WO2010103735A1 - 顔照合装置、電子機器、顔照合装置の制御方法、および顔照合装置制御プログラム - Google Patents
顔照合装置、電子機器、顔照合装置の制御方法、および顔照合装置制御プログラム Download PDFInfo
- Publication number
- WO2010103735A1 WO2010103735A1 PCT/JP2010/001133 JP2010001133W WO2010103735A1 WO 2010103735 A1 WO2010103735 A1 WO 2010103735A1 JP 2010001133 W JP2010001133 W JP 2010001133W WO 2010103735 A1 WO2010103735 A1 WO 2010103735A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- face image
- image
- age
- target
- time
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/5838—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/10—Recognition assisted with metadata
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/178—Human faces, e.g. facial parts, sketches or expressions estimating age from face image; using age information for improving recognition
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/179—Human faces, e.g. facial parts, sketches or expressions metadata assisted face recognition
Definitions
- the present invention relates to a face verification apparatus, an electronic device, a control method of the face verification apparatus, and a face verification apparatus control program, which collate a target face image with a face image of a registrant.
- digital cameras In recent years, with the spread of digital still cameras and digital video cameras (hereinafter collectively referred to as “digital cameras”), a large amount of digital data (images) such as still images and moving images captured by the digital camera It is stored in a storage medium. And, since a large number of images are stored in this storage medium, it is difficult to find out an image in which a desired person appears from the stored images, and it is difficult to classify the images according to the appearing person. It is becoming
- the face image registered and the photographed face image are collated to determine whether they are the same person or not, so-called face authentication It is conceivable to use
- Patent Document 1 generates an image to be newly registered from the image input at the time of matching and the stored image, and automatically updates the registered image to the newly generated image. By doing so, techniques are disclosed that reduce the effects of aging.
- Patent Document 3 discloses a technique for deleting an image of low importance, which has a long elapsed time from the registration time or the time of last verification, from a storage medium.
- Patent Document 4 a feature amount corresponding to the time from the time when the face image is registered to the time when the face is shot is calculated, and the accuracy of authentication is enhanced by performing face authentication using this feature amount.
- Patent Document 5 discloses a technique for performing authentication that matches reality by performing authentication using biometric data (image data) captured at a plurality of points in the past.
- Patent Document 6 discloses a technique for differentially compressing and storing a captured face image based on the difference between a similar face image indicating a degree of similarity equal to or higher than a predetermined reference value and the captured face image. .
- Japanese patent publication Japanese Patent Application Laid-Open No. 11-175718 (July 2, 1999 published)” Japanese Patent Publication "Japanese Patent Application Publication No. 2003-44858 (February 14, 2003)” Japanese Patent Publication "Japanese Patent Application Laid-Open No. 2004-54888 (February 19, 2004)” Japanese Patent Publication "Japanese Patent Application Laid-Open No. 2007-102517 (Apr. 19, 2007)” Japanese Patent Publication "2006-92491 (Apr. 6, 2006)” Japanese Patent Publication "Japanese Patent Application Laid-Open Publication No. 2006-59200 (released on March 2, 2006)"
- Patent Document 1 discloses the image stored in the storage medium, even if the target image is searched and classified using the authentication technique disclosed in Patent Document 1 described above.
- the technology corresponds to the latest state of the person to be authenticated by updating the registration data, and can not correspond to images of the same person at various ages. Therefore, with respect to the image at the age which can not be coped with, the accuracy of the authentication is lowered, and it is not always possible to accurately collate the target image, search and classify.
- Patent Document 2 performs authentication using a plurality of pieces of biological information, and can not cope with images at various ages stored in a storage medium.
- Patent Document 3 is for deleting an image of low importance from a storage medium, and does not lead to accurate matching, search, and classification of images.
- Patent Document 4 calculates a feature corresponding to the time from the time when the face image is registered to the time when the face is shot, and the feature corresponding to this time is It is only for calculation purposes, and is different from the actual face. Therefore, it is not always possible to accurately authenticate, and it is not always possible to accurately collate, search, and classify images.
- Patent Document 5 authenticates the current face using biometric data captured at a plurality of points in the past, and it is possible to accurately authenticate images at various ages. It is not possible to match, search and classify images accurately.
- Patent Document 6 differentially compresses and stores a captured face image, and does not lead to accurate matching, search, and classification of the image.
- the present invention has been made in view of the above problems, and an object thereof is to accurately collate a desired face image even from among a plurality of images including face images at various ages.
- the object is to realize a face matching device or the like.
- the face matching apparatus matches a target face image to be matched included in a captured image with a reference face image which is a reference face image registered in a reference image database.
- a face collating apparatus wherein the reference image database can store a plurality of reference face images for each registrant, and further, photographing the reference face image with respect to each reference face image It is possible to store a reference face image shooting time that is a time, and obtaining means for obtaining a target face image shooting time which is a shooting time of the target face image, and a target face image shooting time obtained by the obtaining means. And selecting means for selecting the reference face image used for the matching based on the reference face image photographing time point stored in the reference image database.
- the control method of the face matching device is a face matching device that matches a target face image to be matched included in a captured image with a reference face image that is a reference face image registered in a reference image database.
- a plurality of reference face images can be stored in the reference image database for each registrant, and further, when the reference face image is captured for each reference face image
- the reference used for matching based on the target face image photographing time which is the photographing time of the acquired target face image and the reference face image photographing time stored in the reference image database A face image is selected.
- the reference face image taken at the time of shooting close to the time of shooting of the target face image is a case where the registrant whose age is close to that of the registrant at the time of shooting of the target face image is shot
- More accurate matching can be performed by selecting the reference face image photographed at the photographing time point close to the time point.
- the face matching apparatus can store a plurality of reference face images for each registrant in the reference image database, and further, for each reference face image described above. It is possible to store the reference face image shooting time which is the shooting time of the reference face image, and the obtaining means for obtaining the target face image shooting time which is the shooting time of the target face image; And a selection unit configured to select the reference face image used for the matching based on the target face image photographing time point and the reference face image photographing time point stored in the reference image database.
- the control method of the face collation device it is possible to store a plurality of reference face images for each registrant in the reference image database, and further, for each reference face image, It is possible to store a reference face image photographing time which is a photographing time of a reference face image, and an acquiring step of acquiring a target face image photographing time which is a photographing time of the target face image; And a selection step of selecting the reference face image to be used for the matching based on the face image photographing time point and the reference face image photographing time point stored in the reference image database.
- the reference face image taken at the time of shooting close to the time of shooting of the target face image is a case where the registrant whose age is close to that of the registrant at the time of shooting of the target face image is shot
- By selecting the reference face image photographed at the photographing time point close to the time point it is possible to perform more accurate matching.
- the face matching device can store a plurality of reference face images for each registrant in the reference image database, and further, the birth time of each registrant and each reference It is possible to store a reference age, which is the age of the registrant at the time of shooting of the face image, and an acquisition means for acquiring the time of photographing of the target face image at the time of photographing of the target face image; And selecting means for selecting the reference face image to be used for the matching based on the target face image photographing time point and the birth time and reference time age stored in the reference image database.
- the control method of the face matching device it is possible to store a plurality of reference face images for each registrant in the reference image database, and further, the birth time of each registrant, And acquiring the reference age, which is the age of the registrant at the shooting time of each reference face image, and obtaining the target face image shooting time at the shooting time of the target face image;
- FIG. 1 showing an embodiment of the present invention, is a block diagram showing an essential configuration of a collation unit and a storage unit. It is a block diagram which shows the principal part structure of the face collation apparatus in the said embodiment. It is a figure which shows the content memorize
- FIG. 2 is a block diagram of the face matching device 1 according to the present embodiment.
- the face verification apparatus (electronic device) 1 is configured to include a display unit (notification means) 11, an external device connection unit 12, an operation unit 13, a storage unit 14, and a control unit 10.
- face collation device 1 is explained as a personal computer below, it is not restricted to this. It may be various devices having a function of collating images, such as a digital camera.
- the face matching device 1 is a person who is shown in an image (target image) to be matched, such as an image acquired from an external device connected via the external device connection unit 12 or an image stored in the storage unit 14.
