
- the output device 16 may be a display device (for example, a display) capable of displaying information about the threshold value calculation system 10.
- FIG. 2 is a block diagram showing a functional configuration of the threshold value calculation system according to the first embodiment.
- the threshold value calculation system 10 includes a collation information acquisition unit 110, an attribute information acquisition unit 120, and a personal information storage unit 130 as processing blocks for realizing the function.
- a sampling unit 140, a population estimation unit 150, and a threshold value calculation unit 160 are provided.
- Each of the collation information acquisition unit 110, the attribute information acquisition unit 120, the sampling unit 140, the population estimation unit 150, and the threshold value calculation unit 160 may be realized by the processor 11 (see FIG. 1) described above.
- the personal information storage unit 130 may be realized by the storage device 14 (see FIG. 1) described above.
- the collation information acquisition unit 110 is configured to be able to acquire collation information used for the authentication operation of the living body (specifically, the collation operation with the registered data).
- the collation information acquisition unit 110 may directly acquire the collation information, or may calculate the collation information using the acquired information. Specific examples of the collation information will be described in other embodiments described later.
- the collation information acquired by the collation information acquisition unit 110 is output to the personal information storage unit 130.
- the attribute information acquisition unit 120 is configured to be able to acquire attribute information indicating the attributes of the living body or the collation information.
- the attribute information acquisition unit 110 may directly acquire the attribute information, or may determine and acquire the attribute from the acquired information. Specific examples of attribute information will be described in other embodiments described later.
- the collation information acquired by the attribute information acquisition unit 120 is output to the personal information storage unit 130.
- the personal information storage unit 130 is configured to be able to store the collation information acquired by the collation information acquisition unit 110 and the attribute information acquired by the attribute information acquisition unit 120.
- the personal information storage unit 130 is configured to be able to store a plurality of collation information and attribute information for each living body (for example, in FIG. 2, the collation information and attribute information of individual A, individual B, and individual X are each storable. An example that is stored separately is shown).
- the collation information and attribute information stored in the personal information storage unit 130 can be appropriately read out by the sampling unit 140. Further, the personal information storage unit 130 may have a function of partially (for example, for each living body) deleting the stored collation information and attribute information.
- the sample extraction unit 140 is configured to be able to extract a part or all of the collation information stored in the personal information storage unit 130 as sample data for estimating the population.
- the sampling unit 140 is configured to be able to extract sample data based on a predetermined condition regarding attribute information (hereinafter, appropriately referred to as “population condition”).
- the population condition is a condition set based on, for example, a population assumed to be an authentication target.
- the parameters set as the population condition are, for example, the confidence coefficient (1- ⁇ ⁇ ) for calculating the confidence interval of the population mean, or the confidence coefficient (1- ⁇ ⁇ ) for calculating the confidence interval of the population. It may be there. More specific examples of population conditions will be described in embodiments described below.
- the sample data extracted by the sampling unit 140 is output to the population estimation unit 150.
- the population estimation unit 150 is configured so that the population can be estimated using the sample data extracted by the sampling unit 140.
- the population here includes an unknown stranger who is not stored in the personal information storage unit 130, and is calculated assuming that the threshold value for the collation information can be calculated.
- Information about the population estimated by the population estimation unit 150 is output to the threshold value calculation unit 160.
- the population estimation unit 150 may calculate, for example, an interval upper limit value according to a confidence coefficient (1- ⁇ ⁇ ). Specifically, the population estimation unit 150 sets the number of elements nE of the sample data, and calculates the sample mean x AVE , the unbiased variance U 2 , and the universal standard deviation U S from the sample data. Subsequently, the population estimation unit 150 calculates a confidence interval of the population mean ⁇ from the calculated sample mean x AVE , unbiased variance U 2 , number of elements n E , and confidence coefficient (1- ⁇ ⁇ ). At this time, if the number of elements n E is sufficiently large, the population mean ⁇ may be the sample mean x AVE . Then, the population estimation unit 150 calculates the section upper limit value according to the confidence coefficient (1- ⁇ ⁇ ) by using the population mean ⁇ and the unbiased standard deviation US, which is an unbiased estimator of the population standard deviation ⁇ .
- the threshold value calculation unit 160 is configured to be able to calculate a threshold value used for biometric authentication using collation information based on the distribution of the population estimated by the population estimation unit 150. For example, when the population estimation unit 150 calculates the section upper limit value according to the confidence coefficient (1- ⁇ ⁇ ), the threshold value calculation unit 160 sets the value equal to or higher than the calculated section upper limit value as the threshold value. You may do it.
- the threshold value calculation unit 160 may store the calculated threshold value in the personal information storage unit 130. When the threshold value is already stored in the personal information storage unit 130, the threshold value calculation unit 160 may rewrite (that is, update) with a new threshold value. Further, the threshold value calculation unit 160 may have a function of notifying the system administrator that the threshold value has been calculated (updated).
- FIG. 3 is a flowchart showing the flow of the registration operation in the threshold value calculation system according to the first embodiment.
- the collation information acquisition unit 110 first acquires the collation information of the living body (step S11). Then, the collation information acquisition unit 110 stores the acquired collation information in the personal information storage unit 130 (step S12).
- the attribute information acquisition unit 120 acquires attribute information indicating the attributes of the living body or the collation information (step S13). Then, the attribute information acquisition unit 120 stores the acquired attribute information in the personal information storage unit 130 (step S14). The collation information and the attribute information are stored in a state of being associated with each living body.
- the personal information storage unit 130 stores the collation information and the attribute information of a plurality of living bodies in the living body unit.
- FIG. 4 is a flowchart showing the flow of the threshold value calculation operation in the threshold value calculation system according to the first embodiment.
- the sample extraction unit 140 first collects the collation information stored in the personal information storage unit based on the population condition as sample data. (Step S101). Subsequently, the population estimation unit 150 estimates the population using the extracted sample data (step S102).
- the threshold value calculation unit 160 calculates the threshold value from the estimated population distribution (step S103).
- the threshold value calculation unit 160 typically calculates a threshold value for each living body.
- the threshold value calculation unit 160 may create a cluster in a specific attribute unit, calculate the maximum value or the average value of the threshold value of the living body belonging to the cluster, and set the threshold value in the cluster unit. Further, the threshold value calculation unit 160 may set one threshold value for the entire personal information storage unit 130.
- the threshold value calculation system 10 determines whether or not the threshold value has been calculated for all the registrants (that is, all the living bodies stored in the personal information storage unit) (step S104). Then, when it is determined that the threshold value has not been calculated for all the registrants (step S104: NO), the threshold value calculation system 10 repeats the process from step S101. On the other hand, when it is determined that the threshold value has been calculated for all the registrants (step S104: YES), the threshold value calculation system 10 ends a series of processes.
- FIG. 5 is a graph showing the relationship between the threshold value and the false acceptance rate and the false rejection rate.
- FIG. 6 is a graph showing the difference in the ideal threshold value due to the difference in the sample distribution.
- the smaller the threshold value used for biometric authentication the higher the FAR (tolerance rate for others).
- the larger the threshold value used for biometric authentication the higher the FRR (false rejection rate). Therefore, it is important to set an appropriate threshold value for biometric authentication.
