WO2007105768A1 - 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 - Google Patents
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 Download PDFInfo
- Publication number
- WO2007105768A1 WO2007105768A1 PCT/JP2007/055125 JP2007055125W WO2007105768A1 WO 2007105768 A1 WO2007105768 A1 WO 2007105768A1 JP 2007055125 W JP2007055125 W JP 2007055125W WO 2007105768 A1 WO2007105768 A1 WO 2007105768A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- face image
- registration
- face
- image
- representative
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
-
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/28—Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
-
- 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
- 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
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/772—Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
-
- 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
Definitions
- Face image registration device face image registration method, face image registration program, and recording medium
- the present invention relates to a face image registration apparatus, a face image registration method, a face image registration program, and a recording medium that register a face image used for face authentication processing in a dictionary.
- the face recognition system currently has various unsolved problems. One of them is that the accuracy of certification is still not high enough. Even if a face image is shown, the level that can be authenticated has not yet been reached. For example, there is a problem that the person cannot be authenticated because the brightness of the background has changed.
- Patent Document 1 detects a frame in which a face is reflected in a video, extracts the frame power face image, and groups the faces of the same character from all the extracted face images.
- An image search method characterized by extracting a representative face image for each character is disclosed.
- Patent Document 2 discloses a means for inputting image data for performing person authentication in units of frames, a means for detecting a face area from the image data, and a means for detecting a feature quantity from the face area. The feature quantity is compared with a reference value to determine whether the feature quantity is effective. Means for storing a plurality of frames of feature quantities determined to be valid, means for obtaining an authentication feature quantity by performing an operation on the feature quantities of the plurality of frames, and the authentication feature quantity. There has been disclosed a person authentication apparatus characterized by comprising a means for authenticating the person by comparing with a facial feature amount stored in advance.
- Patent Document 3 discloses a face authentication database in which a plurality of pieces of face information obtained by photographing an authentication person's face in a plurality of different environments are registered with pattern names corresponding to the environment.
- An authentication person identification table associating the pattern name registered in the face authentication database with the authentication person photographed with this pattern name, a camera for obtaining face information by photographing the face of the person to be authenticated
- authentication processing means for collating face information photographed by this camera with a plurality of pieces of face information registered in the face authentication database to identify the face information having the highest similarity.
- An authentication system comprising: an authentication principal identifying means for identifying an authentication principal corresponding to the pattern name from the pattern name of the face information identified by the authentication processing means using the authentication principal identification table. Is Ru.
- Patent Document 4 in a face image database new registration device for registering a new person in a person's face image database and face feature database, the positions of a plurality of feature points in the face area of the person in the input image, and An estimation means for estimating the orientation of a person's face, and a frame indicating the predetermined orientation of the person's face according to the face orientation information estimated by the estimation means or the face orientation information input from outside
- the selecting means for selecting image power, the user interface means for receiving and displaying a frame selected by the selecting means and its related information, and the estimating means in the frame selected by the selecting means by the estimating means By measuring the feature quantity of the person's face in the vicinity of the estimated feature points, the person data in the input image is measured.
- a face image database new registration device including a calculation means for calculating a base component is disclosed.
- Patent Document 1 Japanese Patent Publication “JP 2001-167110 (Publication Date: June 22, 2001)”
- Patent Document 2 Japanese Patent Publication “JP-A-6-259534 (Publication Date: September 16, 1994)”
- Patent Document 3 Japanese Patent Publication “JP 2005-115481 Publication (Publication Date: April 28, 2005)”
- Patent Document 4 Japanese Patent Publication “JP 2002-133423 (Publication Date: May 10, 2002)”
- Patent Document 1 merely groups the faces of the same characters, and is not devised in that the grouped faces are registered in a dictionary.
- Patent Document 2 merely improves the measurement accuracy of the feature quantity for valid frame determination, and has no ingenuity in registering the measured feature quantity in the dictionary.
- Patent Document 3 With the technique of Patent Document 3, it is necessary for the user to prepare face information in different environments, and this causes a problem that causes the user to take extra effort.
- Patent Document 4 With the technology of Patent Document 4, only the position of the feature point in each frame of the input material is estimated, leaving the final decision to the user. Therefore, there arises a problem that causes the user to spend unnecessary effort.
- the present invention has been made to solve the above-described problems, and an object of the present invention is to register a face image registration device and a face for registering various face images in a dictionary without requiring a user to take time and effort.
- an image registration method a face image registration program, and a recording medium.
- a face image registration device In order to solve the above problems, a face image registration device according to the present invention
- a face image registration device that extracts a face image representing a person's face from an input moving image and registers it in a dictionary
- Representative face image extracting means for extracting at least one face image that matches a representative condition determined from the above moving image as a representative face image
- a face image extracting means Among the other face images of the person represented by the representative face image, at least one other face image that matches the registration conditions determined in advance is extracted from the moving image as a registration face image.
- a face image registration means for registering the registration face image in the dictionary in association with the representative face image.
- the representative face image extracting means extracts at least one face image that matches a predetermined representative condition from the moving image as the representative face image.
- the “face image” here refers to a still image representing a person's face.
- the “representative condition” is a condition that must be satisfied by a face image most suitable for face authentication among various face images captured in a moving image. For example, the face facing forward satisfies the representative condition. Therefore, the representative face image extracting unit extracts, as the representative face image, for example, a front face image that is an image of a face facing the front from the moving image.
- a moving image is an image of the movement of a person or thing.
- the representative face image extracting means can extract the moving image force from the representative face image such as the front face without any problem.
- the registration face image extracting means matches the registration condition that has been determined in advance among other face images of the person represented by the representative face image.
- One other face image is extracted from the moving image as a registration face image.
- the “registration condition” is a condition that applies to a face image suitable for image recognition. For example, a condition that applies to faces in various states, such as a face facing up, a face with a beard, and a face with an open mouth. Therefore, when the representative face image extracting means extracts, for example, a face facing front as a representative face image, the face image registration device removes the face or beard facing upward for the person represented by the front face image. Extract facial images for registration that match various registration conditions, such as raw faces and faces with open mouth, from the same video.
- the face image registration means registers the representative face image and the registration face image extraction means in association with each other in the dictionary. Therefore, for example, various different face images such as a bearded face image are registered in association with the face image facing the front. Registration here is not necessarily limited to registering the image itself. For example, a feature amount representing the feature of the image may be calculated and registered in the dictionary instead of the image.
- the face image registration device can register various face images for the same person in the dictionary for registering face images. Therefore, there is an effect that the types of face images registered in the dictionary can be more reliably increased.
- the face images required for registration are manually checked, or compared with methods that prepare face images for registration in advance by changing the shooting conditions in various ways. Thus, it is possible to further reduce the labor required for the registration process of the face image. In other words, various face images can be registered in the dictionary without requiring the user to spend time.
- the face image registration method according to the present invention provides:
- a face image registration method for extracting a face image representing a person's face from an input moving image and registering it in a dictionary
- At least one other face image that matches the registration conditions determined in advance is extracted from the moving image as a registration face image.
- a face image registration step of registering the registration face image in the dictionary in association with the representative face image is a face image registration step of registering the registration face image in the dictionary in association with the representative face image.
- this method has the same operational effects as the above-described face image registration device. To do.
- the face image registration means The face image registration means
- the face image registration unit registers the feature amount obtained by quantifying the feature of the registration face image in the dictionary.
- the feature amount of the face image is calculated by specifying the position of each organ of the face and calculating the arrangement state of each position.