- the face (target face image) is determined to which face image of the registrant's face images (reference face images) stored in the storage unit 14 the reference of matching is most similar, and the result is displayed It is
- the reference face image and the photographed image to be compared are described as collated and searched, the feature amount stored in the reference face image and the photographed image and the photographed image are stored. It includes the case where the search is performed by collating with the feature quantity of.
- the face of the person appearing in the image to be matched is similar to any of the reference face images by appropriately selecting the reference face image in which the reference face is shown using the application degree described below. It is possible to determine the accuracy of
- the display unit 11 displays the operation performed by the face matching device 1, the state of the face matching device 1, and the like as a result of the matching in the face matching device 1.
- the display unit 11 may be realized by any device as long as it can display information, but can be realized by, for example, an LCD (liquid crystal display), an organic EL (electroluminescence) display, a plasma display, or the like.
- the external device connection unit 12 is connected to an external device to transmit and receive data to and from the external device.
- Examples of the external device include, for example, a scanner and a hard disk.
- the operation unit 13 receives an operation on the face matching device 1 from the user.
- the storage unit 14 stores a reference face image serving as a reference for matching, image data such as a photographed image to be matched, a program for operating the face matching apparatus 1, and the like.
- a flash memory can be mentioned as an example of the storage unit 14. The details of the storage unit 14 will be described later.
- the control unit 10 performs matching, search, and registration of a reference face image in the face matching device 1. Details of the control unit 10 will be described later.
- the control unit 10 is configured to include a collation unit 16, a registration processing unit 17, and an output processing unit 18.
- the collation part 16 is an object image acquisition part (acquisition means) 101, an application degree calculation part (application degree calculation means, object age calculation means) 102, the feature quantity extraction part 103, collation utilization image, as shown in FIG.
- the configuration includes a selection unit (selection unit) 104, a similarity calculation unit 105, a collation reliability calculation unit (weighted collation result calculation unit) 106, a collation result determination unit 107, and a result output unit (inquiry unit).
- FIG. 1 is a block diagram showing the configuration of matching unit 16 and storage unit 14 according to the present embodiment.
- the storage unit 14 is configured to include a reference image database 121, an application degree coefficient table storage unit 122, and an image database 123.
- the reference image database 121 is a database that stores image data of a reference face image serving as a reference for matching.
- FIG. 3 shows the contents stored in the reference image database 121.
- the ID identification
- the date and time when the reference face image was taken shooting date, shooting time, reference time
- the ID is, for example, the name of a person appearing in the reference face image.
- reference time age information either the reference time age itself or information capable of calculating the reference time age will be referred to as reference time age information.
- the application coefficient table storage unit 122 indicates a relationship between an age (target age) and a coefficient at the time of shooting (target time) of a photographed image (target image) of a registrant captured in a reference face image.
- a coefficient table of degrees is stored.
- the application degree indicates how much the reference face image for matching the photographed image can be applied to the matching.
- “applicability 1 / (difference between shooting date of shooting image and shooting date of reference face image ⁇ coefficient)”.
- the value when the difference between the photographed image and the reference face image on the day of the newborn is “1” (no change) is statistically defined as “1” (no change), with this as the upper limit.
- the application degree of the captured image is calculated from the age of the subject person at the time of capturing the captured image.
- FIG. 4 shows an example in which the relationship between the coefficient for calculating the degree of application and the age of the subject at the time of photographing of a photographed image is shown by a graph.
- FIG. 4 is a diagram showing an example of a graph showing a coefficient table of the application degree, which is a relationship between the coefficient for obtaining the application degree and the target age, which is the age of the registrant at the shooting time of the target image.
- the coefficient has a maximum value “1” at the age of 0, and the coefficient decreases as the age increases. And, the lower the age, the greater the change in coefficients. Further, in the example shown in FIG. 4, when the age of the registrant at the shooting date of the shot image is 20, the coefficient is 0.1.
- the method of determining the degree of application is not limited to this, and any method may be used as long as it can indicate how much the reference face image used to collate the photographed image can be applied to collation. May be Further, although the application degree is calculated using the coefficient in the above, the application degree may be calculated using only the difference between the shooting date of the shot image and the shooting date of the reference face image without using the coefficient.
- the image database 123 is a database storing images.
- the image stored in the image database 123 is an image acquired from an external device (for example, a scanner or a digital camera) connected via the external device connection unit 12 or acquired via the Internet when it can be connected to the Internet Image.
- the image database 123 stores the image and the date and time when the image was taken.
- the collation unit 16 includes a target image acquisition unit 101, an application degree calculation unit 102, a feature amount extraction unit 103, a collation use image selection unit 104, a similarity calculation unit 105, and a collation reliability calculation unit 106.
- the configuration includes a matching result determination unit 107 and a result output unit 108. Then, the collation unit 16 collates the photographed image to be collated with the reference face image to be collated, and outputs the result.
- the target image acquisition unit 101 acquires a captured image (target image) to be compared. Then, photographed image data indicating the acquired photographed image is transmitted to the application degree calculation unit 102 and the feature amount extraction unit 103.
- the photographed image to be compared may be acquired from the image database 123 or may be acquired from an external device via the external device connection unit 12.
- the target image acquisition unit 101 acquires information indicating the target time that is the shooting time of the captured image as information of the target age along with the captured image, and transmits the information to the applicability calculation unit 102.
- the information on target age is either the target age itself or information that can calculate the target age.
- the application degree calculation unit 102 uses the coefficient table stored in the application degree coefficient table storage unit 122 to compare the photographed image acquired from the target image acquisition unit 101 with the reference face image stored in the reference image database 121.
- the application degree is calculated by the method described above.
- the application degree of each of the plurality of reference face images and the photographed image is calculated.
- the reference face image and the application degree data indicating the calculated application degree of the reference face image are transmitted to the collation-use image selection unit 104.
- the age of the registrant at the time of shooting of the target image can be calculated from the date of birth of the registrant and the time of shooting of the target image if the reference image database 121 stores the date of birth of the registrant. Further, even if the date of birth is not stored, if the date and time of shooting and the age of the registrant at that time are stored, it can be calculated from the time of shooting of the information and the target image.
- the difference between the shooting date and time of the target image and the reference face image can be calculated from the shooting date and time of both.
- the shooting date and time of the reference face image is not stored, if the age and date of birth of the registrant at the time of shooting the reference face image are stored, the approximate shooting date and time is Since the calculation can be performed, the difference between the shooting date and time of the target image and the reference face image can be calculated using this.
- the collation-use image selection unit 104 selects, from the application degree data acquired from the application degree calculation unit 102, a reference face image whose application degree exceeds a threshold as a reference face image to be used for collation. Then, the selected reference face image and the application degree are transmitted to the similarity calculation unit 105.
- the collation-use image selection unit 104 may select the reference face image to be used for collation by the number of sheets for each ID, or may be selected in consideration of the number of sheets for each ID and a threshold. It may be
- the number required for collation may be determined in advance for each ID, and the required number of reference face images may be selected as the reference face image to be used for collation from the highly applicable reference face image.
- a necessary number of reference face images may be selected as a reference face image to be used for collation from the excess reference face images.
- the feature amount extraction unit 103 extracts the feature amount of the captured image acquired from the target image acquisition unit 101. Then, the feature amount data indicating the extracted feature amount is transmitted to the similarity calculation unit 105.
- the feature amount indicates face features such as the contour of the face, the color of the eyes, the shape of the eyes, the shape of the nose, the shape of the lips, and the like shown in the photographed image.
- the similarity calculation unit 105 uses the feature amount data of the captured image acquired from the feature amount extraction unit 103 and the feature amount data of the reference face image acquired from the matching usage image selection unit 104 to display the person's image captured in the captured image. The degree of similarity between the face and the face of the person appearing in the reference face image is calculated. Then, the similarity degree data indicating the calculated similarity degree, the ID included in the reference face image data on which the similarity degree is calculated, and the application degree are transmitted to the matching reliability calculation unit 106. When a plurality of reference face images to be used for matching is selected in the matching usage image selecting unit 104, similarity is obtained from the feature amount data of each reference face image and the feature amount data acquired from the feature amount extracting unit 103. The degree is calculated, and the result is transmitted to the matching reliability calculation unit 106.