- the threshold value calculated by the distribution of the sample data is different.
- the threshold value X is calculated as an ideal threshold value.
- the threshold value Y is calculated from the sample data shown in FIG. 6 (b).
- the threshold value Z is calculated from the sample data shown in FIG. 6 (c). Therefore, if the sample data used for calculating the threshold value is not properly extracted, the calculated threshold value may become an inappropriate value.
- sample data for calculating the threshold value is extracted based on the population condition regarding the attribute information. That is, appropriate sample data is extracted in consideration of attribute information. With appropriate sample data, the population can be estimated appropriately. Properly estimating the population means that the distribution of the population including unknown strangers that are not remembered can also be properly estimated. As a result, an appropriate threshold value assuming an unknown stranger can be calculated from the estimated population distribution.
- the threshold value calculation system 10 according to the second embodiment will be described with reference to FIGS. 7 and 8.
- the second embodiment is different from the first embodiment described above in a part of the configuration and operation.
- the hardware configuration may be the same as that of the first embodiment (see FIG. 1). .. Therefore, in the following, the description of the portion overlapping with the first embodiment will be omitted as appropriate.
- FIG. 7 is a block diagram showing a functional configuration of the threshold value calculation system according to the second embodiment.
- the same reference numerals are given to the same components as those shown in FIG. 2.
- the threshold value calculation system 10 has an image acquisition unit 50, a feature amount extraction unit 111, a collation score calculation unit 112, an attribute information acquisition unit 120, and a personal information storage unit. It includes 130, a sampling unit 140, a population estimation unit 150, and a threshold value calculation unit 160. That is, the threshold value calculation system 10 according to the second embodiment is configured to further include an image acquisition unit 50 in addition to the configuration of the first embodiment (see FIG. 2). Further, the threshold value calculation system 10 according to the second embodiment is configured to include a feature amount extraction unit 111 and a collation score calculation unit 112 in place of the collation information acquisition unit 110 according to the first embodiment. Each of the image acquisition unit 50, the feature amount extraction unit 111, and the collation score calculation unit 112 may be realized by the processor 11 (see FIG. 1) described above.
- the image acquisition unit 50 is configured to be able to acquire an image including a living body from, for example, a camera or the like.
- the image acquisition unit 50 acquires, for example, a face image, an iris image, a fingerprint image, or the like as an image including a living body.
- the image data acquired by the image acquisition unit 50 is output to the feature amount extraction unit 111 and the attribute information acquisition unit 120.
- the feature amount extraction unit 111 is configured to be able to extract the feature amount of the living body from the image acquired by the image acquisition unit 50. As for the method for extracting the feature amount, the existing technique can be appropriately adopted, and therefore detailed description thereof is omitted here.
- the feature amount extracted by the feature amount extraction unit 111 is stored in the personal information storage unit 130.
- the attribute information acquisition unit 120 determines and acquires an attribute from the image acquired by the image acquisition unit 50.
- the attribute information acquisition unit 120 may acquire the attribute information from other than the image.
- the attribute information acquisition unit 120 may acquire qualitative and quantitative information explicitly input as personal data of a living body as attribute information.
- the collation score calculation unit 112 is configured to be able to calculate a collation score using the feature amount stored in the personal information storage unit 130 (in other words, the feature amount extracted by the feature amount extraction unit 111).
- the collation score here is a score indicating the degree of similarity (or degree of agreement) between the newly registered feature amount of the living body and the already registered feature amount of the living body, and is one of the newly registered living bodies. It is calculated by comparing the feature amount of a person with the feature amount of n people already registered.
- the collation score calculated by the collation score calculation unit 112 is stored in the personal information storage unit 130 for each living body. That is, the collation score is stored in the personal information storage unit 130 in a state of being associated with the feature amount and attribute information that have already been stored.
- FIG. 8 is a flowchart showing the flow of the registration operation in the threshold value calculation system according to the second embodiment.
- the same reference numerals are given to the same processes as those shown in FIG.
- the image acquisition unit 50 first acquires an image including a living body (step S21).
- the image acquisition unit 50 may execute various processes (for example, image processing for facilitating acquisition of feature amounts and attribute information) on the image data.
- the feature amount extraction unit 111 extracts the feature amount of the living body from the image data (step S22). Then, the feature amount extraction unit 111 stores the extracted feature amount in the personal information storage unit 130 (step S23). After that, the collation score calculation unit 112 calculates the collation score from the feature amount (step S24). Then, the collation score calculation unit 112 stores the calculated collation score in the personal information storage unit 130 (step S25).
- the attribute information acquisition unit 120 acquires attribute information indicating the attributes of the living body or the collation information (step S13). Then, the attribute information acquisition unit 130 stores the acquired attribute information in the personal information storage unit 130 (step S14).
- the collation score is stored in the personal information storage unit 130 as collation information.
- the collation score may be stored as a set with the attribute information of the living body used for calculating the collation score.
- the collation score calculated by comparison with Mr. A may be stored as a set with the attribute information of Mr. A.
- N is a natural number
- N sets that is, a set of collation score and attribute information
- the personal information storage unit 130 may store the feature amount as the collation information. In this case, the collation score may not be stored in the personal information storage unit 130.
- the threshold value calculation operation according to the second embodiment is the same flow as the threshold value calculation operation (see FIG. 4) according to the first embodiment. Therefore, a new figure will be omitted, and the description will be given with reference to FIG. 4 as appropriate.
- the sampling unit 140 samples the collation score stored in the personal information storage unit 130 based on the population condition while referring to the corresponding attribute information. Extract as data (step S101). If the collation score is not stored in the personal information storage unit 130 (for example, when the feature amount is stored as collation information), the feature score may be calculated at this stage. Subsequently, the population estimation unit 150 estimates the population using the extracted sample data (step S102). Subsequently, the threshold value calculation unit 160 calculates the threshold value from the estimated population distribution (step S103). The threshold value calculation unit 160 calculates a threshold value for each living body.
- the threshold value calculation system 10 determines whether or not the threshold value has been calculated for all the registrants (that is, all the living bodies stored in the personal information storage unit) (step S104). Then, when it is determined that the threshold value has not been calculated for all the registrants (step S104: NO), the threshold value calculation system 10 repeats the process from step S101. On the other hand, when it is determined that the threshold value has been calculated for all the registrants (step S104: YES), the threshold value calculation system 10 ends a series of processes.
- a feature amount is extracted from an image of a living body, and a collation score is calculated using the feature amount. Therefore, it is possible to easily acquire and store the verification information used for biometric authentication. Further, since the attribute information can be determined and acquired from the image of the living body, it is possible to appropriately acquire the attribute information even when the attribute information is not directly input.
- FIG. 9 is a block diagram showing a functional configuration of a modified example of the threshold value calculation system according to the second embodiment.
- the same reference numerals are given to the same components as those shown in FIG. 7.
- the threshold value calculation system 10 includes an image acquisition unit 50, a feature amount extraction unit 111, a collation score calculation unit 112, an attribute information acquisition unit 120, and an individual. It includes an information storage unit 130, a sampling unit 140, a population estimation unit 150, a threshold value calculation unit 160, and a collation determination unit 170. That is, the threshold value calculation system 10 according to the second embodiment is configured to further include a collation determination unit 170 in addition to the configuration of the second embodiment (see FIG. 7). The collation determination unit 170 may be realized by the processor 11 (see FIG. 1) described above.