- the feature amount calculated in this way is usually smaller in data amount than the image itself.
- the face image registration device has the effect of reducing the data size of the dictionary compared to registering the face image itself in the dictionary.
- the face image registration means The face image registration means
- all of the registration face images extracted for each registration condition are associated with the representative face image and registered in the dictionary.
- the representative face image extraction unit when a plurality of different registration conditions are preliminarily determined, includes at least one registration that matches the registration condition for each registration condition. Extract a facial image. For example, if 10 registration conditions are preliminarily determined, registration face images that apply to all 10 registration conditions are extracted from the moving image.
- the face image registration means extracts the registration face image extracted for each registration condition, All are registered in the dictionary in association with the representative face image. In other words, a plurality of face images that individually match a plurality of registration conditions are comprehensively registered in the dictionary.
- the face image registration device has an effect of further enhancing the degree of fullness of the dictionary.
- the face image registration means The face image registration means
- the face image registration means The face image registration means
- all of the one registration face image extracted for each registration condition is associated with the representative face image and registered in the dictionary.
- the representative face image extracting means when a plurality of different registration conditions are preliminarily determined, for each registration condition, the one registration face that matches the registration condition. Extract images. For example, if 10 registration conditions are preliminarily determined, one registration face image that applies to all 10 registration conditions is extracted from the moving image one by one.
- the face image registration means registers all the registration face images extracted for each registration condition in the dictionary in association with the representative face image. That is, a plurality of face images covering all of the plurality of registration conditions are registered in the dictionary.
- the face image registration device has an effect of further enhancing the degree of fullness of the dictionary. Also, the number of face images registered in the dictionary is limited to one for each registration condition. Therefore, the data size of the dictionary can be made smaller than when a plurality of face images are registered for each registration condition.
- the registration face image extraction means includes: When a plurality of different representative face images are extracted by the representative face image extracting means, it is preferable to extract the registration face image for each representative face image.
- the registration face image extracting unit extracts a registration face image for each extracted representative face image.
- a plurality of different representative face images are a plurality of face images in which each face image represents a different person's face. At this time, the registration face image is extracted for each different person.
- the face image registering means registers the representative face image and the registration face image in the dictionary in association with the representative face image for each person when a plurality of persons are photographed in the moving image. As a result, face images for several different people are accumulated in the dictionary.
- the face image registration device has an effect of further enhancing the degree of fullness of the dictionary.
- the face image registration device further includes:
- the representative face image extracting means extracts the representative face image from the scene
- the registration face image extracting means extracts the registration face image from the scene.
- a scene extraction means extracts a scene from a moving image.
- a scene is a partial moving image in which a plurality of frames in which a specific face continues to exist in a moving image are combined into one.
- the scene extraction means uses face tracking technology to enter a frame with a specific face, continue to exist in a certain number of frames, and finally exit from another frame. Extracted from moving images as one scene.
- the representative face image extracting means extracts a representative face image from the scene. Therefore, a part of the moving image can be a processing target of face image extraction.
- the registration face image extraction means extracts a registration face image from the same scene. As a result, a part of the moving image can be a processing target of face image extraction.
- the face image registration device can keep the moving image targeted for the process of extracting the face image as a part of the whole. As a result, the time required for the face image registration process can be shortened.
- the scene extracting means extracts a plurality of different scenes from the moving image force, and the registration face image extracting means cannot extract the registration face image from a certain scene, and other scenes different from the scene are extracted. Scene power It is preferable to extract the registration face image.
- the scene extracting means also extracts a moving image force from a plurality of different scenes.
- a scene in which a person's face is not shown is divided, and a plurality of divided partial moving images are extracted as different scenes from the moving image column.
- the registration face image extracting means extracts the registration face image from other scenes different from the above scenes. Extract face images. This produces an effect that the types of face images registered in the dictionary can be increased more reliably.
- the registration face image extraction means includes:
- the face image registration means When the recognition reliability of the other registration face image is higher than the recognition reliability of the registered registration face image, the registered face image is replaced with the other registration face image. Is preferred.
- the registration face image extraction unit displays another registration face image that matches the registration condition when the registration face image that matches the registration condition has already been registered in the dictionary. To extract.
- the face image registration means converts the registered face image into another registration image. Replace with a face image.
- the recognition reliability of the face image registered in the dictionary can be further increased.
- the face image registration device has an effect that the recognition accuracy of the face image registered in the dictionary can be further increased.
- the face image registration device may be realized by a computer.
- a face image registration program that realizes the face image registration apparatus in the computer by operating the computer as each of the above means, and a computer-readable recording medium that records the face image registration program. Falls within the scope of the present invention.
- FIG. 1 is a block diagram showing a main configuration of a face image registration system according to an embodiment of the present invention.
- FIG. 2 is a flowchart showing a processing flow when the face image registration device extracts a face image from a moving image and registers it in a face image dictionary.
- FIG. 3 (a) is a diagram showing an example of a moving image processed when the dictionary creation device extracts a face image.
- FIG. 3 (b) is a diagram showing an example of a scene extracted by the scene extraction unit.
- FIG. 3 (c) is a diagram showing an example of a plurality of frames each including various face images and constituting one scene.
- FIG. 4 is a diagram showing an example of a display screen for requesting the user to select a representative face image.
- FIG. 5 is a diagram showing an example of a pickup condition list 50.
- FIG. 6 is a flowchart showing a flow of processing in which a face image extraction unit for registration extracts a scene force from a face image that matches a pickup condition.
- FIG. 7 is a flowchart showing a processing flow when the face image registration unit registers the registration face image extracted by the registration face image extraction unit in the face image dictionary.
- FIG. 8 is a flowchart showing a processing flow when a registration face image extraction unit extracts a face image that matches a pickup condition.
- FIG. 9 is a flowchart showing a processing flow when the registration face image extraction unit updates the pickup condition list.
- FIG. 10 is a flowchart showing a flow of processing when a face image registration device extracts face images of a plurality of persons and registers them in a face image dictionary.
- FIG. 11 is a diagram showing an example of dictionary data stored in a face image dictionary.
- Representative face image extraction unit (Representative face image extraction means)
- Registration face image extraction unit Registration face image extraction means
- Face image registration part (Face image registration means)
- FIG. 1 is a block diagram showing a main configuration of a face image registration system 20 according to an embodiment of the present invention.
- the face image registration system 20 includes a face image registration device 1, an input device 2, a display device 3, and a moving image photographing device 4.
- the face image registration device 1 is a device that registers a face image necessary for face authentication in a dictionary. Details of the configuration and outline will be described later.
- the input device 2 is a device used by the user to input information to the face image registration device 1 or to select information displayed on the display device 3. For example, it is realized as a mouse or a keyboard.
- the display device 3 is a device that displays information and presents it to the user.
- it is realized as a liquid crystal display device or a plasma display device.
- the moving image shooting device 4 is a device for shooting a person or thing in the outside world into a moving image.
- it is realized as a video camera.
- the face image registration device 1 includes a moving image input unit 10, a scene extraction unit 11 (scene Extraction means), representative face image extraction section 12 (representative face image extraction means), registration face image extraction section 1 3 (registration face image extraction means), face image registration section 14 (face image registration means), face image dictionary 15 (dictionary), input processing unit 16, and display processing unit 17.
- scene extraction means scene Extraction means
- representative face image extraction section 12 representative face image extraction means
- registration face image extraction section 1 3 registration face image extraction means
- face image registration section 14 face image registration means
- face image dictionary 15 dictionary
- input processing unit 16 and display processing unit 17.