- the collation reliability calculation unit 106 calculates the collation reliability of the ID from the similarity, the application degree, and the ID acquired from the similarity calculation unit 105. Then, the calculated matching reliability is transmitted to the matching result determination unit 107 together with the ID.
- the matching reliability is calculated by taking the weighted average of the similarity by weighting the application for each ID. This will be described with reference to FIG. FIG. 5 is an explanatory view showing an example of the degree of similarity and the degree of application for each ID, and the degree of matching reliability.
- FIG. 5 is an explanatory view showing an example of the degree of similarity and the degree of application for each ID, and the degree of matching reliability.
- FIG. 5 is an explanatory view showing an example of the degree of similarity and the degree of application for each ID, and the degree of matching reliability.
- ID “A” A1, A2 and A3 there are three reference face images of ID “A” A1, A2 and A3, and the similarity is 0.8, 0.6, 0.9, and the application degree is
- the collation result determination unit 107 determines an ID having the highest collation reliability from the collation reliability and the ID acquired from the collation reliability calculation unit 106, and transmits the ID and the collation reliability to the result output unit 108.
- the result output unit 108 transmits the ID acquired from the matching result determination unit 107 to the output processing unit 18 together with the matching reliability as the matching result.
- FIG. 6 is a block diagram showing the configuration of the registration processing unit 17 according to the present embodiment.
- the registration processing unit 17 includes a reference face image acquisition unit 110, an age / life calculation unit 111, a feature extraction unit 112, a similarity calculation unit 113, a registration unit (image registration unit) 114, and an addition.
- An image recommendation unit (necessary age output means) 115 is included. Then, the registration processing unit 17 registers, in the reference image database 121, an image serving as a reference used when collating images.
- the reference face image acquisition unit 110 acquires image data to be a reference face image and an ID indicating a person appearing in the image data, and transmits it to the age / birth year calculation unit 111 and the feature amount extraction unit 112. .
- the image data includes data indicating an image and data indicating a shooting date and time of the image.
- the image data may be acquired from the image database 123 or may be acquired from an external device connected via the external device connection unit 12.
- the reference face image acquisition unit 110 can acquire the ID.
- the age and birth date calculation unit 111 calculates the age and birth date of the person (target person) appearing in the image from the image data (reference face image) acquired from the reference face image acquisition unit 110. Then, the calculated age and birth year are transmitted to the registration unit 114. In the calculation of the age and the birth year, the age at the time of shooting is calculated from the shooting date and date and the date of birth, and the birth year is calculated from the shooting date and time and the age at the shooting. If the reference face image of the same ID is stored in the reference image database 121, the date of birth of the subject may be acquired from the image data, or if it is a new registration, it may be acquired by the user's input. It is also good.
- the feature amount extraction unit 112 extracts the feature amount of the image data acquired from the reference face image acquisition unit 110, and transmits the feature amount data indicating the feature amount to the similarity calculation unit 113 together with the ID.
- the similarity calculation unit 113 calculates the similarity from the feature amount data acquired from the feature amount extraction unit 112 and the feature amount data of the reference face image stored in the reference image database 121 having the same ID as the acquired ID. Do. Then, similarity data indicating the calculated similarity is transmitted to the registration unit 114 together with the ID.
- the registration unit 114 determines whether the similarity obtained from the similarity calculation unit 113 is too large or too small with respect to the age difference. Then, if there is a similarity within a predetermined range, the acquired image data is registered in the reference image database 121 as a reference face image, along with the feature amount, ID, shooting date, age and age acquired from the age and birth year calculation unit 111. Do. Then, data indicating that the registration has been completed is transmitted to the additional image recommendation unit 115 together with the ID.
- the reference face image of the ID stored in the reference image database 121 is of the age of the ID of the age. Even if an image in which a person is photographed becomes a target of comparison as a photographed image, it is determined whether or not a reference face image of application degree exceeding a predetermined value can be selected. Then, if it is determined that the image can not be selected, the reference face image of the application degree exceeding the predetermined value can be selected even if the image in which the person with the ID of any age is taken as the target of comparison as the photographed image. Information indicating the age required for the output processing unit 18 is transmitted.
- the output processing unit 18 causes the display unit 11 to display the result based on the information acquired from the result output unit 108 or the additional image recommendation unit 115. That is, when the ID of the collation reliability equal to or more than the threshold is acquired from the result output unit 108, the ID is displayed on the display unit 11 as the collation result. In addition, when the ID of the collation reliability less than the threshold is acquired, the display unit 11 displays the ID as a collation result together with a display indicating that the degree of the reliability is low.
- the user is asked whether the result of the collation is correct or not, and if an operation indicating that the user is correct is accepted, the photographed image at this time is used as a reference face image of the ID.
- the registration may be prompted in the database 121.
- the display unit 11 when information indicating the required reference age is acquired from the additional image recommendation unit 115, the display unit 11 causes the display unit 11 to display a message prompting registration of the age.
- FIG. 7 is a flowchart showing the flow of processing when the face matching apparatus 1 performs matching of images.
- the target image acquisition unit 101 determines whether a reference face image is registered in the reference image database 121. (S701).
- the matching process is ended.
- the feature extraction unit 103 tries to extract the feature of the captured image acquired from the target image acquisition unit 101 (S702).
- the application degree calculation unit 102 stores the application degree of the reference face image stored in the reference image database 121 in the application degree coefficient table storage unit 122. It calculates using the coefficient table of the application degree (S703).
- the collation-use image selection unit 104 selects a reference face image whose applicability calculated by the application calculation unit 102 exceeds a predetermined value as a reference face image to be used for collation (S704).
- the similarity calculating unit 105 calculates the similarity between the reference face image selected by the matching-use image selecting unit 104 and the photographed image (S705).
- the matching reliability calculation unit 106 calculates the matching reliability using the degree of application and the degree of similarity (S706).
- the processing from S703 to S706 is performed for all the IDs of the reference face image stored in the reference image database 121 (S707). Then, when the processing from S703 to S706 is completed for all the IDs (YES in S707), the collation result determination unit 107 determines the ID of the reference face image having the highest collation reliability as the collation result (S708). Then, the result output unit 108 outputs the determined matching result as the matching result (S709). Above, the flow of collation processing is completed.
- the age of the person (target person) appearing in the photographed image in the reference image database 121 is approximately the same.
- the user is notified that the photographed image is registered as a reference face image, and the photographed image is registered as a reference face image by the operation of the user It may be one.
- the photographed image may be registered as a reference face image of the subject.
- the collation-use image selection unit 104 matches the reference face image whose applicability is not more than the predetermined value. Select as a reference face image to use. Then, using the selected reference face image, the above-described S705.
- the result output unit 108 performs an output for inquiring whether the collation result is correct. Then, when the result output unit 108 receives an input that the collation result is correct, the registration unit 114 registers the image of the subject (photographed image) in the reference image database 121 as the reference face image of the subject.
- FIG. 8 is a flowchart showing a flow of processing for registering a reference face image in the reference image database 121.
- the feature amount extraction unit 112 tries to extract the feature amount of the reference face image acquired by the reference face image acquisition unit 110 (S801). Then, if the feature amount can not be extracted (NO in S801), the process of registering the reference face image is ended. On the other hand, if the feature amount can be extracted (YES in S801), the registration processing unit 17 determines whether the ID of the reference face image is already registered in the reference image database 121 (S802). Then, if the ID is not registered in the reference image database 121 (YES in S802), the registration processing unit 17 displays a date of birth input screen (S803), and from the user, the date of birth of the person indicated by the ID. The input of the day is accepted (S804).
- the age / birth year calculation unit 111 uses the registered date of birth and the acquired face image data.
- the age of the person (target person) appearing in the reference face image is calculated based on the included shooting date and time (S805). If the captured image is captured in real time, the date and time of capturing may be determined from current time data managed by the apparatus.
- the similarity calculation unit 113 calculates the similarity between the reference face image to be newly registered and the reference face image of the same ID already registered in the reference image database 121 (S806). Then, the registration unit 114 determines whether the similarity is incorrect based on the similarity calculated by the similarity calculation unit 113 and the age difference when appearing in the image for which the similarity is calculated (S807).