- the personal information storage unit 130 is configured to be able to store the threshold value calculated by the threshold value calculation unit 160 for each living body. That is, the threshold value calculated by the threshold value calculation unit 160 is stored in the personal information storage unit 130 in a state of being associated with the feature amount, the attribute information, and the collation score already stored.
- FIG. 10 is a flowchart showing the flow of the authentication operation according to the modified example of the threshold value calculation system according to the second embodiment.
- the same reference numerals are given to the same processes as those shown in FIG.
- the image acquisition unit 50 first acquires an image including a living body to be authenticated (step S21).
- the feature amount extraction unit 111 extracts the feature amount of the living body from the image data (step S22). Then, the feature amount extraction unit 111 stores the extracted feature amount in the personal information storage unit 130 (step S23). After that, the collation score calculation unit 112 calculates the collation score from the feature amount (step S24). Then, the collation score calculation unit 112 stores the calculated collation score in the personal information storage unit 130 (step S25).
- the attribute information acquisition unit 120 acquires attribute information indicating the attributes of the living body or the collation information (step S13). Then, the attribute information acquisition unit 130 stores the acquired attribute information in the personal information storage unit 130 (step S14). If the attribute information is not used at the time of authentication, it is not always necessary to acquire the attribute information in parallel. For example, the attribute information may be acquired after the authentication operation is completed according to the processing load of the system or the like.
- the collation determination unit 170 compares the collation score with the threshold value and determines whether or not there is a collation score exceeding the threshold value (step S201). Then, when there is a collation score exceeding the threshold value (step S201: YES), the collation determination unit 170 has succeeded in biometric authentication (that is, the living body to be authenticated matches the living body corresponding to the collation score exceeding the threshold value). (Step S202). In this case, the collation determination unit 170 may output an instruction to perform processing associated with the success of biometric authentication. For example, a subject who has succeeded in biometric authentication may be instructed to perform a process of opening (that is, making it possible to pass) the gate to be passed.
- the collation score to be authenticated exceeds the threshold value for a plurality of registered living organisms, it may be determined that the collation score matches the one with the higher collation score. Since it is not preferable that there are a plurality of organisms exceeding the threshold value, when a plurality of organisms exceeding the threshold value are detected, the population is estimated and the threshold value is reset (for example, the threshold value is changed to be higher). ) May be executed. Further, when the threshold value is set for each registered living body, the degree of deviation between the collation score and the threshold value (that is, how much the threshold value is exceeded) may be taken into consideration.
- biometric authentication If the biometric authentication is successful, a part or all of the information already stored in the personal information storage unit 130 (that is, feature amount, attribute information, collation score) is newly acquired according to the quality of the data. It may be rewritten to the data.
- the collation determination unit 170 determines that the biometric authentication has failed (that is, the organism to be authenticated does not match any of the registered organisms). (Step S203). In this case, the collation determination unit 170 may output an instruction to perform processing associated with the failure of biometric authentication. For example, a subject who has failed biometric authentication may be instructed to perform a process of closing (that is, making it impossible to pass) the gate to be passed.
- the threshold value used in the above-mentioned authentication operation is calculated in advance by the threshold value calculation operation described with reference to FIG. Therefore, since it is not necessary to calculate the threshold value after the authentication operation is started, it is possible to suppress an increase in the processing load in the authentication operation.
- the threshold value calculation system 10 according to the third embodiment will be described with reference to FIG.
- the third embodiment explains specific examples of the attribute information, and the configuration and operation of the system may be the same as those of the first and second embodiments described above (see FIGS. 1 to 10). ). Therefore, in the following, the description of the parts overlapping with the first and second embodiments will be omitted as appropriate.
- the threshold value calculation system 10 may use personal attribute information indicating an individual attribute of a living body as attribute information. Examples of personal attribute information include race, age, gender, skin color, and the like.
- the threshold value calculation system 10 according to the third embodiment may be used in combination with a plurality of personal attribute information. If the personal attribute information is used, an appropriate threshold value can be calculated from the population distribution considering the attributes of each individual included in the population assumed at the time of system operation. (Conditions regarding personal attribute information) Subsequently, the population condition regarding the personal attribute information will be described with reference to FIG.
- FIG. 11 is a graph showing threshold fluctuations when a sample is extracted along an assumed population distribution.
- the population condition may be set according to the ratio of the population assumed at the time of system operation. For example, if it is assumed that the ratio of males and females to be authenticated by the system is almost the same, the population condition of "male 50%, female 50%" may be set.
- the sampling unit 140 extracts males only as a part of the elements and collects sample data matching the population conditions. It should be extracted.
- the sampling unit 140 may weight the female element and extract it as a plurality of elements, and extract the sample data that matches the population condition.
- the sample data of the corresponding attribute is small and weighting is performed, population estimation is performed using only the sample data of the corresponding attribute in consideration of the bias of the sample data, and the result is used as an element for multiple elements. It may be extracted. That is, instead of using a small amount of data as it is as a plurality of data, the above-mentioned population estimation may be applied to make up for the insufficient data. More specifically, when the number of samples of 3 (average collation score 0.4) is used as the data for 10 people, for example, the collation score of 0.30 is 1 and the collation score is 0, according to the distribution of the population. 35 may be treated as 2 persons, a collation score of 0.40 as 4 persons, a collation score of 0.45 as 2 persons, a collation score of 0.45 as 2 persons, and a collation score of 0.50 as 1 person.
- a threshold value different from the threshold value calculated without considering the attributes is calculated.
- the sampling distribution shown in FIG. 11A is extracted without considering the attributes.
- the sample distribution as shown by the broken line in FIG. 11B is extracted.
- a population distribution as shown in FIG. 11 (c) can be obtained from the sample distribution considering the assumed population distribution (that is, a distribution different from the sample distribution in which the attributes are not considered). Therefore, in this case, the calculated threshold value is higher than that in the case where the attribute is not considered.
- a more appropriate threshold value can be calculated in consideration of the ratio of each attribute in the population.
- the day of the week or time zone may be used as the population condition.
- a different ratio is set as the population condition for each day of the week, for example, a male-female ratio of 40:60 on Monday and a male-female ratio of 60:40 on Saturday. It's okay.
- the ratio of men and women is 40:60 from 10:00 to 12:00, and the ratio of men and women is 60:40 from 12:00 to 14:00.
- the population condition may be set by combining the time zone and the day of the week. By using the day of the week and the time zone as the population condition, it is possible to set an appropriate population condition according to the actual operation.
- the threshold value calculation system 10 according to the fourth embodiment will be described with reference to FIG. Note that the fourth embodiment explains specific examples of attribute information as in the third embodiment, and the configuration and operation of the system may be the same as those of the first and second embodiments described above. (See FIGS. 1 to 10). Therefore, in the following, the description of the parts overlapping with the first and second embodiments will be omitted as appropriate.
- the threshold value calculation system 10 may use the environment attribute information indicating the environment in which the collation information is acquired as the attribute information.