- the face image registration device 1 extracts a face image suitable for dictionary registration from a moving image and registers it.
- the purpose is not to separate the scenes or to make the face image dallet.
- the purpose is to improve dictionary performance and simplify the performance improvement procedure.
- the face image registration device 1 also extracts a face image suitable for face authentication processing, and also inputs the input moving image power, and automatically registers it in the face image dictionary 15. At this time, for example, a face image of a person designated by the user is selectively registered. The point is to extract face images from moving images.
- the face image registration device can extract various face images of the same person from the moving images.
- various face images can be registered in the face image dictionary 15, so that the bias of the face images registered in the face image dictionary 15 can be reduced.
- the face image dictionary 15 can be created easily and accurately.
- the authentication device can greatly improve the performance of human face authentication.
- FIG. 2 is a flowchart showing the flow of processing when the face image registration device 1 extracts a face image from a moving image and registers it in the face image dictionary 15.
- the user Prior to the processing shown in FIG. 2, the user uses the moving image capturing device 4 to capture a moving image. Thereby, the moving image photographing apparatus 4 inputs the moving image to the face image registration apparatus 1 (step S20).
- the moving image input unit 10 receives the input moving image in the face image registration device 1.
- input The processing unit 16 outputs the received moving image to the scene extraction unit 11.
- the moving image input to the face image registration device 1 is not limited to the one captured by the moving image capturing device 4.
- it may be commercially available video software, or may be a moving image that can be obtained free of charge, such as a broadcast drama or sports program.
- the moving image input to the face image registration device 1 can be anything as long as it captures a human face.
- the face image registration device 1 extracts a face image from a moving image. Accordingly, first, an example of a moving image to be processed by the face image registration device 1 will be described below with reference to FIG.
- FIG. 3 (a) is a diagram showing an example of a moving image processed when the dictionary creating device 1 extracts a face image.
- the moving image 30 is composed of a plurality of continuous frames 31.
- Each frame corresponds to a so-called “frame” in terms of video. That is, each frame 31 is one still image.
- the scene extraction unit 11 extracts at least one scene from the input moving image (step S21).
- the “scene” is a partial moving image in which some frames constituting the moving image are continuously connected.
- the scene extraction unit 11 may extract a single scene from the time when the face of the same person enters the frame until it leaves. In other words, a scene may be anything from the time a person's face appears in the first frame until it disappears.
- the scene extraction unit 11 may extract a scene with a moving image force using a known arbitrary method. For example, it is possible to use so-called face tracking technology that detects and tracks a face that changes continuously little by little. Further, a change in the background may be detected, and the moving image may be segmented at a location where a frame showing a background that matches a predetermined specific condition appears. At this time, the frame groups before and after the separation are extracted as different scenes.
- FIG. 3B is a diagram showing an example of the scene extracted by the scene extraction unit 11.
- the scene extraction unit 11 extracts three scenes, that is, a scene 32, a scene 33, and a scene 34 from one moving image.
- the scene 32 has a structure in which a plurality of continuous frames are connected up to the frame 32a force frame 32b. In each of these frames, the same person's face is continuously photographed.
- the scene 33 is composed of two frames, a frame 33a and a frame 33b. Since the face is not shot in the frame immediately before frame 33a and the frame immediately after frame 33b, the scene extraction unit 11 determines the location constituted by the two frames 33a and 33b in the moving image. Extract as one scene 33.
- the scene 34 has a structure in which three frames, that is, frames 34a to 34c are continuously connected. In these frames, face images of the same person are taken continuously. Therefore, the scene extraction unit 11 extracts a portion constituted by three frames 34a to 34c in the moving image as one scene 34.
- FIG. 3 (c) is a diagram showing an example of a plurality of frames each including various face images and constituting one scene 32.
- FIG. 3 (c) is a diagram showing an example of a plurality of frames each including various face images and constituting one scene 32.
- the scene 32 is composed of a plurality of frames. Each frame captures the face of at least one person.
- the scene 32 also includes frames in which other people's faces are shot at the same time. Note that not all frames are taken of a plurality of human face images. However, each frame has a face with various characteristics such as a face facing forward, a face facing sideways, a face with a beard, and a face with open mouth.
- the scene extraction unit 11 outputs the scene extracted in step S21 to the representative face image extraction unit 12.
- the representative face image extraction unit 12 extracts various face images from the input scene. Search (step S22).
- the representative face image extraction unit 12 extracts a face image that matches the representative condition that has been determined in advance.
- the “representative condition” is a condition that should be satisfied by a face image most suitable for face authentication among various face images captured in a moving image.
- representative conditions are preliminarily stored in a memory (not shown).
- the condition that the face image of the face facing the front satisfies is the representative condition. Therefore, the representative face image extraction unit 12 extracts a face facing forward from the input scene (step S23).
- the representative condition is not necessarily limited to the condition that the face image of the face facing the front satisfies.
- any condition may be used as long as the face image suitable for face authentication should be satisfied.
- the representative face image extracting unit 12 may extract all face images that match the representative condition from the scene as much as possible. Therefore, the faces facing the front of the same person may be extracted separately from each other as different representative face images.
- the user may set the representative conditions as appropriate.
- the representative face image extracting unit 12 extracts a face image that matches the representative condition set by the user from the scene as a representative face image.
- the representative face image extraction unit 12 extracts as many representative face images as possible from the input scene regardless of overlapping of the corresponding persons. Then, all the extracted representative face images are output to the display processing unit 17.
- the face image registration device 1 selects one extracted representative face image (step S24).
- the face image registration device 1 of the present embodiment allows the user to select one representative face image of a person to be registered in the face image dictionary 15.
- the display processing unit 17 causes the display device 3 to display all input representative face images.
- the user confirms the display of the representative face image through the display device 3.
- the user selects one of the displayed representative face images using the input device 2. To do.
- This selection is accepted by the input processing unit 16.
- the input processing unit 16 outputs information about the representative face image selected by the user to the registration face image extracting unit 13.
- the registration face image extraction unit 13 selects the representative face image selected by the user as the processing target.
- FIG. 4 is a diagram showing an example of a display screen for requesting the user to select a representative face image.
- the display processing unit 17 displays the screen 40 shown in FIG.
- the screen 40 includes an area for displaying a plurality of front face images including the front face image 41 and a message box 42.
- a message prompting the user to select a face image is displayed.
- the message “Please select the face image of the person you want to register” is displayed.
- the representative face image extraction unit 12 can also extract different front face images for the same person a plurality of times as the representative face image.
- the display processing unit 17 causes the display device 3 to simultaneously display a plurality of different front face images for the same person.
- the registration face image extraction unit 13 registers the front face image selected by the user in the face image dictionary 15 (step S25). At this time, a unique identifier for identifying the selected front face image may be registered in the face image dictionary 15. Alternatively, the selected front face image itself may be registered. Alternatively, the feature amount of the face image that can be calculated by a method described later may be registered. In any case, the storage area for registering the face image representing the face of the selected person is newly set in the face image dictionary 15.
- the registration face image extraction unit 13 searches various face images of the same person for the person represented by the front face image selected by the user, as well as the input scene power (step S26). At this time, all face images that can be determined to be the same person's face as the extracted front face regardless of the condition are searched.
- the determination method may be a known one. In the present embodiment, the registration face image extraction unit 13 collectively searches for images that can be determined as face images.