- whether the degree of similarity is incorrect is determined from the relationship between the age difference and the degree of similarity. That is, since it is conceivable that the degree of similarity is small when the age difference is large, it is considered that the degree of similarity is not incorrect. On the other hand, although the difference in age is small, the fact that the degree of similarity is small is not generally considered, and the degree of similarity is incorrect. Also, if the age difference is small and the degree of similarity is greater than a predetermined value, the degree of similarity is determined to be incorrect. This is because the advantage of registering a plurality of similar images with similar age is considered to be small.
- the registration unit 114 determines whether to hold an existing reference face image or register a new reference face image. (S808). Then, if the existing reference face image is held (YES in S808), the registration process is ended. On the other hand, if the existing reference face image is not held (NO in S808), the registration processing unit 17 deletes the reference face image registered in the reference image database 121 (S809). Then, a new reference face image is registered (S810).
- the additional image recommendation unit 115 determines the age of the reference face image having the same ID as the ID of the newly registered reference face image registered in the reference image database 121. It is determined whether it is possible to select as the reference face image of the application degree exceeding the threshold also for the photographed image of. And when it is judged that the reference face image in which the person of the ID of another age is taken is necessary to be able to select as the reference face image of the application degree exceeding the threshold value with respect to the photographed image of any age. In step S811, the user is notified that a reference face image in which a person of the required age is photographed is registered (S811).
- the reference face image whose applicability exceeds the threshold is selected as the reference face image used for matching
- the reference face using all the reference face images for matching without using the applicability as the threshold It may be configured to select as an image.
- the user may be notified that the reliability is low. If the user determines that the output result is correct even though the matching reliability is lower than the threshold, the photographed image used for this matching is newly registered as a reference face image. May be
- the present invention may have the following configuration. That is, a face matching apparatus that matches a target face image that is a target face image with a reference face image that is a face image of a registrant's reference and outputs a matching result, wherein the target face image is captured Target age at the time of the registrant at a certain target time and target age related information that is any of information that can calculate the target age, and for each of a plurality of the above reference face images, the standard at the shooting time Acquisition means for acquiring a union at the reference time, which is the age of the registrant at the time, and reference age-related information, which is any of information that can calculate the reference age, and a sum acquired by the acquisition means
- a face collating apparatus comprising: selection means for selecting the standard face image to be used for the collation using the target age related information and the standard age related information from a set.
- the storage unit stores a plurality of reference times of the reference face image and the birth time of the registrant, and the target age-related information includes the target time and the birth time of the registrant. And the reference age-related information includes a reference time of the reference face image and a birth time of the registrant.
- the storage unit stores a reference age of the plurality of reference face images and a birth date of the registrant, and the target age-related information includes the target time and the birth time of the registrant.
- the reference time age related information is a reference time age of the reference face image.
- the storage unit stores a plurality of reference times and reference times of the reference face image, and the target age-related information includes the target time, the reference time and reference times of the reference face, and the like. And the reference time age related information is a reference time age of the reference face image.
- the face matching apparatus calculates and calculates the reference age, which is the age of the reference face image of the registrant, from the photographing time of the reference face image and the birth time of the registrant.
- the age / birth year calculation unit 111 is provided to store the reference time age in the storage unit.
- the age and birth year calculation unit 111 calculates the reference age of the registrant from the photographing date of the reference face image and the birth date of the registrant, and stores the reference age of the registrant in the storage unit. As a result, it is possible to omit the task of the user manually inputting the reference age.
- the face matching apparatus matches the target face image to be matched included in the captured image with the reference face image that is the reference face image registered in the reference image database.
- a plurality of reference face images can be stored in the reference image database for each registrant, and the reference face image is captured at each reference face image
- a reference face image shooting time point can be stored, an acquisition unit for acquiring a target face image shooting time point which is a shooting time point of the target face image, a target face image shooting time point acquired by the acquisition means, and the reference And selecting means for selecting the reference face image to be used for the comparison based on the reference face image photographing time point stored in the image database.
- the control method of the face matching device is a face matching device that matches a target face image to be matched included in a captured image with a reference face image that is a reference face image registered in a reference image database.
- a plurality of reference face images can be stored in the reference image database for each registrant, and further, when the reference face image is captured for each reference face image
- the reference used for matching based on the target face image photographing time which is the photographing time of the acquired target face image and the reference face image photographing time stored in the reference image database A face image is selected.
- the reference face image taken at the time of shooting close to the time of shooting of the target face image is a case where the registrant whose age is close to that of the registrant at the time of shooting of the target face image is shot
- More accurate matching can be performed by selecting the reference face image photographed at the photographing time point close to the time point.
- the reference image database can further store, for each of the reference face images described above, a reference time age which is the age of the registrant at the time of shooting the reference face image.
- the selection means may further select the reference face image to be used for the matching, using the reference age.
- age may include not only the number of years elapsed since birth but also the date and time.
- the various time information includes the date, it may include the time, or may include only the year and month.
- the reference image database can further store the birth time of each registrant, and the selection means further uses the birth time to perform the collation.
- the reference face image to be used may be selected.
- the face matching apparatus matches a target face image to be matched included in a captured image with a reference face image which is a reference face image registered in a reference image database.
- a face collation apparatus wherein a plurality of reference face images can be stored for each registrant in the reference image database, and further, photographing of each birth date of each registrant and each reference face image It is possible to store a reference time age, which is the age of a registrant at a time, and an acquisition means for acquiring a target face image photographing time which is a photographing time of the target face image, and a target face image acquired by the acquisition means And selecting means for selecting the reference face image to be used for the matching on the basis of the photographing time and the birth time and the reference time age stored in the reference image database.
- the control method of the face matching device is a face matching device that matches a target face image to be matched included in a captured image with a reference face image that is a reference face image registered in a reference image database.
- the reference image database can store a plurality of reference face images for each registrant, and further, the birth time of each registrant and each reference face image. It is possible to store a reference age, which is the age of the registrant at the time of shooting, and an acquisition step of acquiring a target face image photographing time which is the photographing time of the target face image; And a selection step of selecting the reference face image to be used for the matching based on the image photographing time point and the birth time and reference time age stored in the reference image database. It is characterized in.
- the matching is performed based on the shooting time of the target face image that is the shooting time of the obtained target face image and the birth time and reference age of the registrant stored in the reference image database.
- a reference face image to be used is selected.
- the reference face image can be calculated using the age of the registrant at the time of shooting of the target face image and the reference age. Can be selected.
- the target face image photographing time point acquired by the acquisition means is an applicability indicating how much the reference face image can be applied to collation with a target face image included in the photographed image.
- Application degree calculation means for calculating based on the difference between the reference face image photographing time point stored in the reference image database and the selection degree, and the selection degree calculated by the application degree calculation means exceeds a predetermined threshold It is preferable to select a reference face image.
- the application degree calculation means calculates based on the difference between the target face image shooting time point acquired by the acquisition means and the reference face image shooting time point stored in the reference image database. Then, the selection unit selects a reference face image in which the application degree calculated by the application degree calculation unit exceeds a threshold.
- the degree of application is calculated based on the difference between the target face image shooting time which is the shooting time of the target face image and the reference face image shooting time which is the shooting time of the reference face image. Since the reference face image used for collation is selected in consideration of the period up to the reference face image photographing time, more accurate collation can be performed.
- the photographing of the target face image is performed based on the target face image photographing time acquired by the acquisition means and the reference face image photographing time and reference age stored in the reference image database.
- Target age calculation means for calculating target age, which is the age of the registrant at the time, and applicability indicating the extent to which the reference face image can be applied to matching with the target face image included in the captured image Is calculated based on the difference between the target face image shooting time acquired by the acquisition means and the reference face image imaging time stored in the reference image database, and the target age calculated by the target age calculation means.
- the image processing apparatus may further include application degree calculation means, and the selection means may select a reference face image in which the application degree calculated by the application degree calculation means exceeds a predetermined threshold.
- the application degree is calculated based on the difference between the target face image shooting time point acquired by the acquisition unit and the reference face image shooting time point, and the target age. Then, the selection unit selects a reference face image in which the application degree calculated by the application degree calculation unit exceeds a threshold.