- environmental attribute information the environment when an image of a living body is taken (for example, how the living body is reflected, camera specifications, imaging parameter information, presence / absence and intensity of a light source, background type, image quality, color). Taste, pixel brightness, etc.).
- the threshold value calculation system 10 according to the fourth embodiment may be used in combination with a plurality of environmental attribute information. If the environment attribute information is used, only the sample data acquired in an appropriate environment can be extracted, so that the influence of the data acquired in an inappropriate environment can be eliminated and an appropriate threshold value can be calculated. (Conditions related to environmental attribute information) Subsequently, the population condition regarding the environmental attribute information will be described with reference to FIG. 12.
- FIG. 12 is a graph showing threshold fluctuations when a sample is extracted by excluding the influence of the pixel luminance difference.
- the population condition may be set to extract only those having a similar degree of similarity to the environment in which the image of the living body is captured. In other words, if the environment in which the image of the living body is captured is significantly different, it may be excluded from the extraction target. Further, those obtained from those having an extremely poor index indicating the captured environment or the quality of the image may be excluded from the extraction target. That is, the information whose environmental attribute information (for example, imaging parameter, environmental parameter, resolution, etc.) does not reach a certain level may be excluded from the extraction target.
- environmental attribute information for example, imaging parameter, environmental parameter, resolution, etc.
- a threshold value different from the threshold value calculated without considering the attributes is calculated.
- the sampling distribution shown in FIG. 12A is extracted without considering the attributes.
- the sample distribution as shown by the broken line in FIG. 12B is extracted by excluding those having a large difference in pixel luminance.
- a population distribution as shown in FIG. 12 (c) can be obtained from the sample distribution excluding those having a large difference in the brightness of the image pixels (that is, a distribution different from the sample distribution in which the attributes are not considered) can be obtained. ). Therefore, in this case, the calculated threshold value is lower than that in the case where the attribute is not considered.
- the day of the week or the time zone may be used as the population condition.
- the day of the week or the time zone may be used as the population condition.
- the shape of the distribution of the population is stored as a history, and if a similar distribution is calculated in a specific time zone or day of the week, the subsequent threshold update is stopped in that specific time zone or day of the week.
- the frequency of threshold update may be reduced.
- the threshold value calculation system 10 according to the fifth embodiment will be described with reference to FIGS. 13 and 14.
- the fourth embodiment is different from the first to fourth embodiments described above in a part of the configuration and operation, and the hardware configuration and each operation described above (registration operation, threshold value calculation operation, authentication).
- the operation may be the same as each embodiment described above. Therefore, in the following, the description of the parts overlapping with the above-mentioned first to fourth embodiments will be omitted as appropriate.
- the threshold value calculation system 10 includes an image acquisition unit 50, a feature amount extraction unit 111, a collation score calculation unit 112, an attribute information acquisition unit 120, and a personal information storage unit. It includes 130, a sampling unit 140, a population estimation unit 150, a threshold value calculation unit 160, a collation determination unit 170, an authentication status storage unit 180, and a condition change unit 190. That is, the threshold value calculation system 10 according to the fifth embodiment is configured to further include an authentication status storage unit 180 and a condition change unit 190 in addition to the configuration of the modified example of the second embodiment (see FIG. 9). There is.
- the authentication status storage unit 180 may be realized by the storage device 14 (see FIG. 1) described above.
- the condition changing unit 190 may be realized by the processor 11 (see FIG. 1) described above.
- the authentication status storage unit 180 is configured to be able to store the authentication result in the collation determination unit 170.
- the authentication status storage unit 180 stores, for example, the number of registered living organisms, the number of living organisms determined to be unregistered, the ratio by attribute information, and the like.
- Various information stored in the authentication status storage unit 180 is appropriately readable by the condition change unit 190.
- the condition changing unit 190 is configured to be able to change the population condition on condition that sufficient information is accumulated in the authentication status storage unit 180.
- the condition changing unit 190 sets the population condition according to the actual operation status of the system by feeding back the information stored in the authentication status storage unit 180, for example. For example, when it can be determined from the information stored in the authentication status storage unit 180 that the population ratio has changed, the condition changing unit 190 changes the population condition according to the changed ratio. Change to.
- FIG. 14 is a flowchart showing the flow of the fluctuation operation of the population condition by the threshold value calculation system according to the fifth embodiment.
- the condition changing unit 190 first determines whether or not sufficient authentication history is accumulated in the authentication status storage unit 180 (step S301). .. Specifically, the condition changing unit 190 may determine whether or not a large amount of authentication history is accumulated so that the population in actual operation can be estimated.
- condition changing unit 190 may be configured so that the conditions can be changed manually (for example, by an operation of a system administrator or the like). For example, if an instruction to change the population condition is input immediately, the condition changing unit 190 may change the population condition according to the authentication history even if sufficient authentication history is not accumulated. good. Further, when the specific condition of the population condition is input, the condition changing unit 190 may adopt the input population condition as it is without using the accumulated authentication history.
- the population condition is appropriately changed, so that sample data is newly extracted from the changed population condition and the threshold value is set. Will be updated. Therefore, it is possible to update the threshold value more appropriately than when the population condition is not changed. Specifically, it is possible to realize threshold update according to the population in actual operation.
- the threshold value calculation system 10 according to the sixth embodiment will be described with reference to FIG.
- the sixth embodiment explains a specific example of the timing of executing the threshold value calculation operation, and the system configuration and other operations may be the same as those of the first to fifth embodiments described above. Therefore, in the following, the description of the part that overlaps with the part already described will be omitted as appropriate.
- FIG. 15 is a flowchart showing the flow of the threshold value calculation operation in the threshold value calculation system according to the sixth embodiment.
- the same reference numerals are given to the same processes as those shown in FIG.
- the example shown in FIG. 15 is premised on a configuration including the condition changing unit 190 as in the fourth embodiment.
- the threshold value calculation system 10 first determines whether or not the population condition has been changed by the condition changing unit 190 (step S401). If it is determined that the population condition has not been changed (step S401: NO), the subsequent processing is omitted, and the series of processing ends.
- step S401 when it is determined that the population condition has been changed (step S401: YES), the threshold value calculation system 10 executes the threshold value calculation operation described in FIG. 4 (that is, the processing of steps S101 to S104). Therefore, the threshold value is calculated (updated) at the timing when the population condition is changed. In this way, the threshold value can be updated at an appropriate timing according to the change in the population that should be assumed in actual operation.
- FIG. 16 is a flowchart showing the flow of the threshold value calculation operation in the threshold value calculation system according to the seventh embodiment.
- the same reference numerals are given to the same processes as those shown in FIG.
- the threshold value calculation system 10 first determines whether or not new biological information is stored in the personal information storage unit (step S501). That is, the threshold value calculation system 10 determines whether or not the registration operation described with reference to FIGS. 3 and 8 has been executed. If it is determined that new biological information is not stored (step S501: NO), the subsequent processing is omitted, and a series of processing is completed.
- step S501 when it is determined that new biological information has been stored (step S501: YES), the threshold value calculation system 10 executes the threshold value calculation operation described in FIG. 4 (that is, the processing of steps S101 to S104). Therefore, the threshold value is calculated (updated) at the timing when new biological information is stored. By doing so, it becomes possible to update the threshold value at the timing when the elements that can be extracted as sample data increase. Further, since the threshold value is calculated at the stage of registering the biometric information, it is possible to suppress an increase in the processing load when performing biometric authentication (that is, it is not necessary to execute the process of calculating the threshold value at the time of collation).