- the registration face image extraction unit 13 determines whether or not the searched face image matches a pickup condition (registration condition) that has been determined in advance. To do. Then, the face image determined to be satisfied is extracted as a face image to be registered in the face image dictionary 15 (step S27). At this time, a pickup condition list preliminarily stored in a memory (not shown) is used.
- the “pickup condition list” is a list in which various pickup conditions that should be satisfied by a face image suitable for registration in the face image dictionary 15 are combined.
- FIG. 5 is a diagram showing an example of the pickup condition list 50.
- a main condition As shown in this figure, in the pickup condition list 50, a main condition, a sub condition A, a sub condition B, and a sub condition C are defined.
- the main condition is a condition that roughly classifies the features of the face image.
- the face direction, brightness, facial expression, and accessory are defined as main conditions of the pickup condition.
- the force sub-condition A is a further definition of the main condition.
- “Face” is “Up”, “Down”, “Right”, “Left”, “Brightness” is “Flash”, “Background ratio”, “Face” is “Eye”, “Mouth”, In the “accompaniment”, “glasses” and “ ⁇ ” are defined as sub-conditions A, respectively.
- the force subcondition B is obtained by further defining the subcondition A. For example, “Up” is “0-30 degrees”, “Flash” is “Yes”, “No”, “Background ratio” is “Bright”, “Dark”, “Eye” is “Open”, “Open” Sub-conditions B such as “Closed” are defined.
- the force subcondition C is obtained by further defining the subcondition B. For example, “0-30 degrees” is defined as “5-degree increments”.
- the level of the sub-condition to be set varies depending on the pickup condition.
- the registration face image extraction unit 13 generates various pickup conditions by appropriately combining the main conditions defined in the pickup condition list 50 with the sub-conditions belonging to the main conditions. For example, when talking about “face orientation,” one “pick face up 5 degrees” Treat as a condition.
- the registration face image extraction unit 13 sets at least one face image that matches at least one pickup condition defined in the pickup condition list 50 as a registration face image. Extract scene power.
- the registration face image extraction unit 13 outputs the registration face image from which the scene power is also extracted to the face image registration unit 14.
- the face image registration unit 14 registers the input registration face image in the face image dictionary 15 (step S28). Details of this will be described later.
- the registration face image extraction unit 13 determines whether or not the scene power has also been extracted from the face image (step S29). At this time, the registration face image extraction unit 13
- Condition B A certain number of scenes were processed
- Condition E A certain number of moving images were processed
- Condition F The extraction process starts and the force has reached a certain time
- step S27 the registration face image extraction unit 13 determines whether or not the face image retrieved from the scene matches the pickup condition, and selects one new face image. Then, it is determined whether or not the force matches the pickup condition described above.
- the registration face image extraction unit 13 performs the pickup condition as the face image power registration face image searched by the scene camera. Is repeatedly determined for each face image. [Details of condition judgment and extraction processing]
- FIG. 6 is a flowchart showing a flow of processing in which the registration face image extraction unit 13 extracts a scene force from a face image that matches the pickup condition.
- the registration face image extraction unit 13 of the present embodiment extracts a face image that matches the pickup conditions from the scene using, for example, a technique disclosed in JP-A-2005-49854. Therefore, a face image (template) corresponding to each pickup condition is not required in advance. That is, if a condition defined as a pickup condition is given, a face image that matches the condition can be automatically extracted for each person.
- the registration face image extraction unit 13 selects one unregistered pickup condition (step S60).
- the “unregistered pickup condition” here refers to the condition that the face image satisfies the face image registered in the face image dictionary 15 among the face images of the person represented by the selected representative face image. That is. For example, when the face image power matching the face orientation pickup condition is not registered in the face image dictionary 15, the pickup condition for the face orientation is selected from the pickup condition list 50.
- the registration face image extraction unit 13 executes processing according to the selected condition. In the example of Fig. 6, jump to the block for processing the selected condition.
- a flag indicating that a face image that matches the pickup condition is registered in the face image dictionary 15 is set in the pickup condition list 50 for each pickup condition. Therefore, the registration face image extraction unit 13 should select a condition for which the flag is not set among the pickup conditions defined in the pickup condition list 50 as an unregistered pickup condition.
- the registration face image extraction unit 13 calculates the orientation of the face represented by the face image to be processed. For example, the direction and numerical value such as 40 degrees upward are calculated.
- the registration face image extraction unit 13 performs, for example, • Triangle shape connecting eyes and mouth
- the registration face image extraction unit 13 determines whether or not the calculated face orientation satisfies the selected pickup condition (step S62). When it is determined that the image is satisfied (Yes), the registration face image extraction unit 13 updates the pickup condition (step S69). Details of the update will be described later.
- step S62 When it is determined in step S62 that the face image force does not match the pick-up condition for the face orientation (No), the registration face image extraction unit 13 next calculates the brightness of the face image. (Step S63).
- the brightness of the face image is selected. Since these calculation methods have already been established, detailed description thereof is omitted.
- the registration face image extraction unit 13 determines whether the calculated brightness satisfies the selected pickup condition or not (step S64). When it is determined that the condition is satisfied (Yes), the registration face image extraction unit 13 updates the pickup conditions (step S69). Details of the update will be described later.
- step S64 when it is determined that the brightness power that the face image power is calculated does not match the pick-up condition for brightness (No), the registration face image extraction unit 13 then calculates the facial image expression ( Step S65).
- the registration face image extraction unit 13 performs, for example, 'Edge amount from eye center
- Registration face image extraction unit 13 determines whether the calculated facial expression satisfies the pickup condition for the facial expression or not (step S68). When it is determined that the image is satisfied (Yes), the registration face image extraction unit 13 updates the pickup condition (step S69). Details of the update will be described later.
- step S66 when it is determined that the facial expression power also does not match the pickup condition for the facial expression (No), the registration facial image extraction unit 13 calculates the presence or absence of an accessory of the facial image (step S67).
- the registration face image extraction unit 13 performs, for example,
- the registration face image extraction unit 13 determines whether or not the calculated accessory is a color that satisfies the pickup condition for the accessory (step S68). When it is determined that the condition is satisfied (Yes), the registration face image extraction unit 13 updates the pickup condition (step S69). Details will be described later.
- the registration facial image extraction unit 13 if it is determined that the conditions for calculating facial image power satisfy the selected pickup conditions (Yes), the registration facial image extraction unit 13 will Then, the pickup condition list 50 is updated (step S69). Specifically, the selected pick As for the cleanup condition, a flag indicating that a matching face image is registered in the face image dictionary 15 is set in the pickup condition list 50. At this time, the registration face image extraction unit 13 will not select the condition registered this time again when processing the next face image.
- the face image registration device 1 automatically generates the face image dictionary 15 most suitable for face authentication of a person when the user selects a person to be registered in the face image dictionary 15. That is, various different face images are automatically extracted from the moving image (scene) and registered in the face image dictionary 15. Therefore, the labor required for the face image registration process can be reduced as compared with the prior art.
- the face image registration device 1 may leave the selection of the registration face image to be registered in the face image dictionary 15 to the user. At this time, the face image registration device 1 causes the display device 3 to display a list of various searched face images. As a result, the face image selected by the user is registered in the face image dictionary 15. In this case, since the determination by the user is added to the selection of the registered face image, the accuracy of the face image dictionary 15 can be further increased.
- the registration face image extraction unit 13 updates the pickup condition list 50 each time a face image that matches the pickup condition is extracted.
- the registration face image extraction unit 13 may first store all face images that match the pickup conditions in the memory. At this time, the registration face image extraction unit 13 may select one face image to be registered in the face image dictionary 15 at random after completing the face image extraction. Or let the user select the face image to register.
- FIG. 7 is a flowchart showing a processing flow when the face image registration unit 14 registers the registration face image extracted by the registration face image extraction unit 13 in the face image dictionary 15.
- the face image registration unit 14 first applies the face represented by the face image. Then, the positions of various organs are detected (step S70). Specifically, the center of the eye, the corner of the eye, the head of the eye, the mouth, both ends of the mouth, both ends of the eyebrows, etc. are detected. Since the organ detection method is known, a detailed description thereof will be omitted.
- the organs detected by the face image registration unit 14 are not limited to these, and may be any facial organs.
- the number to be detected is not particularly limited. The number of organs to be detected and how many are detected is predetermined in the memory, not shown in the face image registration apparatus 1.
- the face image registration unit 14 calculates the feature amount of the face image using the detected information on each organ position (step S71). This digitizes the features of the face image. The numerical method is already established and will not be described in detail here.
- the face image registration unit 14 stores the calculated feature value in association with the representative face image already registered in the face image dictionary 15 (step S72).
- the face image registration unit 14 registers data with a smaller amount in the face image dictionary 15 as compared to registering the face image itself in the face image dictionary 15. Therefore, the data size of the generated face image dictionary 15 can be made smaller.
- the face image registration device 1 extracts a face image that matches a pickup condition that is not registered in the face image dictionary 15 and registers it in the face image dictionary 15.
- the face image registration device 1 may extract a face image that matches the pickup condition regardless of whether or not the registered power is used.
- the registration face image extraction unit 13 determines which one of the face image registered in the face image dictionary 15 and the newly extracted registration face image is suitable for face authentication. If it is determined that the newly extracted face image is suitable for face recognition, the existing face image is replaced with the newly extracted face image.
- FIG. 8 is a flowchart showing a processing flow when the registration face image extraction unit 13 extracts a face image that matches the pickup condition.
- the registration face image extraction unit 13 selects one face image to be processed.
- the face image registration unit 14 calculates the face orientation for the selected face image (step S80). Since the face orientation calculation processing has been described above, a description thereof will be omitted.
- the registration face image extraction unit 13 determines whether or not the calculated face orientation is a force that matches at least one of the pickup conditions for the face orientation defined in the pickup condition list 50 (step S81). ).
- step S81 If it is determined in step S81 that the calculated face orientation matches the pickup condition (Yes), the registration face image extraction unit 13 updates the pickup condition list 50. Details of this processing will be described later.
- step S81 When it is determined in step S81 that the calculated face orientation does not match the pickup condition (No), the registration face image extraction unit 13 next calculates the brightness of the selected face image. (Step S82). Since the brightness calculation process has been described above, a description thereof will be omitted. The registration face image extraction unit 13 determines whether or not the calculated brightness matches at least one of the pickup conditions for brightness defined in the pickup condition list 50 (step S83).
- step S83 When it is determined in step S83 that the calculated brightness matches the pickup condition (Yes), the registration face image extraction unit 13 updates the pickup condition list 50. Details of this processing will be described later.
- step S83 When it is determined in step S83 that the calculated face orientation does not match the pickup condition (No), the registration face image extraction unit 13 next calculates the facial expression of the selected face image (Ste S82). Since the facial expression calculation process has been described above, a description thereof will be omitted. The registration face image extraction unit 13 determines whether the calculated facial expression is a force that matches at least one of the pickup conditions for the facial expression defined in the pickup condition list 50 (step S85).
- step S83 when it is determined that the calculated brightness matches the pickup condition (Yes), the registration face image extraction unit 13 updates the pickup condition list 50. Details of this processing will be described later.
- step S85 it is determined that the calculated facial expression does not match the pickup conditions. (No), the registration face image extraction unit 13 then calculates an accessory for the selected face image (step S86). Since the facial expression calculation process has been described above, a description thereof will be omitted. The registration face image extraction unit 13 determines whether or not the calculated accessory matches at least one of the pickup conditions for the accessory defined in the pickup condition list 50 (step S87).
- the registration face image extraction unit 13 updates the pickup condition list 50 when the face image matches the pickup condition (step S88).
- FIG. 9 is a flowchart showing the flow of processing when the registration face image extraction unit 13 updates the pickup condition list.
- the registration face image extraction unit 13 determines whether or not the pick-up condition determined to match the selected face image is a force registered in the pickup condition list 50 (step S 90). Specifically, it is determined whether or not a flag for the pickup condition is set in the pickup condition list 50.
- step S90 When the determination result in step S90 is “false” (No), the pickup condition list 50 is updated. Specifically, a flag is set for the pickup condition determined to match the face image.
- step S90 determines whether or not the newly recognized recognition reliability is greater than the stored recognition reliability (STEP).
- the “recognition reliability” is a numerical value indicating how well the pickup condition is satisfied.
- the registration face image extraction unit 13 stores the calculated numerical value in the pickup condition list 50 in association with the pickup condition, such as 5 degrees and 8 degrees. Alternatively, it may be registered in the face image dictionary 15 in association with the registration face image.
- registration face image extraction unit 13 updates pickup condition list 50. Since the flag indicating that the face image matching the pickup condition is registered in the face image dictionary 15 is already set in the pickup condition list 50, it is left as it is. On the other hand, the recognition reliability stored so far is replaced with the recognition reliability of the newly extracted face image. Thereafter, the face image registration unit 14 replaces the face image registered in the face image dictionary 15 with the face image newly extracted by the registration face image extraction unit 13.
- the face image registration device 1 extracts a registration face image for one person selected by the user from a plurality of representative face images and registers it in the face image dictionary 15.
- the face image registration device 1 extracts a face image suitable for registration from one moving image or one scene
- the face image registration device 1 extracts face images for registration for a plurality of persons, and continues to the face image dictionary 15. You may register.
- FIG. 10 is a flowchart showing the flow of processing when the face image registration device 1 extracts face images for a plurality of persons and registers them in the face image dictionary 15.
- Steps S100 to S109 shown in FIG. 10 are essentially the same as steps S20 and S29, respectively. Therefore, the description is omitted for details.
- step S104 the display processing unit 17 sets a flag indicating that the person selected by the user has been selected in a memory (not shown).
- step S109 it is determined that the extraction of the registration face image for a person has ended. If it is determined (Yes), the registration face image extraction unit 13 is a person who has not registered in the face image dictionary 15 among the representative face images presented to the user, that is, has not yet been processed. Whether or not force ⁇ is determined (step S110). Specifically, it is determined whether there is a face image for which the flag is not set.
- step S110 determines whether the result of determination in step S110 is "true" (Yes)
- the processing in Fig. 10 returns to step S104, and face image registration device 1 displays the representative face image displayed on display device 3.
- the user is allowed to select an unregistered one.
- the selected face image may be displayed in a single scale to clearly indicate that it has been selected.
- the user selects one unselected representative face image.
- the registration face image extraction unit 13 extracts a registration face image for the person represented by the newly selected representative face image from the scene.
- the face image registration unit 14 registers the registration face image selected by the registration face image extraction unit 13 in the face image dictionary 15.
- the face image registration device 1 repeats the above processing until there is no more representative face image that can be selected or until the user explicitly instructs the end of registration.
- the registration face image extraction unit 13 extracts a registration face image for each representative face image when the representative face image extraction unit 12 extracts a plurality of different representative face images.
- a plurality of different representative face images are a plurality of face images in which each face image represents a face of a different person. At this time, the registration face image extraction unit 13 extracts a registration face image for each different person.
- the face image registration unit 14 associates the representative face image with the registration face image and registers them in the face image dictionary 15 for each person when a plurality of persons are photographed in the moving image. To do. Therefore, face images about a plurality of different persons are accumulated in the face image dictionary 15.
- the face image registration device 1 can further enhance the degree of fulfillment of the face image dictionary 15.
- FIG. 11 is a diagram showing an example of dictionary data stored in the face image dictionary 15.
- the dictionary data generated by the face image registration device 1 includes header data H, feature quantity data dl, feature quantity data d2,..., Feature quantity data dn.
- the header data H is composed of various definition data related to the entire dictionary data, such as version information of the face image dictionary 15 and the number of people registered in the face image dictionary 15.
- the feature values are integrated for each registration face image.
- the face image registration device 1 does not extract a matching face image by comparing a template image that matches a specific pickup condition with a face image in the moving image. Instead, a face image that matches each registered pickup condition is extracted, and the feature values calculated from the extracted face image data are combined into one to generate a face image dictionary 15. Therefore, it is possible to create a face image dictionary 15 that can handle a wider range than the registered conditions.
- Various authentication devices use the face image dictionary 15 generated by the face image registration device 1 as described above.
- the authentication device extracts a photograph of a specific person from a large number of given still images.
- the authentication accuracy can be increased by using the face image dictionary 15 at the time of personal authentication.
- scenes where a specific person appears can be extracted from the input video.
- the face image dictionary 15 with higher accuracy than in the past can be used, authentication of the face image can be executed with higher accuracy than in the past.
- the registration face image extraction unit 13 can search another scene if there is no registration face image necessary for one scene.
- the scene extraction unit 11 extracts a plurality of different scenes from the moving image.
- a scene in which a person's face is not shown is separated.
- the registration face image extraction unit 13 extracts a registration face from other scenes different from the scene. Extract images.
- the types of face images registered in the face image dictionary 15 can be increased more reliably.
- the representative face image extracting unit 12 can also extract face images of the same size for each person when extracting the representative face image with scene power. By doing so, the authentication accuracy of the representative face image registered in the face image dictionary 15 can be stabilized.
- the registration face image extraction unit 13 can also extract face images of the same size for each pickup condition when extracting the scene power registration. In this way, the authentication accuracy of the registration face image registered in the face image dictionary 15 can be stabilized.
- the face image registration device 1 may automatically select the representative face image extracted by the representative face image extraction unit 12 without allowing the user to select it. At this time, the registration face image extraction unit 13 selects any one of the extracted representative face images, and extracts a registration face image for the person represented by the selected representative face image from the scene. According to this method, the face image dictionary 15 can be generated completely automatically, and the user's hand is not troubled at all.
- each block included in the face image registration device 1 may be configured by hardware logic.
- a CPU Central Processing Unit
- the face image registration device 1 executes a command of a control program that realizes each function.
- CPU Read Only Memory
- RAM Random Access Memory
- memory that stores the control program and various data
- a device recording medium
- the object of the present invention can also be achieved by a predetermined recording medium.
- the program code (execution format program, intermediate code program, source program) of the control program of the face image registration apparatus 1 which is software for realizing the above-described functions is recorded so as to be readable by a computer. Just do it.
- This recording medium is supplied to the face image registration device 1.
- the face image registration device 1 (or CCU or MPU) as a computer may read and execute the program code recorded on the supplied recording medium.
- the recording medium that supplies the program code to the face image registration device 1 is not limited to a specific structure or type. That is, this recording medium includes, for example, a tape system such as a magnetic tape or a cassette tape, a magnetic disk such as a floppy (registered trademark) disk / hard disk, or an optical disk such as a CD-ROM / MO / MD / DVD / CD-R. Disk system, IC card (including memory card) Z optical card or other card system, or mask ROMZ EPROMZEEPROMZ flash ROM or other semiconductor memory system
- the object of the present invention can be achieved.
- the above program code is supplied to the face image registration device 1 via the communication network.
- the communication network is not limited to a specific type or form as long as it can supply the program code to the face image registration device 1.
- the Internet, intranet, extranet, LAN, ISDN, VAN, CATV communication network, virtual private network, telephone line network, mobile communication network, satellite communication network, etc. may be used.
- the transmission medium constituting this communication network is not limited to a specific configuration or type as long as it is an arbitrary medium capable of transmitting the program code.
- 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.
- the face image registration apparatus registers the face image for registration extracted from the moving image in the dictionary in association with the representative face image extracted from the moving image. Because it is equipped with means, it is possible to register various face images in the dictionary without requiring the user to spend time and effort.
- the present invention can be widely used as a face image registration device for generating a face image dictionary used by various authentication devices for authenticating a person's face image.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Library & Information Science (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Collating Specific Patterns (AREA)
- Processing Or Creating Images (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP07738595A EP1998286A4 (en) | 2006-03-15 | 2007-03-14 | FACE IMAGE RECORDING DEVICE, METHOD, AND RECORDING MEDIUM, AND RECORDING MEDIUM |
US12/282,631 US8848985B2 (en) | 2006-03-15 | 2007-03-14 | Face-image registration device, face-image registration method, face-image registration program, and storage medium |
CN2007800090665A CN101401126B (zh) | 2006-03-15 | 2007-03-14 | 面部图像注册装置、面部图像注册方法 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2006-071851 | 2006-03-15 | ||
JP2006071851A JP4725377B2 (ja) | 2006-03-15 | 2006-03-15 | 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 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2007105768A1 true WO2007105768A1 (ja) | 2007-09-20 |
Family
ID=38509584
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2007/055125 WO2007105768A1 (ja) | 2006-03-15 | 2007-03-14 | 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 |
Country Status (6)
Country | Link |
---|---|
US (1) | US8848985B2 (ja) |
EP (1) | EP1998286A4 (ja) |
JP (1) | JP4725377B2 (ja) |
KR (1) | KR100996066B1 (ja) |
CN (1) | CN101401126B (ja) |
WO (1) | WO2007105768A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017047090A1 (ja) * | 2015-09-18 | 2017-03-23 | 日本電気株式会社 | 指紋撮像システム、指紋撮像装置、画像処理装置、指紋撮像方法、及び記録媒体 |
CN109784231A (zh) * | 2018-12-28 | 2019-05-21 | 广东中安金狮科创有限公司 | 安防信息管理方法、装置及存储介质 |
Families Citing this family (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4697106B2 (ja) * | 2006-09-25 | 2011-06-08 | ソニー株式会社 | 画像処理装置および方法、並びにプログラム |
JP4941124B2 (ja) * | 2007-06-25 | 2012-05-30 | オムロン株式会社 | 監視システムおよび方法、情報処理装置、並びにプログラム |
JP4983558B2 (ja) * | 2007-11-13 | 2012-07-25 | マツダ株式会社 | 車両用運転支援装置 |
KR100840023B1 (ko) * | 2007-11-13 | 2008-06-20 | (주)올라웍스 | 셀프 촬영 시 얼굴의 구도를 잡도록 도와주는 방법 및시스템 |
CN101465961B (zh) * | 2007-12-19 | 2013-10-30 | 神基科技股份有限公司 | 以特征影像辨识控制快门的摄像装置及方法 |
JP2010027035A (ja) | 2008-06-16 | 2010-02-04 | Canon Inc | 個人認証装置及び個人認証方法 |
JP2014116033A (ja) * | 2008-06-16 | 2014-06-26 | Canon Inc | 個人認証装置及び個人認証方法、カメラ、プログラム並びに記憶媒体 |
US9405995B2 (en) * | 2008-07-14 | 2016-08-02 | Lockheed Martin Corporation | Method and apparatus for facial identification |
JP5371083B2 (ja) * | 2008-09-16 | 2013-12-18 | Kddi株式会社 | 顔識別特徴量登録装置、顔識別特徴量登録方法、顔識別特徴量登録プログラム及び記録媒体 |
JP5164767B2 (ja) * | 2008-09-25 | 2013-03-21 | キヤノン株式会社 | 画像処理装置、画像処理装置の制御方法及びプログラム |
DE102009023306A1 (de) * | 2009-05-29 | 2010-12-09 | Siemens Aktiengesellschaft | Verfahren zum rechnergestützten Erkennen der Identität eines Objekts |
JP5438419B2 (ja) * | 2009-07-29 | 2014-03-12 | 富士フイルム株式会社 | 人物照合装置及び人物照合方法 |
JP5881932B2 (ja) * | 2009-09-30 | 2016-03-09 | カシオ計算機株式会社 | 顔画像データ生成装置、顔画像データ生成方法及びプログラム |
JP4844670B2 (ja) * | 2009-11-13 | 2011-12-28 | 日本ビクター株式会社 | 映像処理装置および映像処理方法 |
KR101185243B1 (ko) * | 2009-12-15 | 2012-09-21 | 삼성전자주식회사 | 얼굴 인식을 위한 복수의 얼굴 이미지를 등록하기 위한 장치 및 방법 |
KR101673032B1 (ko) * | 2010-01-25 | 2016-11-04 | 엘지전자 주식회사 | 화상 통신 방법 및 그를 이용한 디지털 tv |
KR101214732B1 (ko) * | 2010-03-09 | 2012-12-21 | 삼성전자주식회사 | 복수의 얼굴 영상을 이용한 얼굴 인식 장치 및 방법 |
JP5434708B2 (ja) * | 2010-03-15 | 2014-03-05 | オムロン株式会社 | 照合装置、デジタル画像処理システム、照合装置制御プログラム、コンピュータ読み取り可能な記録媒体、および照合装置の制御方法 |
JP5479198B2 (ja) * | 2010-04-23 | 2014-04-23 | 株式会社東芝 | 電子機器及び画像処理プログラム |
WO2012029150A1 (ja) * | 2010-09-01 | 2012-03-08 | 富士通株式会社 | 生体認証システム、生体認証方法、及びプログラム |
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 |
JP5801601B2 (ja) | 2011-05-10 | 2015-10-28 | キヤノン株式会社 | 画像認識装置、画像認識装置の制御方法、およびプログラム |
JP5311239B2 (ja) | 2011-06-17 | 2013-10-09 | カシオ計算機株式会社 | 売上データ処理装置及びプログラム |
JP5755046B2 (ja) | 2011-06-22 | 2015-07-29 | キヤノン株式会社 | 画像認識装置、画像認識方法及びプログラム |
US9875431B2 (en) | 2013-06-28 | 2018-01-23 | Nec Corporation | Training data generating device, method, and program, and crowd state recognition device, method, and program |
JP5673743B2 (ja) * | 2013-06-28 | 2015-02-18 | カシオ計算機株式会社 | データ処理装置及びプログラム |
JP6018029B2 (ja) * | 2013-09-26 | 2016-11-02 | 富士フイルム株式会社 | 撮像画像の主要顔画像決定装置ならびにその制御方法およびその制御プログラム |
JP6287527B2 (ja) * | 2014-04-17 | 2018-03-07 | 富士通株式会社 | 情報処理装置、方法及びプログラム |
JP6754574B2 (ja) * | 2016-01-18 | 2020-09-16 | 株式会社日立情報通信エンジニアリング | 移動体計測システム、及び計測対象の領域における人物の数を特定する方法 |
WO2017146161A1 (ja) * | 2016-02-26 | 2017-08-31 | 日本電気株式会社 | 顔照合システム、顔照合装置、顔照合方法、及び記録媒体 |
JP2017021844A (ja) * | 2016-09-29 | 2017-01-26 | 富士フイルム株式会社 | 撮像画像の主要顔画像決定装置ならびにその制御方法およびその制御プログラム |
US10621419B2 (en) * | 2016-12-06 | 2020-04-14 | Robert William Kocher | Method and system for increasing biometric acceptance rates and reducing false accept rates and false rates |
JP7104291B2 (ja) * | 2017-06-30 | 2022-07-21 | キヤノンマーケティングジャパン株式会社 | 情報処理システム、情報処理方法、プログラム |
JP7299543B2 (ja) * | 2017-06-30 | 2023-06-28 | キヤノンマーケティングジャパン株式会社 | 情報処理システム、情報処理方法、プログラム |
JP6769475B2 (ja) | 2018-12-04 | 2020-10-14 | 日本電気株式会社 | 情報処理システム、認証対象の管理方法、及びプログラム |
JP7054847B2 (ja) * | 2019-03-04 | 2022-04-15 | パナソニックIpマネジメント株式会社 | 顔認証登録装置および顔認証登録方法 |
JP7327571B2 (ja) | 2020-09-24 | 2023-08-16 | 日本電気株式会社 | 情報処理システム、端末装置、認証対象の管理方法、及びプログラム |
JP7067593B2 (ja) * | 2020-09-24 | 2022-05-16 | 日本電気株式会社 | 情報処理システム、認証対象の管理方法、及びプログラム |
WO2023032157A1 (ja) * | 2021-09-03 | 2023-03-09 | 日本電気株式会社 | 情報処理システム、情報処理装置、情報処理方法、及び記録媒体 |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06259534A (ja) | 1993-03-03 | 1994-09-16 | Toshiba Corp | 人物認証装置 |
JPH10232934A (ja) * | 1997-02-18 | 1998-09-02 | Toshiba Corp | 顔画像登録装置及びその方法 |
JPH11167632A (ja) * | 1997-12-02 | 1999-06-22 | Omron Corp | 本人特定装置 |
JP2001167110A (ja) | 1999-12-08 | 2001-06-22 | Matsushita Electric Ind Co Ltd | 画像検索方法及びその装置 |
JP2002133423A (ja) | 2000-10-19 | 2002-05-10 | Nippon Hoso Kyokai <Nhk> | 顔画像データベース新規登録装置および記録媒体 |
JP2002269563A (ja) * | 2001-03-08 | 2002-09-20 | Matsushita Electric Ind Co Ltd | 顔画像照合システム |
JP2005049854A (ja) | 2003-07-15 | 2005-02-24 | Omron Corp | 対象決定装置及び撮像装置 |
JP2005115481A (ja) | 2003-10-03 | 2005-04-28 | Toshiba Social Automation Systems Co Ltd | 認証システム及びコンピュータ読み取り可能な記憶媒体 |
JP2006031387A (ja) * | 2004-07-15 | 2006-02-02 | Yamaha Motor Co Ltd | 画像認識装置、画像認識方法、画像認識プログラムおよび画像認識プログラムを記録した記録媒体 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10257436A (ja) * | 1997-03-10 | 1998-09-25 | Atsushi Matsushita | 動画像の自動階層構造化方法及びこれを用いたブラウジング方法 |
TWI278782B (en) * | 2001-08-24 | 2007-04-11 | Toshiba Corp | Personal recognition apparatus |
US7587068B1 (en) * | 2004-01-22 | 2009-09-08 | Fotonation Vision Limited | Classification database for consumer digital images |
JP4230870B2 (ja) * | 2003-09-25 | 2009-02-25 | 富士フイルム株式会社 | 動画記録装置、動画記録方法、及びプログラム |
JP4189811B2 (ja) * | 2003-09-30 | 2008-12-03 | 日本電気株式会社 | 画像検出装置、画像検出方法及び画像検出用プログラム |
JP4230959B2 (ja) * | 2004-05-19 | 2009-02-25 | 株式会社東芝 | メディアデータ再生装置、メディアデータ再生システム、メディアデータ再生プログラムおよび遠隔操作プログラム |
-
2006
- 2006-03-15 JP JP2006071851A patent/JP4725377B2/ja not_active Expired - Fee Related
-
2007
- 2007-03-14 EP EP07738595A patent/EP1998286A4/en not_active Ceased
- 2007-03-14 US US12/282,631 patent/US8848985B2/en active Active
- 2007-03-14 CN CN2007800090665A patent/CN101401126B/zh active Active
- 2007-03-14 WO PCT/JP2007/055125 patent/WO2007105768A1/ja active Application Filing
- 2007-03-14 KR KR1020087023721A patent/KR100996066B1/ko active IP Right Grant
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06259534A (ja) | 1993-03-03 | 1994-09-16 | Toshiba Corp | 人物認証装置 |
JPH10232934A (ja) * | 1997-02-18 | 1998-09-02 | Toshiba Corp | 顔画像登録装置及びその方法 |
JPH11167632A (ja) * | 1997-12-02 | 1999-06-22 | Omron Corp | 本人特定装置 |
JP2001167110A (ja) | 1999-12-08 | 2001-06-22 | Matsushita Electric Ind Co Ltd | 画像検索方法及びその装置 |
JP2002133423A (ja) | 2000-10-19 | 2002-05-10 | Nippon Hoso Kyokai <Nhk> | 顔画像データベース新規登録装置および記録媒体 |
JP2002269563A (ja) * | 2001-03-08 | 2002-09-20 | Matsushita Electric Ind Co Ltd | 顔画像照合システム |
JP2005049854A (ja) | 2003-07-15 | 2005-02-24 | Omron Corp | 対象決定装置及び撮像装置 |
JP2005115481A (ja) | 2003-10-03 | 2005-04-28 | Toshiba Social Automation Systems Co Ltd | 認証システム及びコンピュータ読み取り可能な記憶媒体 |
JP2006031387A (ja) * | 2004-07-15 | 2006-02-02 | Yamaha Motor Co Ltd | 画像認識装置、画像認識方法、画像認識プログラムおよび画像認識プログラムを記録した記録媒体 |
Non-Patent Citations (1)
Title |
---|
See also references of EP1998286A4 |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017047090A1 (ja) * | 2015-09-18 | 2017-03-23 | 日本電気株式会社 | 指紋撮像システム、指紋撮像装置、画像処理装置、指紋撮像方法、及び記録媒体 |
US10692201B2 (en) | 2015-09-18 | 2020-06-23 | Nec Corporation | Fingerprint capture system, fingerprint capture device, image processing apparatus, fingerprint capture method, and storage medium |
US10699399B2 (en) | 2015-09-18 | 2020-06-30 | Nec Corporation | Fingerprint capture system, fingerprint capture device, image processing apparatus, fingerprint capture method, and storage medium |
US10984514B2 (en) | 2015-09-18 | 2021-04-20 | Nec Corporation | Fingerprint capture system, fingerprint capture device, image processing apparatus, fingerprint capture method, and storage medium |
US11004185B2 (en) | 2015-09-18 | 2021-05-11 | Nec Corporation | Fingerprint capture system, fingerprint capture device, image processing apparatus, fingerprint capture method, and storage medium |
US11501422B2 (en) | 2015-09-18 | 2022-11-15 | Nec Corporation | Fingerprint capture system, fingerprint capture device, image processing apparatus, fingerprint capture method, and storage medium |
US11844609B2 (en) | 2015-09-18 | 2023-12-19 | Nec Corporation | Fingerprint capture system, fingerprint capture device, image processing apparatus, fingerprint capture method, and storage medium |
CN109784231A (zh) * | 2018-12-28 | 2019-05-21 | 广东中安金狮科创有限公司 | 安防信息管理方法、装置及存储介质 |
CN109784231B (zh) * | 2018-12-28 | 2023-07-25 | 广东中安金狮科创有限公司 | 安防信息管理方法、装置及存储介质 |
Also Published As
Publication number | Publication date |
---|---|
JP2007249588A (ja) | 2007-09-27 |
KR20080104034A (ko) | 2008-11-28 |
US8848985B2 (en) | 2014-09-30 |
EP1998286A4 (en) | 2011-02-02 |
JP4725377B2 (ja) | 2011-07-13 |
KR100996066B1 (ko) | 2010-11-25 |
CN101401126B (zh) | 2011-10-26 |
US20090060295A1 (en) | 2009-03-05 |
CN101401126A (zh) | 2009-04-01 |
EP1998286A1 (en) | 2008-12-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP4725377B2 (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 | |
WO2019128508A1 (zh) | 图像处理方法、装置、存储介质及电子设备 | |
US8285006B2 (en) | Human face recognition and user interface system for digital camera and video camera | |
JP4497236B2 (ja) | 検出用情報登録装置、電子機器、検出用情報登録装置の制御方法、電子機器の制御方法、検出用情報登録装置制御プログラム、電子機器の制御プログラム | |
EP2339536B1 (en) | Image processing system, image processing apparatus, image processing method, and program | |
JP5366756B2 (ja) | 情報処理装置及び情報処理方法 | |
US20060008173A1 (en) | Device and method for correcting image including person area | |
CN103632130B (zh) | 验证被摄体的验证装置及验证方法 | |
JP5662670B2 (ja) | 画像処理装置、画像処理方法、及びプログラム | |
JP2009514107A (ja) | 集合体からの特定の人物の割り出し | |
JP6941966B2 (ja) | 人物認証装置 | |
US11710347B2 (en) | Information processing apparatus, information processing method, and program | |
JP2010118868A (ja) | 情報処理装置およびその制御方法 | |
JP2013065118A (ja) | 顔認識装置、及び顔認識方法 | |
JP2013232060A (ja) | 画像認識装置、その制御方法、および制御プログラム、並びに撮像装置および表示装置 | |
JP2024045460A (ja) | 情報処理システム、情報処理装置、情報処理方法、およびプログラム | |
US9286707B1 (en) | Removing transient objects to synthesize an unobstructed image | |
JP5895624B2 (ja) | 画像処理装置、画像処理方法、制御プログラムおよび記録媒体 | |
WO2020115910A1 (ja) | 情報処理システム、情報処理装置、情報処理方法、およびプログラム | |
JP5861530B2 (ja) | 利用者検知装置、方法、及びプログラム | |
JP4781248B2 (ja) | 撮像装置、撮像装置の制御方法、プログラム及び記録媒体 | |
JP2008148262A (ja) | 撮像装置及び撮像装置の制御方法、プログラム及び記憶媒体 | |
JP7371806B2 (ja) | 情報処理装置、情報処理方法、およびプログラム | |
US20220392256A1 (en) | Authentication device, registration device, authentication method, registration method, and storage medium | |
JP2017033390A (ja) | 画像解析装置及びプログラム |
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: 07738595 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 12282631 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 200780009066.5 Country of ref document: CN |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1020087023721 Country of ref document: KR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2007738595 Country of ref document: EP |