- the application degree is calculated based on the difference between the target face image shooting time that is the shooting time of the target image and the reference face image shooting time that is the shooting time of the reference face image, and the age at the target time. The age and the time period from the target face image shooting time to the reference face image shooting time are added.
- the reference face image whose applicability exceeds the threshold that is, the age at which the target face image was taken and the period from the target face image photographing point to the reference face image photographing point can be applied to matching the target face image. Only the reference face image whose L exceeds a threshold is used for matching with the target face image, and more accurate matching can be performed.
- the registrant's at the photographing time of the target face image based on the photographing time of the target face image acquired by the acquiring means and the birth time stored in the reference image database.
- the acquisition means for calculating the target time age calculating means for calculating the target time age which is the age, and the application degree indicating how much the reference face image can be applied to the comparison with the target face image included in the photographed image. It has an application degree calculation means to calculate based on the difference between the acquired target face image photographing time and the reference face image photographing time stored in the reference image database and the target age calculated by the target age calculating means.
- the selection unit may select a reference face image in which the application degree calculated by the application degree calculation unit exceeds a predetermined threshold.
- the age of the registrant at the photographing time of the target face image based on the photographing time of the target face image acquired by the acquiring unit and the birth time stored in the reference image database.
- the target image age calculating means for calculating the target object age and the application degree indicating the extent to which the reference face image can be applied to the comparison with the target face image included in the captured image are stored in the reference image database
- Application degree calculating means for calculating based on the difference between the standard age at the time of calculation and the target age calculated by the age calculating means for the target, and the selecting A reference face image exceeding a predetermined threshold may be selected.
- the application rate has a larger change rate with respect to the difference between the target face image photographing time and the reference face image photographing time as the target age decreases.
- the application degree becomes the reference for matching from the time of shooting of the target face image at the time of shooting of the target face image.
- the applicability changes significantly with respect to the difference between the target face image shooting time and the reference face image shooting time Do.
- the change does not significantly change with respect to the difference between the target face image photographing time and the reference face image photographing time.
- the application degree also changes significantly during the period in which the change in the face (the secular change) is large with respect to the elapsed period, and the application degree does not change significantly in the period with the small secular change.
- the degree can be calculated.
- the matching can be performed using the reference face image whose application degree has been calculated more finely, and accurate matching can be performed more precisely corresponding to the secular change of the human face.
- a calculation means may be provided.
- the weighted matching result calculation means calculates the weighted matching result weighted by the application degree of the reference face image with respect to the comparison result between the reference face image selected by the selection means and the target face image. Do. As a result, the result of the matching using the highly applicable reference face image is heavier, the result of the matching using the less applicable reference face image is lighter, and the weighted result is calculated. Can output the result of the matching.
- the range of subject age at which the application degree is equal to or higher than the threshold is calculated from the baseline age of the reference face image stored in the reference image database. If the target age is an age not included in the category, the required age output means is provided to calculate the range of the standard age at which the application degree is equal to or higher than the above threshold and output the calculated standard age range. It may be one.
- the required age output unit calculates the range of target age at which the application degree is equal to or higher than the threshold from the standard age, and sets the age not included in the calculated target age range to the target age. If it is determined that the range of the reference age at which the application degree is equal to or more than the threshold is calculated, the calculated range of the reference age is output.
- the user can recognize the reference age necessary for the selection means to select the reference face image of the application degree equal to or higher than the threshold regardless of the age at the target age. Thereby, the user can be prompted to register a reference face image of the reference age.
- the selecting means can select the reference face image having a threshold value or more regardless of the age of the registrant at the time of photographing the target face image. Therefore, more accurate matching can be performed.
- the application degree indicating the degree to which the reference face image can be applied to the collation of the face image included in the photographed image is the target face image photographing time acquired by the acquiring unit and An application degree calculation unit that calculates based on the difference from the reference face image shooting time stored in the reference face image database is provided, and the selection unit is a reference face image whose applicability calculated by the application degree calculation unit exceeds a threshold If the reference face image below the threshold is selected and the inquiry means for inquiring whether the comparison result is correct or not and the inquiry means receive that it is correct, the target face image is registered as the reference face image. And registration means.
- the selection means selects the reference face image below the predetermined threshold, and when the matching result is correct, the image registration means The target face image is registered as the reference face image.
- the target face image is registered as the reference face image, so that in the subsequent matching, the target face image
- more accurate matching can be performed.
- an electronic device concerning the present invention is an electronic device provided with the above-mentioned face collation device, and when there is no standard face image whose above-mentioned applicability exceeds a threshold to the above-mentioned object face image. And a notification unit for notifying the user.
- the notification means if there is no reference face image whose degree of applicability exceeds the threshold value, the notification means notifies the user. From this, the user can recognize that there is no reference face image whose applicability exceeds the threshold.
- the face collation device may be realized by a computer.
- a control program of the face collation device for realizing the face collation device by the computer by operating the computer as the respective means is also the present invention. Fall into the category of
- each block of the face matching device 1, particularly the matching unit 16 of the control unit 10, the registration processing unit 17, the output processing unit 18, the target image obtaining unit 101 of the matching unit 16, the application degree calculation unit 102, the feature quantity extraction unit 103, collation utilization image selection unit 104, similarity calculation unit 105, collation reliability calculation unit 106, collation result determination unit 107, result output unit 108, reference face image acquisition unit 110 of registration processing unit 17, age / life calculation unit 111, the feature extraction unit 112, the similarity calculation unit 113, the registration unit 114, and the additional image recommendation unit 115 may be configured by hardware logic, or using a CPU (central processing unit) as follows. It may be realized by software.
- the face verification apparatus 1 includes a CPU that executes an instruction of a control program that implements each function, a ROM (read only memory) that stores the program, a random access memory (RAM) that expands the program, the program, and various programs.
- a storage device such as a memory for storing data is provided.
- the object of the present invention is a recording medium in which a program code (an execution form program, an intermediate code program, a source program) of a control program of the face verification apparatus 1 which is software that realizes the functions described above is readable by computer. This can also be achieved by supplying the face verification apparatus 1 and the computer (or CPU or MPU (microprocessor unit)) reads out and executes the program code stored in the recording medium.
- the recording medium examples include tape systems such as magnetic tapes and cassette tapes, magnetic disks such as floppy (registered trademark) disks / hard disks, and compact disc read-only memory (CD-ROM) / MO (magneto-optical) / Disk system including an optical disk such as MD (Mini Disc) / DVD (digital versatile disk) / CD-R (CD Recordable), IC card (including memory card) / card system such as optical card, or mask ROM / EPROM ( A semiconductor memory system such as an erasable programmable read-only memory (EEPROM) / EEPROM (electrically erasable and programmable read-only memory) / flash ROM can be used.
- tape systems such as magnetic tapes and cassette tapes
- magnetic disks such as floppy (registered trademark) disks / hard disks
- CD-ROM compact disc read-only memory
- MO magneticto-optical
- Disk system including an optical disk such as MD (Mini Disc) / DVD (
- the face verification apparatus 1 may be configured to be connectable to a communication network, and the program code may be supplied via the communication network.
- the communication network is not particularly limited.
- the Internet intranet, extranet, LAN (local area network), ISDN (integrated services digital network), VAN (value-added network), CATV (community antenna television) communication
- a network a virtual private network, a telephone network, a mobile communication network, a satellite communication network, etc. can be used.
- the transmission medium constituting the communication network is not particularly limited.
- IEEE institute of electrical and electronic engineers 1394, USB, power line carrier, cable TV line, telephone line, ADSL (asynchronous digital subscriber loop) line
- wired such as infrared (IRDA) (infrared data association) or remote control such as infrared, Bluetooth (registered trademark), 802.11 radio, HDR (high data rate), wireless such as mobile phone network, satellite link, terrestrial digital network etc.
- IRDA infrared
- Bluetooth registered trademark
- 802.11 radio high data rate
- wireless such as mobile phone network, satellite link, terrestrial digital network etc.
- the present invention can also be realized in the form of a computer data signal embedded in a carrier wave, in which the program code is embodied by electronic transmission.