- the threshold value calculation system 10 according to the eighth embodiment will be described with reference to FIGS. 17 and 18. It should be noted that the eighth embodiment differs from the above-mentioned first to seventh embodiments only in a part of the operation (specifically, the operation of registering the unregistered person data), and the system configuration and other operations are different. The operation may be the same as each embodiment described above. Therefore, in the following, the description of the parts overlapping with the first to fifth embodiments will be omitted as appropriate.
- FIG. 17 is a flowchart showing the flow of the authentication operation according to the modified example of the threshold value calculation system according to the eighth embodiment.
- the same reference numerals are given to the same processes as those shown in FIG.
- the threshold value calculation system 10 executes the authentication operation described with reference to FIG. That is, the threshold value calculation system 10 executes the processes of steps S21 to S25 and steps S201 to S203 of FIG.
- the collation determination unit 170 determines whether or not the collation score of the organism to be authenticated is equal to or less than the second threshold value (step S203).
- the "second threshold value” here is a threshold value for determining whether or not the living body to be authenticated is an unregistered person (that is, a living body in which data is not stored in the personal information storage unit 130). It is set as a value lower than the threshold value used for biometric authentication.
- the second threshold value may be, for example, the value of the population average.
- the collation determination unit 170 determines that the living body is an unregistered person, and stores the attribute information and the collation score in the personal information storage unit 130.
- the collation determination unit 170 may store personal information such as a feature amount in the personal information storage unit 130.
- the collation determination unit 170 determines that the living body is not an unregistered person, and does not store the above-mentioned information.
- the determination using the above-mentioned second threshold value is only an example, and it may be determined whether or not the person is an unregistered person by using another method. For example, if the collation score is below all the threshold values, it may be determined that the living body is an unregistered person. That is, in the above-mentioned example, the determination process using the second threshold value is performed when the collation score is lower than all the threshold values, but when the collation score is lower than all the threshold values, the second threshold value is used. The determination process may be omitted to determine that the living body is an unregistered person.
- FIG. 18 is a diagram showing an example of UI display when storing unregistered person data.
- the UI may be realized by the output device 16 (see FIG. 1) described above.
- a display requesting the consent of the unregistered person may be displayed on the UI display.
- the unregistered person touches "Yes” it is determined that the consent to the use of the personal information has been obtained, and the personal information of the unregistered person is stored in the personal information storage unit 130.
- the unregistered person touches "No” it is determined that the consent to use the personal information has not been obtained, and the personal information of the unregistered person is not stored in the personal information storage unit 130.
- the unregistered person is made to judge only whether or not there is consent, but for example, it may be made to judge whether or not it may be used (remembered) for each type of information.
- a list of information to be used may be displayed on the UI, and an unregistered person may be asked to select information that may be used.
- the information of the unregistered person is stored in the personal information storage unit 130.
- the information of unregistered persons can also be used as sample data, so that the population can be estimated more accurately.
- the threshold calculation system includes a first acquisition means for acquiring collation information used for collation of a living body, a second acquisition means for acquiring attribute information indicating the living body or the attribute of the collation information, and each living body.
- a storage means for storing the collation information and the attribute information, a sample extraction means for extracting a plurality of the collation information as sample data from the storage means based on a predetermined condition regarding the attribute information, and a sample data.
- the threshold calculation system according to Appendix 2 further includes an image acquisition means for acquiring an image including a living body and a feature amount extracting means for extracting the feature amount of the living body from the image, and the storage means is the feature amount.
- the threshold value calculation system according to the appendix 3 is the threshold value calculation system according to the appendix 1 or 2, wherein the attribute information includes personal attribute information indicating an individual attribute of the living body.
- the threshold value calculation system according to the appendix 4 is the threshold value calculation system according to the appendix 3, wherein the predetermined condition relates to a ratio to an attribute indicated by the personal attribute information.
- the threshold value calculation system according to the appendix 5 is the threshold value calculation system according to any one of the appendices 1 to 4, wherein the collation information includes environment attribute information indicating the acquired environment.
- the threshold value calculation system according to the appendix 6 is the threshold value calculation system according to the appendix 5, wherein the predetermined condition relates to the similarity of the environmental attribute information or the level of the environmental attribute information. ..
- the threshold value calculation system can change the predetermined conditions based on the storage means for accumulating the result of the authentication process by comparing the collation information with the threshold value and the accumulated result.
- the threshold value calculation system according to any one of Supplementary note 1 to 6, further comprising various condition changing means.
- the threshold value calculation system according to the appendix 8 is the threshold value calculation system according to the appendix 7, wherein the sample extraction means extracts the sample data when the predetermined condition is changed.
- the threshold value calculation system according to Appendix 9 is characterized in that the sample extraction means extracts the sample data when new collation information and attribute information are stored in the storage means.
- the threshold value calculation system according to any one of the above items.
- the threshold calculation system according to Appendix 10 is characterized in that the storage means stores the collation information and the attribute information for the living body in which the authentication process by comparing the collation information with the threshold has failed.
- the threshold value calculation system according to any one of Items 1 to 9.
- the threshold calculation method acquires collation information used for collation of a living body, acquires attribute information indicating an attribute of the living body or the collation information, and stores the collation information and the attribute information for each living body.
- a plurality of the matching information is extracted as sample data from the storage means based on a predetermined condition regarding the attribute information, a population is estimated from the sample data, and the distribution of the estimated population is obtained.
- This is a threshold calculation method characterized by calculating a threshold related to the collation information based on the above.
- Appendix 12 The computer program according to Appendix 12 acquires collation information used for collation of a living body, acquires attribute information indicating the attributes of the living body or the collation information, and stores the collation information and the attribute information for each living body.
- a plurality of the matching information is extracted as sample data from the storage means based on a predetermined condition regarding the attribute information, a population is estimated from the sample data, and the distribution of the estimated population is obtained.
- it is a computer program characterized in that a computer is operated so as to calculate a threshold value related to the collation information.
- Appendix 13 The recording medium described in Appendix 13 is a recording medium characterized in that the computer program described in Appendix 12 is recorded.