- Face recognition device electronic equipment
- control unit 11 display unit (notification means) 12 external device connection unit 13 operation unit 14 storage unit 16 collation unit 17 registration processing unit 18 output processing unit 101 target image acquisition unit (acquisition means) 102 Applicability calculation unit (applicability calculation means, target age calculation means) 103 Feature extraction unit 104 Matching-use image selection unit (selection means) 105 Similarity calculation unit 106 Comparison reliability calculation unit (weighted comparison result calculation means) 107 Verification Result Determination Unit 108 Result Output Unit (Inquiry Means) 110 Reference face image acquisition unit 111 Age / life calculation unit 112 Feature quantity extraction unit 113 Similarity calculation unit 114 Registration unit (image registration means) 115 Additional Image Recommendation (Required Age Output Means) 121 Reference Image Database 122 Application Coefficient Table Storage Unit 123 Image Database
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Collating Specific Patterns (AREA)
- Image Analysis (AREA)
- Processing Or Creating Images (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Processing (AREA)
Abstract
Description
~S708の処理を行い、S709で結果を出力するときに、結果出力部108は、照合結果が正しいか否かを問い合わせる出力を行う。そして、結果出力部108が、照合結果は正しいとの入力を受け付けると、登録部114は、対象者の画像(撮影画像)を該対象者の基準顔画像として、基準画像データベース121へ登録する。
10 制御部
11 表示部(報知手段)
12 外部機器接続部
13 操作部
14 記憶部
16 照合部
17 登録処理部
18 出力処理部
101 対象画像取得部(取得手段)
102 適用度算出部(適用度算出手段、対象時年齢算出手段)
103 特徴量抽出部
104 照合利用画像選択部(選択手段)
105 類似度算出部
106 照合信頼度算出部(加重照合結果算出手段)
107 照合結果決定部
108 結果出力部(問い合わせ手段)
110 基準顔画像取得部
111 年齢・生年算出部
112 特徴量抽出部
113 類似度算出部
114 登録部(画像登録手段)
115 追加画像推奨部(必要年齢出力手段)
121 基準画像データベース
122 適用度係数テーブル記憶部
123 画像データベース
Claims (17)
- 撮影画像に含まれる照合対象となる対象顔画像を、基準画像データベースに登録された基準の顔画像である基準顔画像と照合する顔照合装置であって、
上記基準画像データベースには、各登録者に対して複数の基準顔画像を記憶することが可能であり、さらに、各基準顔画像に対して該基準顔画像の撮影時点である基準顔画像撮影時点を記憶することが可能であり、
上記対象顔画像の撮影時点である対象顔画像撮影時点を取得する取得手段と、
上記取得手段が取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された基準顔画像撮影時点とに基づいて、上記照合に利用する上記基準顔画像を選択する選択手段と、を備えていることを特徴とする顔照合装置。 - 上記基準画像データベースには、さらに、各基準顔画像に対して該基準顔画像の撮影時点における登録者の年齢である基準時年齢を記憶することが可能であり、
上記選択手段は、さらに上記基準時年齢を用いて、上記照合に利用する上記基準顔画像を選択することを特徴とする請求項1に記載の顔照合装置。 - 上記基準画像データベースには、さらに、各登録者の生誕時点を記憶することが可能であり、
上記選択手段は、さらに上記生誕時点を用いて、上記照合に利用する上記基準顔画像を選択することを特徴とする請求項1に記載の顔照合装置。 - 撮影画像に含まれる照合対象となる対象顔画像を、基準画像データベースに登録された基準の顔画像である基準顔画像と照合する顔照合装置であって、
上記基準画像データベースには、各登録者に対して複数の基準顔画像を記憶することが可能であり、さらに、上記各登録者の生誕時点、および各基準顔画像の撮影時点における登録者の年齢である基準時年齢を記憶することが可能であり、
上記対象顔画像の撮影時点である対象顔画像撮影時点を取得する取得手段と、
上記取得手段が取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された生誕時点および基準時年齢とに基づいて、上記照合に利用する上記基準顔画像を選択する選択手段と、を備えていることを特徴とする顔照合装置。 - 上記基準顔画像を、上記撮影画像に含まれる対象顔画像との照合にどの程度適用できるかを示す適用度を、上記取得手段が取得した対象顔画像撮影時点と上記基準画像データベースに記憶された基準顔画像撮影時点との差に基づいて算出する適用度算出手段を備え、
上記選択手段は、上記適用度算出手段が算出した適用度が所定の閾値を超える基準顔画像を選択することを特徴とする請求項1に記載の顔照合装置。 - 上記取得手段が取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された基準顔画像撮影時点および基準時年齢とに基づいて、上記対象顔画像の撮影時点における上記登録者の年齢である対象時年齢を算出する対象時年齢算出手段と、
上記基準顔画像を、上記撮影画像に含まれる対象顔画像との照合にどの程度適用できるかを示す適用度を、上記取得手段が取得した対象顔画像撮影時点と上記基準画像データベースに記憶された基準顔画像撮影時点との差と、上記対象時年齢算出手段が算出した対象時年齢とに基づいて算出する適用度算出手段を備え、
上記選択手段は、上記適用度算出手段が算出した適用度が所定の閾値を超える基準顔画像を選択することを特徴とする請求項2に記載の顔照合装置。 - 上記取得手段が取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された上記生誕時点とに基づいて、上記対象顔画像の撮影時点における上記登録者の年齢である対象時年齢を算出する対象時年齢算出手段と、
上記基準顔画像を、上記撮影画像に含まれる対象顔画像との照合にどの程度適用できるかを示す適用度を、上記取得手段が取得した対象顔画像撮影時点と上記基準画像データベースに記憶された基準顔画像撮影時点との差と、上記対象時年齢算出手段が算出した対象時年齢とに基づいて算出する適用度算出手段を備え、
上記選択手段は、上記適用度算出手段が算出した適用度が所定の閾値を超える基準顔画像を選択することを特徴とする請求項3に記載の顔照合装置。 - 上記取得手段が取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された生誕時点とに基づいて、上記対象顔画像の撮影時点における上記登録者の年齢である対象時年齢を算出する対象時年齢算出手段と
上記基準顔画像を、上記撮影画像に含まれる対象顔画像との照合にどの程度適用できるかを示す適用度を、上記基準画像データベースに記憶された上記基準時年齢と上記対象時年齢算出手段が算出した対象時年齢とに基づいて算出する適用度算出手段を備え、
上記選択手段は、上記適用度算出手段が算出した適用度が所定の閾値を超える基準顔画像を選択することを特徴とする請求項4に記載の顔照合装置。 - 上記適用度は、上記対象時年齢が低くなるほど、対象顔画像撮影時点と基準顔画像撮影時点との差に対して変化率が大きいことを特徴とする請求項6または7に記載の顔照合装置。
- 上記選択手段が選択した基準顔画像と上記対象顔画像との照合の結果に対し、上記基準顔画像の適用度で重み付けした加重照合結果を算出する加重照合結果算出手段を備えていることを特徴とする請求項5~9のいずれか1項に記載の顔照合装置。
- 上記基準画像データベースに記憶されている基準顔画像の基準時年齢から、閾値以上の適用度となる対象時年齢の範囲を算出し、算出した対象時年齢の範囲に含まれない年齢を対象時年齢とした場合に、上記閾値以上の適用度となる基準時年齢の範囲を算出し、算出した基準撮影時年齢の範囲を出力する必要年齢出力手段を備えていることを特徴とする請求項6記載の顔照合装置。
- 上記基準顔画像を上記撮影画像に含まれる顔画像の照合にどの程度適用できるかを示す適用度を、上記取得手段が取得した上記対象顔画像撮影時点と上記基準画像データベースに記憶された基準顔画像撮影時点との差に基づいて算出する適用度算出手段を備え、
上記選択手段は、上記適用度算出手段が算出した適用度が閾値を超える基準顔画像が存在しない場合、閾値を下回る基準顔画像を選択しており、
照合結果の正否を問い合わせる問い合わせ手段と、
上記問い合わせ手段が正しいと受け付けると、当該対象顔画像を上記基準顔画像として登録する画像登録手段とを備えていることを特徴とする請求項1~4のいずれか1項に記載の顔照合装置。 - 請求項5~9のいずれか1項に記載の顔照合装置を備えた電子機器であって、
上記対象顔画像に対し、上記適用度が閾値を超える基準顔画像が無い場合に、ユーザに対し報知を行う報知手段を備えていることを特徴とする電子機器。 - 撮影画像に含まれる照合対象となる対象顔画像を、基準画像データベースに登録された基準の顔画像である基準顔画像と照合する顔照合装置の制御方法であって、
上記基準画像データベースには、各登録者に対して複数の基準顔画像を記憶することが可能であり、さらに、各基準顔画像に対して該基準顔画像の撮影時点である基準顔画像撮影時点を記憶することが可能であり、
上記対象顔画像の撮影時点である対象顔画像撮影時点を取得する取得ステップと、
上記取得ステップで取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された基準顔画像撮影時点とに基づいて、上記照合に利用する上記基準顔画像を選択する選択ステップと、を含むことを特徴とする顔照合装置の制御方法。 - 撮影画像に含まれる照合対象となる対象顔画像を、基準画像データベースに登録された基準の顔画像である基準顔画像と照合する顔照合装置の制御方法であって、
上記基準画像データベースには、各登録者に対して複数の基準顔画像を記憶することが可能であり、さらに、上記各登録者の生誕時点、および各基準顔画像の撮影時点における登録者の年齢である基準時年齢を記憶することが可能であり、
上記対象顔画像の撮影時点である対象顔画像撮影時点を取得する取得ステップと、
上記取得ステップで取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された生誕時点および基準時年齢とに基づいて、上記照合に利用する上記基準顔画像を選択する選択ステップと、を含むことを特徴とする顔照合装置の制御方法。 - 撮影画像に含まれる照合対象となる対象顔画像を、基準画像データベースに登録された基準の顔画像である基準顔画像と照合する顔照合装置制御プログラムであって、
上記基準画像データベースには、各登録者に対して複数の基準顔画像を記憶することが可能であり、さらに、各基準顔画像に対して該基準顔画像の撮影時点である基準顔画像撮影時点を記憶することが可能であり、
上記対象顔画像の撮影時点である対象顔画像撮影時点を取得する取得ステップと、
上記取得ステップで取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された基準顔画像撮影時点とに基づいて、上記照合に利用する上記基準顔画像を選択する選択ステップと、をコンピュータに実行させるための顔照合装置制御プログラム。 - 撮影画像に含まれる照合対象となる対象顔画像を、基準画像データベースに登録された基準の顔画像である基準顔画像と照合する顔照合装置制御プログラムであって、
上記基準画像データベースには、各登録者に対して複数の基準顔画像を記憶することが可能であり、さらに、上記各登録者の生誕時点、および各基準顔画像の撮影時点における登録者の年齢である基準時年齢を記憶することが可能であり、
上記対象顔画像の撮影時点である対象顔画像撮影時点を取得する取得ステップと、
上記取得ステップで取得した対象顔画像撮影時点と、上記基準画像データベースに記憶された生誕時点および基準時年齢とに基づいて、上記照合に利用する上記基準顔画像を選択する選択ステップと、をコンピュータに実行させるための顔照合装置制御プログラム。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/059,381 US8396264B2 (en) | 2009-03-13 | 2010-02-22 | Face matching device, electronic device, face matching device control method, and face matching device control program |
CN2010800023253A CN102124492B (zh) | 2009-03-13 | 2010-02-22 | 脸部对照装置、电子设备、脸部对照装置的控制方法,以及脸部对照装置控制程序 |
EP10750494.6A EP2407929B1 (en) | 2009-03-13 | 2010-02-22 | Face collation device, electronic appliance, method of controlling face collation device, and control program of face collation device |
KR1020117003462A KR101168546B1 (ko) | 2009-03-13 | 2010-02-22 | 얼굴 대조 장치, 전자 기기, 얼굴 대조 장치의 제어 방법 및 얼굴 대조 장치 제어 프로그램 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2009062039A JP4636190B2 (ja) | 2009-03-13 | 2009-03-13 | 顔照合装置、電子機器、顔照合装置の制御方法、および顔照合装置制御プログラム |
JP2009-062039 | 2009-03-13 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010103735A1 true WO2010103735A1 (ja) | 2010-09-16 |
Family
ID=42728034
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2010/001133 WO2010103735A1 (ja) | 2009-03-13 | 2010-02-22 | 顔照合装置、電子機器、顔照合装置の制御方法、および顔照合装置制御プログラム |
Country Status (6)
Country | Link |
---|---|
US (1) | US8396264B2 (ja) |
EP (1) | EP2407929B1 (ja) |
JP (1) | JP4636190B2 (ja) |
KR (1) | KR101168546B1 (ja) |
CN (1) | CN102124492B (ja) |
WO (1) | WO2010103735A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2657908A4 (en) * | 2010-12-21 | 2018-01-17 | Omron Corporation | Image authentication apparatus, image processing system, control program for image authentication apparatus, computer-readable recording medium, and image authentication method |
JP2022015263A (ja) * | 2020-07-08 | 2022-01-21 | 株式会社スクウェア・エニックス | 顔認証プログラム及び顔認証システム |
Families Citing this family (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5753966B2 (ja) * | 2010-08-05 | 2015-07-22 | パナソニックIpマネジメント株式会社 | 顔画像登録装置および方法 |
JP5618295B2 (ja) * | 2010-10-13 | 2014-11-05 | 株式会社日立情報通信エンジニアリング | 認証システム及び認証の信頼度の判定方法 |
US8965067B2 (en) * | 2010-10-28 | 2015-02-24 | Telefonaktiebolaget L M Ericsson (Publ) | Face data acquirer, end user video conference device, server, method, computer program and computer program product for extracting face data |
JP5649425B2 (ja) * | 2010-12-06 | 2015-01-07 | 株式会社東芝 | 映像検索装置 |
WO2012124252A1 (ja) | 2011-03-14 | 2012-09-20 | 株式会社ニコン | 電子機器、電子機器の制御方法およびプログラム |
US9373021B2 (en) * | 2011-03-25 | 2016-06-21 | Sony Corporation | Method, apparatus and system for outputting a group of images |
JP5791364B2 (ja) * | 2011-05-16 | 2015-10-07 | キヤノン株式会社 | 顔認識装置、顔認識方法、顔認識プログラム、およびそのプログラムを記録した記録媒体 |
CN103814545B (zh) * | 2011-06-30 | 2016-10-26 | 东莞市瑞腾电子科技有限公司 | 手机用户身份认证方法 |
US8838647B2 (en) * | 2011-12-06 | 2014-09-16 | International Business Machines Corporation | Automatic multi-user profile management for media content selection |
WO2013100697A1 (en) * | 2011-12-29 | 2013-07-04 | Intel Corporation | Method, apparatus, and computer-readable recording medium for authenticating a user |
US9030545B2 (en) * | 2011-12-30 | 2015-05-12 | GNR Resound A/S | Systems and methods for determining head related transfer functions |
JP6043706B2 (ja) * | 2013-09-25 | 2016-12-14 | 日本電信電話株式会社 | マッチング処理装置およびマッチング方法 |
CN104639823A (zh) * | 2013-11-14 | 2015-05-20 | 索尼公司 | 图像形成方法、装置以及电子设备 |
JP6215774B2 (ja) * | 2014-05-29 | 2017-10-18 | 富士フイルム株式会社 | 同一人物判定装置ならびにその方法およびその制御プログラム |
JP6149015B2 (ja) * | 2014-09-10 | 2017-06-14 | 富士フイルム株式会社 | 画像処理装置、画像処理方法、プログラムおよび記録媒体 |
CN105678127A (zh) | 2014-11-21 | 2016-06-15 | 阿里巴巴集团控股有限公司 | 一种身份信息的验证方法和装置 |
US9836669B2 (en) | 2016-02-22 | 2017-12-05 | International Business Machines Corporation | Generating a reference digital image based on an indicated time frame and searching for other images using the reference digital image |
US10410045B2 (en) * | 2016-03-23 | 2019-09-10 | Intel Corporation | Automated facial recognition systems and methods |
CN108228871A (zh) * | 2017-07-21 | 2018-06-29 | 北京市商汤科技开发有限公司 | 人脸图像动态入库方法和装置、电子设备、介质、程序 |
JP7204596B2 (ja) * | 2019-06-28 | 2023-01-16 | 富士フイルム株式会社 | 画像処理装置,画像処理方法,画像処理プログラムおよびそのプログラムを格納した記録媒体 |
US20210151169A1 (en) * | 2019-11-18 | 2021-05-20 | Fujifilm Medical Systems U.S.A., Inc. | Methods And Systems For Remote Analysis Of Medical Image Records |
WO2023218441A1 (en) * | 2022-05-11 | 2023-11-16 | Inspekto A.M.V. Ltd. | Optimizing a reference group for visual inspection |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11175718A (ja) | 1997-12-12 | 1999-07-02 | Toshiba Corp | 人物認識装置及びその方法 |
JP2003044858A (ja) | 2001-07-31 | 2003-02-14 | Matsushita Electric Ind Co Ltd | 個人認証装置および方法 |
JP2004054888A (ja) | 2002-05-31 | 2004-02-19 | Omron Corp | 顔照合装置 |
JP2006059200A (ja) | 2004-08-20 | 2006-03-02 | Fuji Photo Film Co Ltd | 顔画像格納システム、顔画像格納方法、及び顔画像格納プログラム |