- Threshold calculation system 11 Processor 14 Storage device 16 Output device 50 Image acquisition unit 110 Collation information acquisition unit 111 Feature quantity acquisition unit 112 Collation score calculation unit 120 Attribute information acquisition unit 130 Personal information storage unit 140 Sampling unit 150 Population estimation unit 160 Threshold calculation unit 170 Collation judgment unit 180 Authentication status storage unit 190 Condition change unit
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Human Computer Interaction (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Computer Security & Cryptography (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Collating Specific Patterns (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Image Analysis (AREA)
Abstract
Description
第1実施形態に係る閾値算出システムについて、図1から図6を参照して説明する。
まず、図1を参照しながら、第1実施形態に係る閾値算出システムのハードウェア構成について説明する。図1は、第1実施形態に係る閾値算出システムのハードウェア構成を示すブロック図である。
次に、図2を参照しながら、第1実施形態に係る閾値算出システム10の機能的構成について説明する。図2は、第1実施形態に係る閾値算出システムの機能的構成を示すブロック図である。
次に、図3を参照しながら、第1実施形態に係る閾値算出システム10による個人情報の登録動作(即ち、個人情報記憶部130に生体の照合情報及び属性情報を記憶する動作)の流れについて説明する。図3は、第1実施形態に係る閾値算出システムにおける登録動作の流れを示すフローチャートである。
次に、図4を参照しながら、第1実施形態に係る閾値算出システム10による閾値算出動作の流れについて説明する。図4は、第1実施形態に係る閾値算出システムにおける閾値算出動作の流れを示すフローチャートである。
次に、図5及び図6を参照しながら、第1実施形態に係る閾値算出システム10によって得られる技術的効果について説明する。図5は、閾値と他人許容率及び本人拒否率との関係を示すグラフである。図6は、標本分布の違いによる理想閾値の違いを示すグラフである。
第2実施形態に係る閾値算出システム10について、図7及び図8を参照して説明する。なお、第2実施形態は、上述した第1実施形態と比べて一部の構成及び動作が異なるのみであり、例えばハードウェア構成については第1実施形態(図1参照)と同一であってよい。このため、以下では、第1実施形態と重複する部分については適宜説明を省略するものとする。
まず、図7を参照しながら、第2実施形態に係る閾値算出システム10の機能的構成について説明する。図7は、第2実施形態に係る閾値算出システムの機能的構成を示すブロック図である。なお、図7では、図2で示した構成要素と同様の要素に同一の符号を付している。
次に、図8を参照しながら、第2実施形態に係る閾値算出システム10による個人情報の登録動作の流れについて説明する。図8は、第2実施形態に係る閾値算出システムにおける登録動作の流れを示すフローチャートである。なお、図8では、図3で示した処理と同様の処理に同一の符号を付している。
次に、第2実施形態に係る閾値算出システム10による閾値算出動作の流れについて説明する。なお、第2実施形態に係る閾値算出動作は、第1実施形態に係る閾値算出動作(図4参照)と同様のフローである。このため、新たな図については割愛し、適宜図4を参照して説明を進める。
次に、第2実施形態に係る閾値算出システム10によって得られる技術的効果について説明する。
次に、第2実施形態に係る閾値算出システム10の変形例について、図9及び図10を参照して説明する。なお、以下で説明する変形例は、上述した第2実施形態で算出した閾値を用いて、生体認証動作を行うものである。
まず、図9を参照しながら、第2実施形態の変形例に係る閾値算出システム10の機能的構成について説明する。図9は、第2実施形態に係る閾値算出システムの変形例の機能的構成を示すブロック図である。なお、図9では、図7で示した構成要素と同様の要素に同一の符号を付している。
続いて、図10を参照しながら、第2実施形態の変形例に係る閾値算出システム10による認証動作について説明する。図10は、第2実施形態に係る閾値算出システムの変形例による認証動作の流れを示すフローチャートである。なお、図10では、図8で示した処理と同様の処理に同一の符号を付している。
第3実施形態に係る閾値算出システム10について、図11を参照して説明する。なお、第3実施形態は、属性情報に関する具体例を説明するものであり、システムの構成や動作については、上述した第1及び第2実施形態と同一であってよい(図1から図10参照)。このため、以下では、第1及び第2実施形態と重複する部分については適宜説明を省略するものとする。
まず、第3実施形態に係る閾値算出システム10で用いられる個人属性情報について説明する。
(個人属性情報に関する条件)
続いて、図11を参照しながら、個人属性情報に関する母集団条件について説明する。図11は、想定される母集団分布に沿って標本を抽出した場合の閾値変動を示すグラフである。
第4実施形態に係る閾値算出システム10について、図12を参照して説明する。なお、第4実施形態は、第3実施形態と同様に属性情報に関する具体例を説明するものであり、システムの構成や動作については、上述した第1及び第2実施形態と同一であってよい(図1から図10参照)。このため、以下では、第1及び第2実施形態と重複する部分については適宜説明を省略するものとする。
まず、第4実施形態に係る閾値算出システム10で用いられる環境属性情報について説明する。
(環境属性情報に関する条件)
続いて、図12を参照しながら、環境属性情報に関する母集団条件について説明する。図12は、画素輝度差による影響を排除して標本を抽出した場合の閾値変動を示すグラフである。
第5実施形態に係る閾値算出システム10について、図13及び14を参照して説明する。なお、第4実施形態は、上述した第1から第4実施形態と比べて一部の構成及び動作が異なるものであり、ハードウェア構成、及び上述した各動作(登録動作、閾値算出動作、認証動作)については、すでに説明した各実施形態と同一であってよい。このため、以下では、上述した第1から第4実施形態と重複する部分については適宜説明を省略するものとする。
図13に示すように、第5実施形態に係る閾値算出システム10は、画像取得部50と、特徴量抽出部111と、照合スコア算出部112と、属性情報取得部120と、個人情報記憶部130と、標本抽出部140と、母集団推定部150と、閾値算出部160と、照合判定部170と、認証状況記憶部180と、条件変更部190とを備えている。即ち、第5実施形態に係る閾値算出システム10は、第2実施形態の変形例の構成(図9参照)に加えて、認証状況記憶部180、及び条件変更部190を更に備えて構成されている。なお、認証状況記憶部180は、上述した記憶装置14(図1参照)によって実現されてよい。また、条件変更部190は、上述したプロセッサ11(図1参照)によって実現されてよい。
次に、図14を参照しながら、第5実施形態に係る閾値算出システム10による条件変更動作について説明する。図14は、第5実施形態に係る閾値算出システムによる母集団条件の変動動作の流れを示すフローチャートである。
次に、第5実施形態に係る閾値算出システム10によって得られる技術的効果について説明する。
第6実施形態に係る閾値算出システム10について、図15を参照して説明する。なお、第6実施形態は、閾値算出動作を実行するタイミングの具体例を説明するものであり、システム構成やその他の動作については、上述した第1から第5実施形態と同一であってよい。このため、以下では、すでに説明した部分と重複する部分については適宜説明を省略するものとする。
まず、図15を参照しながら、母集団条件の変更に伴う閾値算出動作について説明する。図15は、第6実施形態に係る閾値算出システムにおける閾値算出動作の流れを示すフローチャートである。なお図15では、図4に示した処理と同様の処理に同一の符号を付している。また、図15に示す例は、第4実施形態のように条件変更部190を備えた構成を前提としている。
第7実施形態に係る閾値算出システム10について、図16を参照して説明する。なお、第7実施形態は、閾値算出動作を実行するタイミングの具体例を説明するものであり、システム構成やその他の動作については、上述した第1から第5実施形態と同一であってよい。このため、以下では、すでに説明した部分と重複する部分については適宜説明を省略するものとする。
(個人情報記憶に伴う閾値算出)
図16を参照しながら、新たな個人情報の記憶に伴う閾値算出動作について説明する。図16は、第7実施形態に係る閾値算出システムにおける閾値算出動作の流れを示すフローチャートである。なお図16では、図4に示した処理と同様の処理に同一の符号を付している。
第8実施形態に係る閾値算出システム10について、図17及び図18を参照して説明する。なお、第8実施形態は、上述した第1から第7実施形態と比べて一部の動作(具体的には、未登録者データを登録する動作)が異なるのみであり、システム構成やその他の動作については、すでに説明した各実施形態と同一であってよい。このため、以下では、第1から第5実施形態と重複する部分については適宜説明を省略するものとする。
まず、図17を参照しながら、第8実施形態に係る閾値算出システム10による認証動作について説明する。図17は、第8実施形態に係る閾値算出システムの変形例による認証動作の流れを示すフローチャートである。なお、図17では、図10で示した処理と同様の処理に同一の符号を付している。
次に、図18を参照しながら、未登録者データを記憶する場合のUI(User Interface)の表示例について説明する。図18は、未登録者データを記憶する場合のUI表示例を示す図である。なお、UIは、上述した出力装置16(図1参照)によって実現されてよい。
次に、第8実施形態に係る閾値算出システム10によって得られる技術的効果について説明する。
以上説明した実施形態に関して、更に以下の付記のようにも記載されうるが、以下には限られない。
付記1に記載の閾値算出システムは、生体の照合に用いる照合情報を取得する第1取得手段と、前記生体又は前記照合情報の属性を示す属性情報を取得する第2取得手段と、前記生体毎に前記照合情報及び前記属性情報を記憶する記憶手段と、前記属性情報に関する所定の条件に基づいて、前記記憶手段から複数の前記照合情報を標本データとして抽出する標本抽出手段と、前記標本データから母集団を推定する母集団推定手段と、前記推定された母集団の分布に基づいて、前記照合情報に関する閾値を算出する閾値算出手段とを備えることを特徴とする閾値算出システムである。
付記2に記載の閾値算出システムは、生体を含む画像を取得する画像取得手段と、前記画像から前記生体の特徴量を抽出する特徴量抽出手段とを更に備え、前記記憶手段は、前記特徴量及び前記特徴量を前記生体間で比較することで得られる照合スコアの少なくとも一方を、前記照合情報として記憶することを特徴とする付記1に記載の閾値算出システムである。
付記3に記載の閾値算出システムは、前記属性情報は、前記生体の個人の属性を示す個人属性情報を含むことを特徴とする付記1又は2に記載の閾値算出システムである。
付記4に記載の閾値算出システムは、前記所定の条件は、前記個人属性情報が示す属性に割合に関するものであることを特徴とする付記3に記載の閾値算出システムである。
付記5に記載の閾値算出システムは、前記照合情報が取得された環境を示す環境属性情報を含むことを特徴とする付記1から4のいずれか一項に記載の閾値算出システムである。
付記6に記載の閾値算出システムは、前記所定の条件は、前記環境属性情報の類似度、又は前記環境属性情報の水準に関するものであることを特徴とする付記5に記載の閾値算出システムである。
付記7に記載の閾値算出システムは、前記照合情報と前記閾値との比較による認証処理の結果を蓄積する蓄積手段と、前記蓄積された結果に基づいて、前記所定の条件を変更することが可能な条件変更手段とを更に備えることを特徴とする付記1から6のいずれか一項に記載の閾値算出システムである。
付記8に記載の閾値算出システムは、前記標本抽出手段は、前記所定の条件が変更された場合に、前記標本データを抽出することを特徴とする付記7に記載の閾値算出システムである。
付記9に記載の閾値算出システムは、前記標本抽出手段は、前記記憶手段に新たな前記照合情報及び属性情報が記憶された場合に、前記標本データを抽出することを特徴とする付記1から8のいずれか一項に記載の閾値算出システムである。
付記10に記載の閾値算出システムは、前記記憶手段は、前記照合情報と前記閾値との比較による認証処理が失敗した前記生体について、前記照合情報及び前記属性情報を記憶することを特徴とする請求項1から9のいずれか一項に記載の閾値算出システムである。
付記11に記載の閾値算出方法は、生体の照合に用いる照合情報を取得し、前記生体又は前記照合情報の属性を示す属性情報を取得し、前記生体毎に前記照合情報及び前記属性情報を記憶手段に記憶し、前記属性情報に関する所定の条件に基づいて、前記記憶手段から複数の前記照合情報を標本データとして抽出し、前記標本データから母集団を推定し、前記推定された母集団の分布に基づいて、前記照合情報に関する閾値を算出することを特徴とする閾値算出方法である。
付記12に記載のコンピュータプログラムは、生体の照合に用いる照合情報を取得し、前記生体又は前記照合情報の属性を示す属性情報を取得し、前記生体毎に前記照合情報及び前記属性情報を記憶手段に記憶し、前記属性情報に関する所定の条件に基づいて、前記記憶手段から複数の前記照合情報を標本データとして抽出し、前記標本データから母集団を推定し、前記推定された母集団の分布に基づいて、前記照合情報に関する閾値を算出するようにコンピュータを動作させることを特徴とするコンピュータプログラムである。
付記13に記載の記録媒体は、付記12に記載のコンピュータプログラムが記録されていることを特徴とする記録媒体である。
11 プロセッサ
14 記憶装置
16 出力装置
50 画像取得部
110 照合情報取得部
111 特徴量取得部
112 照合スコア算出部
120 属性情報取得部
130 個人情報記憶部
140 標本抽出部
150 母集団推定部
160 閾値算出部
170 照合判定部
180 認証状況記憶部
190 条件変更部
Claims (12)
- 生体の照合に用いる照合情報を取得する第1取得手段と、
前記生体又は前記照合情報の属性を示す属性情報を取得する第2取得手段と、
前記生体毎に前記照合情報及び前記属性情報を記憶する記憶手段と、
前記属性情報に関する所定の条件に基づいて、前記記憶手段から複数の前記照合情報を標本データとして抽出する標本抽出手段と、
前記標本データから母集団を推定する母集団推定手段と、
前記推定された母集団の分布に基づいて、前記照合情報に関する閾値を算出する閾値算出手段と
を備えることを特徴とする閾値算出システム。 - 生体を含む画像を取得する画像取得手段と、
前記画像から前記生体の特徴量を抽出する特徴量抽出手段と
を更に備え、
前記記憶手段は、前記特徴量及び前記特徴量を前記生体間で比較することで得られる照合スコアの少なくとも一方を、前記照合情報として記憶する
ことを特徴とする請求項1に記載の閾値算出システム。 - 前記属性情報は、前記生体の個人の属性を示す個人属性情報を含むことを特徴とする請求項1又は2に記載の閾値算出システム。
- 前記所定の条件は、前記個人属性情報が示す属性に割合に関するものであることを特徴とする請求項3に記載の閾値算出システム。
- 前記属性情報は、前記照合情報が取得された環境を示す環境属性情報を含むことを特徴とする請求項1から4のいずれか一項に記載の閾値算出システム。
- 前記所定の条件は、前記環境属性情報の類似度、又は前記環境属性情報の水準に関するものであることを特徴とする請求項5に記載の閾値算出システム。
- 前記照合情報と前記閾値との比較による認証処理の結果を蓄積する蓄積手段と、
前記蓄積された結果に基づいて、前記所定の条件を変更することが可能な条件変更手段と
を更に備えることを特徴とする請求項1から6のいずれか一項に記載の閾値算出システム。 - 前記標本抽出手段は、前記所定の条件が変更された場合に、前記標本データを抽出することを特徴とする請求項7に記載の閾値算出システム。
- 前記標本抽出手段は、前記記憶手段に新たな前記照合情報及び属性情報が記憶された場合に、前記標本データを抽出することを特徴とする請求項1から8のいずれか一項に記載の閾値算出システム。
- 前記記憶手段は、前記照合情報と前記閾値との比較による認証処理が失敗した前記生体について、前記照合情報及び前記属性情報を記憶することを特徴とする請求項1から9のいずれか一項に記載の閾値算出システム。
- 生体の照合に用いる照合情報を取得し、
前記生体又は前記照合情報の属性を示す属性情報を取得し、
前記生体毎に前記照合情報及び前記属性情報を記憶手段に記憶し、
前記属性情報に関する所定の条件に基づいて、前記記憶手段から複数の前記照合情報を標本データとして抽出し、
前記標本データから母集団を推定し、
前記推定された母集団の分布に基づいて、前記照合情報に関する閾値を算出する
ことを特徴とする閾値算出方法。 - 生体の照合に用いる照合情報を取得し、
前記生体又は前記照合情報の属性を示す属性情報を取得し、
前記生体毎に前記照合情報及び前記属性情報を記憶手段に記憶し、
前記属性情報に関する所定の条件に基づいて、前記記憶手段から複数の前記照合情報を標本データとして抽出し、
前記標本データから母集団を推定し、
前記推定された母集団の分布に基づいて、前記照合情報に関する閾値を算出する
ようにコンピュータを動作させることを特徴とするコンピュータプログラム。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2020/037286 WO2022070349A1 (ja) | 2020-09-30 | 2020-09-30 | 閾値算出システム、閾値算出方法、及びコンピュータプログラム |
US18/023,932 US20240029466A1 (en) | 2020-09-30 | 2020-09-30 | Threshold calculation system, threshold calculation method, and computer program |
JP2022553343A JP7485063B2 (ja) | 2020-09-30 | 2020-09-30 | 閾値算出システム、閾値算出方法、及びコンピュータプログラム |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2020/037286 WO2022070349A1 (ja) | 2020-09-30 | 2020-09-30 | 閾値算出システム、閾値算出方法、及びコンピュータプログラム |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022070349A1 true WO2022070349A1 (ja) | 2022-04-07 |
Family
ID=80949933
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2020/037286 WO2022070349A1 (ja) | 2020-09-30 | 2020-09-30 | 閾値算出システム、閾値算出方法、及びコンピュータプログラム |
Country Status (3)
Country | Link |
---|---|
US (1) | US20240029466A1 (ja) |
JP (1) | JP7485063B2 (ja) |
WO (1) | WO2022070349A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024195114A1 (ja) * | 2023-03-23 | 2024-09-26 | 日本電気株式会社 | 入場管理システム、認証装置、認証方法、および記録媒体 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010108074A (ja) * | 2008-10-28 | 2010-05-13 | Hitachi Ltd | 生体認証システム及び方法 |
JP2013122679A (ja) * | 2011-12-12 | 2013-06-20 | Hitachi Ltd | 安全性の高い生体認証システム |
JP2016072836A (ja) * | 2014-09-30 | 2016-05-09 | 株式会社日立製作所 | 逐次バイオメトリック暗号システムおよび逐次バイオメトリック暗号処理方法 |
US20170286746A1 (en) * | 2016-03-31 | 2017-10-05 | Synaptics Incorporated | Efficient determination of biometric attribute for fast rejection of enrolled templates and other applications |
JP2018028815A (ja) * | 2016-08-18 | 2018-02-22 | 富士通株式会社 | 評価装置、評価方法および評価プログラム |
-
2020
- 2020-09-30 JP JP2022553343A patent/JP7485063B2/ja active Active
- 2020-09-30 WO PCT/JP2020/037286 patent/WO2022070349A1/ja active Application Filing
- 2020-09-30 US US18/023,932 patent/US20240029466A1/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010108074A (ja) * | 2008-10-28 | 2010-05-13 | Hitachi Ltd | 生体認証システム及び方法 |
JP2013122679A (ja) * | 2011-12-12 | 2013-06-20 | Hitachi Ltd | 安全性の高い生体認証システム |
JP2016072836A (ja) * | 2014-09-30 | 2016-05-09 | 株式会社日立製作所 | 逐次バイオメトリック暗号システムおよび逐次バイオメトリック暗号処理方法 |
US20170286746A1 (en) * | 2016-03-31 | 2017-10-05 | Synaptics Incorporated | Efficient determination of biometric attribute for fast rejection of enrolled templates and other applications |
JP2018028815A (ja) * | 2016-08-18 | 2018-02-22 | 富士通株式会社 | 評価装置、評価方法および評価プログラム |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024195114A1 (ja) * | 2023-03-23 | 2024-09-26 | 日本電気株式会社 | 入場管理システム、認証装置、認証方法、および記録媒体 |
Also Published As
Publication number | Publication date |
---|---|
US20240029466A1 (en) | 2024-01-25 |
JPWO2022070349A1 (ja) | 2022-04-07 |
JP7485063B2 (ja) | 2024-05-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10719694B2 (en) | Biometric identification and verification | |
US10049361B2 (en) | Dynamic authentication technology | |
CN106022032A (zh) | 个人识别装置、识别阈值设定方法、以及程序 | |
JP5185358B2 (ja) | 行動履歴検索装置 | |
Kuru et al. | Biomedical visual data analysis to build an intelligent diagnostic decision support system in medical genetics | |
US8242881B2 (en) | Method of adjusting reference information for biometric authentication and apparatus | |
JP5633080B2 (ja) | 属性値推定装置、属性値推定方法、プログラム及び記録媒体 | |
US8754747B2 (en) | Authentication device, authentication system, and authentication method | |
EP2416274A1 (en) | Biometric authentication device, biometric authentication method, and storage medium | |
WO2022070349A1 (ja) | 閾値算出システム、閾値算出方法、及びコンピュータプログラム | |
JP2019121188A (ja) | 生体認証装置、生体認証方法および生体認証プログラム | |
JP2018124999A (ja) | 情報処理装置及びプログラム | |
US20200294651A1 (en) | Sleep improvement assistance system, method, and program | |
US20230329610A1 (en) | Device control apparatus, non-transitory computer-readable medium, and device control method | |
JP7029726B2 (ja) | 判定システム、判定方法及びプログラム | |
EP4131148A1 (en) | Authentication device, authentication method, and recording medium | |
JP2011086202A (ja) | 照合装置、照合方法および照合プログラム | |
JPWO2021199114A5 (ja) | 推薦装置、システム、方法及びプログラム | |
JP6728699B2 (ja) | 生体認証装置、生体認証方法および生体認証プログラム | |
JP5737435B2 (ja) | 生体認証装置、生体認証方法、および記憶媒体 | |
WO2023188332A1 (ja) | 人物特定装置、人物特定方法、および人物特定プログラム | |
JP2019197434A (ja) | 提供方法及び提供装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20956277 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18023932 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 2022553343 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20956277 Country of ref document: EP Kind code of ref document: A1 |