JP2006092491A (ja) | 2004-09-27 | 2006-04-06 | Glory Ltd | 本人認証装置、本人認証システム、本人認証方法および本人認証プログラム |
JP2007102517A (ja) | 2005-10-05 | 2007-04-19 | Sony Corp | 顔認識装置および方法、係数生成装置および方法、プログラム、並びに記録媒体 |
JP2008129830A (ja) * | 2006-11-21 | 2008-06-05 | Sony Corp | 個人識別装置、個人識別方法、識別用辞書データの更新方法および識別用辞書データの更新プログラム |
JP2008165314A (ja) * | 2006-12-27 | 2008-07-17 | Sony Corp | 個人識別装置および個人識別方法 |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008077446A (ja) * | 2006-09-22 | 2008-04-03 | Fujifilm Corp | 画像検索装置ならびにその制御方法およびその制御プログラム |
JP4561914B2 (ja) * | 2008-09-22 | 2010-10-13 | ソニー株式会社 | 操作入力装置、操作入力方法、プログラム |
KR20100055946A (ko) * | 2008-11-18 | 2010-05-27 | 삼성전자주식회사 | 동영상 썸네일 생성 방법 및 장치 |
-
2009
- 2009-03-13 JP JP2009062039A patent/JP4636190B2/ja active Active
-
2010
- 2010-02-22 EP EP10750494.6A patent/EP2407929B1/en active Active
- 2010-02-22 WO PCT/JP2010/001133 patent/WO2010103735A1/ja active Application Filing
- 2010-02-22 KR KR1020117003462A patent/KR101168546B1/ko active IP Right Grant
- 2010-02-22 CN CN2010800023253A patent/CN102124492B/zh active Active
- 2010-02-22 US US13/059,381 patent/US8396264B2/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11175718A (ja) | 1997-12-12 | 1999-07-02 | Toshiba Corp | 人物認識装置及びその方法 |
JP2003044858A (ja) | 2001-07-31 | 2003-02-14 | Matsushita Electric Ind Co Ltd | 個人認証装置および方法 |
JP2004054888A (ja) | 2002-05-31 | 2004-02-19 | Omron Corp | 顔照合装置 |
JP2006059200A (ja) | 2004-08-20 | 2006-03-02 | Fuji Photo Film Co Ltd | 顔画像格納システム、顔画像格納方法、及び顔画像格納プログラム |
JP2006092491A (ja) | 2004-09-27 | 2006-04-06 | Glory Ltd | 本人認証装置、本人認証システム、本人認証方法および本人認証プログラム |
JP2007102517A (ja) | 2005-10-05 | 2007-04-19 | Sony Corp | 顔認識装置および方法、係数生成装置および方法、プログラム、並びに記録媒体 |
JP2008129830A (ja) * | 2006-11-21 | 2008-06-05 | Sony Corp | 個人識別装置、個人識別方法、識別用辞書データの更新方法および識別用辞書データの更新プログラム |
JP2008165314A (ja) * | 2006-12-27 | 2008-07-17 | Sony Corp | 個人識別装置および個人識別方法 |
Non-Patent Citations (1)
Title |
---|
See also references of EP2407929A4 |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2657908A4 (en) * | 2010-12-21 | 2018-01-17 | Omron Corporation | Image authentication apparatus, image processing system, control program for image authentication apparatus, computer-readable recording medium, and image authentication method |
JP2022015263A (ja) * | 2020-07-08 | 2022-01-21 | 株式会社スクウェア・エニックス | 顔認証プログラム及び顔認証システム |
JP7181256B2 (ja) | 2020-07-08 | 2022-11-30 | 株式会社スクウェア・エニックス | 顔認証プログラム及び顔認証システム |
US11763600B2 (en) | 2020-07-08 | 2023-09-19 | Square Enix Co., Ltd. | Non-transitory computer-readable medium and face authentication system including face authentication program |
Also Published As
Publication number | Publication date |
---|---|
JP2010218059A (ja) | 2010-09-30 |
US8396264B2 (en) | 2013-03-12 |
CN102124492B (zh) | 2013-06-19 |
KR101168546B1 (ko) | 2012-07-27 |
CN102124492A (zh) | 2011-07-13 |
EP2407929B1 (en) | 2020-02-12 |
KR20110036935A (ko) | 2011-04-12 |
EP2407929A4 (en) | 2015-10-21 |
US20110150302A1 (en) | 2011-06-23 |
JP4636190B2 (ja) | 2011-02-23 |
EP2407929A1 (en) | 2012-01-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2010103735A1 (ja) | 顔照合装置、電子機器、顔照合装置の制御方法、および顔照合装置制御プログラム | |
US9098385B2 (en) | Content managing apparatus, content managing method, content managing program, and integrated circuit | |
US9996735B2 (en) | Facial recognition | |
US8650242B2 (en) | Data processing apparatus and data processing method | |
US8810366B2 (en) | Service providing device, service providing method, biometric information authentication server, biometric information authentication method, program, and service providing system | |
CN104008317B (zh) | 认证设备和认证方法 | |
US20150339518A1 (en) | Electronic device, electronic device control method, and computer-readable recording medium having stored thereon electronic device control program | |
WO2007105768A1 (ja) | 顔画像登録装置、顔画像登録方法、顔画像登録プログラム、および記録媒体 | |
JP2009223580A (ja) | 優先対象決定装置、電子機器、優先対象決定方法、プログラム、および記録媒体 | |
US20110179052A1 (en) | Pattern identification apparatus and control method thereof | |
US11816923B2 (en) | Face image candidate determination apparatus for authentication, face image candidate determination method for authentication, program, and recording medium | |
US20140023246A1 (en) | Intelligent biometric identification of a participant associated with a media recording | |
WO2017170384A1 (ja) | 生体データ処理装置、生体データ処理システム、生体データ処理方法、生体データ処理プログラム、生体データ処理プログラムを記憶する記憶媒体 | |
JP2013041416A (ja) | 情報処理装置及び方法、プログラム、並びに情報処理システム | |
US8065528B2 (en) | Authentication device, authentication method, authentication program and computer readable recording medium | |
WO2020152840A1 (ja) | 証明書認定システム、証明書認定方法及びプログラム | |
JP4501531B2 (ja) | 画像認識装置およびプログラム | |
CA2827639C (en) | Facial recognition | |
JP4661980B2 (ja) | 画像認識装置およびプログラム | |
JP7033228B1 (ja) | 認証システム、認証方法および認証プログラム | |
WO2022215328A1 (ja) | 認証システム、処理方法及びプログラム | |
JP2022073249A (ja) | 情報処理システム、情報処理システムの制御方法、およびプログラム | |
JP2024011177A (ja) | ユーザ認証システム、ユーザ認証方法、プログラム | |
JP2022187205A (ja) | 認証装置、登録装置、認証方法、登録方法、コンピュータプログラム、及び記憶媒体 | |
JP2014123166A (ja) | 認証装置および撮像装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201080002325.3 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 10750494 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 20117003462 Country of ref document: KR Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13059381 Country of ref document: US Ref document number: 2010750494 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |