WO2006080755A1 - Procede, support, et appareil pour le groupage de photographies basees sur des personnes dans un album de photographies numeriques, et procede, support et appareil de realisation d'album de photographies numeriques basees sur des personnes - Google Patents

Procede, support, et appareil pour le groupage de photographies basees sur des personnes dans un album de photographies numeriques, et procede, support et appareil de realisation d'album de photographies numeriques basees sur des personnes Download PDF

Info

Publication number
WO2006080755A1
WO2006080755A1 PCT/KR2005/003401 KR2005003401W WO2006080755A1 WO 2006080755 A1 WO2006080755 A1 WO 2006080755A1 KR 2005003401 W KR2005003401 W KR 2005003401W WO 2006080755 A1 WO2006080755 A1 WO 2006080755A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
person
photo
group
clustering
Prior art date
Application number
PCT/KR2005/003401
Other languages
English (en)
Inventor
Sang-Kyun Kim
Ji-Yeun Kim
Young-Su Moon
Yong-Man Ro
Seung-Ji Yang
Original Assignee
Samsung Electronics Co., Ltd.
Research & Industrial Cooperation Group
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from KR1020050048399A external-priority patent/KR100601997B1/ko
Application filed by Samsung Electronics Co., Ltd., Research & Industrial Cooperation Group filed Critical Samsung Electronics Co., Ltd.
Priority to EP05856532A priority Critical patent/EP1805684A4/fr
Publication of WO2006080755A1 publication Critical patent/WO2006080755A1/fr

Links

Classifications

    • 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
    • G06F16/55—Clustering; Classification
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/30—Scenes; Scene-specific elements in albums, collections or shared content, e.g. social network photos or video
    • 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
    • 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

  • Embodiments of the present invention relate to a digital photo album, and more particularly, to a method, medium, and apparatus for person-based photo clustering, and a person-based digital photo albuming method, medium, and apparatus
  • a digital photo album can be used to transfer photos taken by a user from a digital camera or a memory card to a local storage apparatus of the user and to manage the photos conveniently in a computer. Users may cluster and index many photos in a time series, by person, or even by specific category, using the photo album. A correspondingly constructed photo album enables the users to easily browse photo data or share photos with other users later.
  • clustering of photo data by person i.e., clustering based on the people in the photo.
  • Automatic clustering of a large volume of photo data in relation to persons taken in the photos allows users to easily share photo data with others or easily browse photo data later.
  • Face recognition can be explained as a process of building a facial database containing facial data of different faces of different people, and then, comparing an input facial image with facial image data stored in the already built database to determine who the input facial image belongs to.
  • a facial image can be detected within photo data and, by comparing the facial image with facial image data stored in the database, the input facial image can be recognized.
  • a process almost identical to that of performing pattern recognition in an image processing method is performed, with such a method including image acquisition, pre-processing, face extraction, and face recognition.
  • the research has focused on a facial recognition method that is robust against external lighting, facial expression and pose of a face in which the face of a person can be more accurately extracted.
  • a camera may be used in different manners (e.g., though use of a zoom function or a flash), or the direction of a camera when a photo is taken may change, a variety of backgrounds can be shown in an identical space, and changes of external lighting or illumination are much greater.
  • identification parameters related to individual facial images stored in a database are defined and these are then used for face-based photo clustering.
  • identification parameters there are names, relations, sexes, ages, and so on.
  • a newly input image is found in the database by using an identification parameter input by the user, and then clustered.
  • the user since the user must directly input the identification parameter, a great amount of time is required for inputting the identification parameters, and therefore this method cannot be practically used.
  • Embodiments of the present invention provide a method, medium, and apparatus for person-based digital photo clustering.
  • embodiments of the present invention also provide person-based digital photo albuming method, medium, apparatus.
  • embodiments of the present invention include a person-based digital photo clustering method, including detecting a face area from photo data, extracting predetermined content-based face feature value information from the detected face area, obtaining predetermined person- based photo clustering hint information, and performing person-based clustering by applying the photo clustering hint information to the extracted content-based face feature value information.
  • the detecting of the face area further may further include normalizing a size of the face area by using detected face area information, to generate a normalized face area, and in the extracting of the predetermined content-based face feature value information, predetermined content-based face feature value information may be extracted from the normalized face area.
  • the predetermined content-based feature value information may include an item expressing color, texture, and/or shape feature values extracted by using a visual descriptor of a moving picture experts group (MPEG)-7, and/or an item expressing an audio feature value extracted by using an audio descriptor of the MPEG-7.
  • MPEG moving picture experts group
  • the person-based clustering hint information may include at least one of a personal appearance hint indicating appearance feature information of a person, a situation hint indicating information on a result of clustering based on situation information including a place or background of a photo, a camera hint indicating camera information, photographing information, and/or photo information included in a photo file, and an audio hint indicating audio information of a user stored together with the photo file.
  • the personal appearance hint may include at least one of clothes information, skin tone information, and hair style information of the person.
  • the situation hint may include a situation ID expressing an identifier of a situation to which the photo belongs.
  • the camera hint may include at least one of a taken time expressing information on a time when the photo was taken, a focal length expressing information on a focal length when the photo was taken, an exposure time expressing information on an exposure time when the photo was taken, a zoom info expressing information on a camera zoom when the photo was taken, a flash info expressing information on a flash when the photo was taken, and a GPS info expressing global positioning system (GPS) information when the photo was taken.
  • the audio hint may include a speech info expressing speech information extracted from audio information stored together with photo data and recognized keyword information.
  • the performing of person-based clustering may include performing clustering based on a situation in which a photo is taken, by applying the photo clustering hint information to the extracted content-based face feature value information, to generate a situation-based cluster, and performing person-based clustering in the situation-based cluster, by applying the photo clustering hint information to the extracted content- based person feature value information.
  • the method may further include indexing each of multiple person-based clusters if person-based clustering is finished.
  • a person-based cluster may be compared with predetermined indexed personal photos stored in a database and stored according to a closest corresponding index.
  • the predetermined indexed personal photos stored in the database may have respective person information, and the person information may be formed with person list information including at least one of an identifier of the person, the name of the person, importance of the person, relations of the person with other people in a group of the person, and information on a higher-level group to which the person belongs.
  • the predetermined indexed personal photos stored in the database may further include information on the group to which the person belong, and the information of the person is formed with person group list information including at least one of an identifier of the group, the name of the group, importance of the group, relations of the group with other person groups, and information on a higher-level group to which the group belongs.
  • embodiments of the present invention include a person-based digital photo clustering apparatus including a face area detection unit to detect a face area from photo data, a face feature value extraction unit to extract predetermined content-based face feature value information from the detected face area, a clustering hint information obtaining unit to obtain predetermined person-based photo clustering hint information, and a clustering unit to perform person-based clustering by applying the photo clustering hint information to the extracted content-based face feature value information.
  • the apparatus may further include a face area normalization unit to normalize a size of the face area by using detected face area information to generate a normalized face area, wherein the face feature value extraction unit extracts predetermined content-based face feature value information from the normalized face area.
  • the predetermined content-based feature value information of the face feature value extraction unit may include an item expressing color, texture, and/or shape feature values extracted by using a visual descriptor of an MPEG-7, and/or an item expressing an audio feature value extracted by using an audio descriptor of the MPEG- 7.
  • the clustering hint information obtaining unit may include at least one of a personal appearance hint extraction unit extracting a personal appearance hint indicating appearance feature information of a person, a situation hint extraction unit extracting a situation hint indicating information on a result of clustering based on situation information including a place or background of a photo, a camera hint extraction unit extracting a camera hint indicating camera information, photographing information, and/or photo information included in a photo file, and an audio hint extraction unit extracting an audio hint indicating audio information of a user stored together with the photo file.
  • the clustering unit may include a situation-based clustering unit to perform clustering based on a situation in which a photo is taken, by applying the photo clustering hint information to the extracted content-based face feature value information, to generate a situation-based cluster, and a person-based clustering unit to perform person-based clustering in the situation-based cluster, by applying the photo clustering hint information to the extracted content-based person feature value information.
  • the apparatus may include a person-based indexing unit to index each of multiple person-based clusters if person-based clustering is finished.
  • the apparatus may further include a group preference hint obtaining unit to obtain a group preference value, wherein by referring to the group preference value obtained by the group preference hint obtaining unit, the person-based indexing unit compares a person-based cluster with predetermined indexed personal photos stored in a database and stored according to a closest corresponding index.
  • the predetermined indexed personal photos stored in the database may have respective person information, and information on a group to which other people and the person belong, and the person information may be formed with person list information comprising at least one of an identifier of the person, the name of the person, importance of the person, relations of the person with other people, and information on a higher-level group to which the person belongs, and the person group information is formed with person group list information comprising at least one of an identifier of the group, the name of the group, importance of the group, relations of the group with other groups, and information on a higher-level group to which the group belongs.
  • embodiments of the present invention include a person-based digital photo albuming method, including generating photo description information including a photo identifier for identifying at least a photo, generating albuming tool description information supporting person- based photo clustering and including at least a parameter for person-based photo clustering, performing photo albuming including the person-based clustering by using at least the photo description information and albuming tool description information, generating predetermined photo group description information from photo albumed data from the performing of the photo albuming, and generating predetermined photo album information by using the photo description information and the predetermined photo group description information.
  • the method may further include receiving an input of photo data from an internal memory device of a digital camera or a portable memory device.
  • EXIF information such as camera information and photographing information
  • existence of EXIF information may be confirmed in a photo file of an input photo, and if the EXIF information exists in the photo file, the EXIF information may be extracted and expressed according to a predetermined photo description scheme.
  • the photo description information may include at least a photo identifier, information on a photographer having taken the photo, photo file information, camera information, photographing information and a content-based feature value, with the content-based feature value including a visual descriptor generated by using pixel information of the photo, including color, texture, and/or shape feature values, and/or an audio descriptor including a speech feature value.
  • the parameter for person-based photo clustering may include a person-based group list item describing an indexed person-based group list, and a person-based clustering hint item for improvement of performance of person-based clustering.
  • the person-based group list item may include group list information indicating a list of groups after person-based clustering and indexing are performed, and person list information indicating a list of persons after person-based clustering and indexing are performed.
  • Individual group information on each of multiple groups forming the group list information may include at least one of an identifier of a group, a name of the group, importance of the group, relationships of the group, and an identifier of a higher-level group to which the group belongs.
  • the importance of the group may be determined by a frequency that a user watches photos included in the group or people in the group.
  • Individual person information forming the person list information may include at least one of an identifier of the person, a name of the person, relationships of the person with other people, an importance of the person, and an identifier of an upper- level group to which the person belongs.
  • the importance of the person may be determined by a frequency that the user watches photos including the person.
  • the person-based clustering hint item may include at least one of a personal appearance hint indicating appearance feature information of a person, a situation hint indicating information on a result of situation-clustering based on situation information including a place or background of the photo, a camera hint indicating camera information, photographing information, and/or photo information included in a photo file, a group preference hint indicating information on a group list, selected by the user, which photos to be clustered belong, and an audio hint indicating audio information of a user stored together with the photo file.
  • the situation hint may include an identifier of a situation to which the photo belongs as the result of the situation-based clustering of the photo.
  • the personal appearance hint may include at least one of information on clothes worn by the person, skin tone information of the person, and hair style information of the person.
  • the camera hint may include at least one of information on a time when the photo was taken, information on a focal length when the photo was taken, information on an exposure time when the photo was taken, information on a camera zoom when the photo was taken, information on a flash when the photo was taken, and GPS information when the photo was taken.
  • the group preference hint may include a group preference value numerically expressing groups selected by the user and a degrees of importance of the groups.
  • the audio hint may include a speech info expressing speech information extracted from audio information stored together with the photo file and recognized keyword information.
  • the performing of the photo albuming including person-based clustering may include person-based clustering of digital photo data.
  • the person-based photo clustering may include detecting a face area from photo data, extracting predetermined content-based face feature value information from the detected face area, obtaining predetermined person-based photo clustering hint information, and performing person-based clustering by applying the photo clustering hint information to the extracted content-based face feature value information.
  • the person-based photo clustering may further include normalizing a size of the face area by using information of the detected face area, to generate a normalized face area, and in the extracting of the content-based face feature value information, predetermined content-based face feature value information is extracted from the normalized face area.
  • the method may further include indexing each of multiple person-based clusters if person-based clustering is finished.
  • a person-based cluster may be compared with predetermined indexed personal photos stored in a database and is stored with according to a closest corresponding index.
  • the predetermined indexed personal photos stored in the database may have respective person information, and the person information may be formed with person list information including at least one of an identifier of the person, a name of the person, importance of the person, relations of the person with other people, and information on a higher-level group to which the person belongs.
  • the predetermined indexed personal photos stored in the database may further include information on a group to which the person and the other people belong, and the person group information may be formed with person group list information including at least one of an identifier of the group, the name of the group, importance of the group, relationship of the group with other groups, and information on a higher- level group to which the group belongs.
  • embodiments of the present invention include a person-based digital photo albuming apparatus, including a photo description information generation unit to generate photo description information including a photo identifier for identifying at least a photo, an albuming tool description information generation unit to generate albuming tool description information supporting person-based photo clustering and including at least a parameter for person-based photo clustering, an albuming tool to perform photo albuming including the person-based clustering by using at least the photo description information and the albuming tool description information, photo group information generation unit to generate predetermined photo group description information based on an output of the albuming tool, and photo album information generation unit to generate predetermined photo album information by using the photo description information and the predetermined photo group description information.
  • the apparatus may further include a photo input unit to receive an input of photo data from an internal memory device of a digital camera or a portable memory device.
  • the photo description information generation unit may confirms an existence of
  • EXIF information such as camera information and photographing information
  • a photo file of an input photo and if the EXIF information exists in the photo file, extracts the EXIF information and expresses according to a predetermined photo description scheme.
  • the photo description information may include at least a photo identifier, information on a photographer having taken the photo, photo file information, camera information, photographing information and a content-based feature value, with the content-based feature value potentially including a visual descriptor generated by using pixel information of the photo, and including color, texture, and/or shape feature values, and/or an audio descriptor including a speech feature value.
  • the parameter for person-based photo clustering of the albuming tool description information generation unit may include a person-based group list item describing an indexed person-based group list, and a person-based clustering hint item for improvement of the performance of person-based clustering.
  • the person-based group list item may include group list information indicating a 1 ist of groups after person-based clustering and indexing are performed, and person list information indicating a list of persons after person-based clustering and indexing are performed.
  • Individual group information on each of multiple groups forming the group list information may include at least one of an identifier of a group, a name of the group, importance of the group, relationships of the group, and an identifier of a higher-level group to which the group belongs.
  • Individual person information forming the person list information may include at least one of an identifier of the person, a name of the person, relationships of the person with other people, an importance of the person, and an identifier of an upper- level group to which the person belongs.
  • the person-based clustering hint item may include at least one of a personal appearance hint indicating appearance feature information of a person, a situation hint indicating information on a result of situation-clustering based on situation information including a place or background of the photo, a camera hint indicating camera information, photographing information, and/or photo information included in a photo file, a group preference hint indicating information on a group list, selected by the user, which photos to be clustered belong, and an audio hint indicating audio information of a user stored together with the photo file.
  • the albuming tool may include a person-based photo clustering tool performing person-based clustering of digital photo data.
  • the person-based photo clustering tool may include a face area detection unit to detect a face area from photo data, a face feature value extraction unit to extract predetermined content-based face feature value information from the detected face area, a clustering hint information obtaining unit to obtain predetermined person-based photo clustering hint information, and a clustering unit to perform person-based clustering by applying the photo clustering hint information to the extracted content-based face feature value information.
  • the person-based photo clustering tool may further include a face area normalizing unit to normalize a size of the detected face area by using information of the detected face area, to generate a normalized face area, and the face feature value extraction unit to extract predetermined content-based face feature value information from the normalized face area.
  • the apparatus may include a person-based indexing unit to index each of multiple person-based clusters if person-based clustering is finished.
  • the person- based indexing unit may further include a group preference hint obtaining unit to obtain a group preference value, and by referring to the group preference value, obtained by the group preference hint obtaining unit, the person-based indexing unit may compare a person-based cluster with predetermined indexed personal photos stored in a database and is stored according to a closest corresponding index.
  • the predetermined indexed personal photos stored in the database may have respective person information, and information on a group to which the person and other people belong, and the person information may be formed with person list information comprising at least one of an identifier of the person, a name of the person, importance of the person, relationships of the person with the other people, and information on a higher-level group to which the person belongs, and the person group information may be formed with person group list information comprising at least one of an identifier of the group, a name of the group, importance of the group, relationships of the group with other groups, and information on a higher-level group to which the group belongs.
  • embodiments of the present invention include a medium including computer readable code to implement embodiments of the present invention.
  • Embodiments of the present invention include a person-based digital photo clustering method, medium, and apparatus, and a person-based digital photo albuming method, medium, and apparatus, shifting away from the face recognition methods used in conventional security systems.
  • Photo data taken in the daily lives of users may be clustered, by person, such that an album can be effectively formed with a large volume of photo data. That is, person-based digital photo clustering can be effectively performed by using additional information including content-based feature information of face parts included in photo data, appearance feature information, such as the clothes color, hair style, skin tone of each person, and situation information such as a taken time, and backgrounds.
  • FlG. 1 is a block diagram of a person-based digital photo albuming apparatus, according to an embodiment of the present invention.
  • FlG. 2 is a block diagram of a person-based photo clustering tool , according to an embodiment of the present invention.
  • FlG. 3 illustrates a structural diagram of photo description information generated in a photo description information generation unit, according to an embodiment of the present invention
  • FlG. 4 illustrates a description scheme expressing parameters for person-based photo clustering in a process for person-based photo clustering, e.g., by using the photo information described in FlG. 3, according to an embodiment of the present invention
  • FlG. 5 is a block diagram showing a description scheme expressing an item
  • person-based group list describing a person-based group list among parameters for person-based photo clustering, e.g., as described in FlG. 4, according to an embodiment of the present invention
  • FlG. 6 is a block diagram showing a description scheme of person-based photo clustering hint items to improve a performance of person-based photo clustering, according to an embodiment of the present invention
  • FlG. 7 is a block diagram showing a description scheme (photo group description) expressing photo group information after photos are person-based clustered, according to an embodiment of the present invention
  • FlG. 8 is a block diagram of a photo information description scheme, according to an embodiment of the present invention, expressed as an XML schema;
  • FlG. 9 is a block diagram of a parameter description scheme for photo albuming, according to an embodiment of the present invention, expressed as an XML schema;
  • FlG. 10 is a block diagram of a photo group description scheme, according to an embodiment of the present invention, expressed as an XML schema;
  • FlG. 11 is a block diagram of an entire description scheme for digital photo albuming using person-based photo clustering, according to an embodiment of the present invention, expressed as an XML schema;
  • FlG. 12 is a flowchart of operations performed by a person-based digital photo album generation method, according to an embodiment of the present invention.
  • FlG. 13 is a detailed flowchart of an operation 1230 of FlG. 12, according to an embodiment of the present invention.
  • FlG. 14 is a detailed flowchart of an operation 1330 of FlG. 13, according to an embodiment of the present invention.
  • FlG. 15 is a flow diagram showing an example of a method by which photo data may be first situation-based clustered and then indexed, according to an embodiment of the present invention
  • FlG. 16 illustrates an embodiment of a person list and a person group list stored in a database
  • FlG. 17 illustrates an embodiment of relations among persons expressed in a table.
  • a person-based digital photo albuming apparatus may include a photo description information generation unit 110, an albuming tool description information generation unit 120, an albuming tool 130, a photo group information generation unit 140, and a photo albuming information generation unit 150.
  • the apparatus may further include a photo input unit 100.
  • the photo input unit 100 may receive photo data, e.g., a series of photos, from an internal memory apparatus of a digital camera or a portable memory apparatus, for example.
  • the photo description information generation unit 110 may generate photo description information describing a photo and including at least a photo descriptor.
  • the photo description information generation unit 110 can confirm from each input photo whether or not there is EXIF information, such as camera information and photographing information, stored in a photo file, and if the information is in the photo file, the information may be extracted and expressed according to a photo description scheme. At the same time, content-based feature values may be extracted from the pixel information of the photo and expressed according to the photo description scheme.
  • the photo description information may be input to the photo albuming tool 130 for grouping photos.
  • the albuming tool description information generation unit 120 may generate albuming tool description information, including predetermined parameters, for more efficiently clustering and indexing photos by using the variety of generated photo de- scription information items.
  • the albuming tool 130 may perform photo albuming, including person-based clustering, by using at least the photo description information and the albuming tool description information, and may include a person-based photo clustering tool 135 clustering digital photo data based on individual people.
  • the person-based photo clustering tool 135 is a person-based digital photo clustering apparatus which clusters digital photo data based on individual people.
  • the person-based photo clustering tool 135 may be formed with a face area detection unit 200, a face feature value extraction unit 220, a clustering hint information obtaining unit 230, and a person-based clustering unit 240. Also, the person-based clustering tool 135 may further include a face area normalization unit 210. It may also be preferable for the person-based clustering tool 135 to further include a person-based indexing unit 250.
  • the face area detection unit 200 may detect a face area from photo data.
  • the face area normalization unit 210 may normalize the size of the face area by using the detected face area information.
  • the face feature value extraction unit 200 may extract predetermined content-based face feature value information from the normalized face area.
  • the clustering hint information obtaining unit 230 may obtain predetermined person-based photo clustering hint information, and include at least one of a personal appearance hint extracted unit 232, a camera hint extraction unit 234, a situation hint extraction unit 236, and an audio hint extraction unit 238.
  • the personal appearance hint extraction unit 232 may extract a personal appearance hint indicating appearance feature information of a person.
  • the camera hint extraction unit 234 may extract a camera hint indicating camera information, photographing information, and photo information included in a photo file.
  • the situation hint extraction unit 236 may extract a situation hint indicating information on the result clustered based on situation information, such as the place where a photo was taken, or the background.
  • the audio hint extraction unit 238 may extract an audio hint indicating audio information of a user, stored together with a photo file.
  • the person-based clustering unit 240 may perform person-based clustering by applying the photo clustering hint information to the extracted content-based face feature value information.
  • the person-based indexing unit 250 may compare the person-based photo cluster with predetermined personal photos stored in a database 260, and store as a closest index.
  • Each of the predetermined indexed personal photos stored in the database 260 can have an individual person's information and information on a group to which the person belongs.
  • the individual person information can be formed with person list in- formation, including at least one of the identifier (ID) of the person, the name of the person, importance of the person, relations with other people, and information on a higher-level group to which the person belongs.
  • the person group information may be formed with person group list information including at least one of the ID of the person group, the name of the person group, importance of the person group, relations of the person group with other person's groups, and information on a higher-level group to which the person group belongs.
  • a database updating unit 280 can update indexed photos of the database 280 by using the indexed photos.
  • the photo group information generation unit 140 can generate predetermined photo group description information from the output of the albuming tool 130.
  • the photo albuming information generation unit 150 may then generate predetermined photo albuming information by using the photo description information and the photo group description information.
  • FlG. 3 illustrates a structural diagram of photo description information generated in the photo description information generation unit 110.
  • the photo description information can express camera information and photographing information stored in a file and content-based feature value information extracted from the contents of photos.
  • the photo information description information 30 may include a photo identifier (Photo ID) 300 identifying each photo, an item (Author) 320 expressing the author having taken the photo, an item (File information) 340 expressing file information stored in the photo file, an item (Camera information) 360 expressing camera information stored in the photo file, and an item (Content-based information) 380 expressing a content-based feature value.
  • the photo information description information 30 may also include an item (File name) 342 expressing the name of the photo file, an item (File format) 344 expressing the format of the photo file, an item (File size) 346 expressing the capacity of the photo file and an item (File creation date/time) 348 expressing the date and time when the photo file was created.
  • an item (File name) 342 expressing the name of the photo file
  • an item (File format) 344 expressing the format of the photo file
  • an item (File size) 346 expressing the capacity of the photo file
  • an item (File creation date/time) 348 expressing the date and time when the photo file was created.
  • the photo information description information 30 may also include an item (IsEXIFlnformation) 362 expressing whether or not the photo file includes EXIF information, an item (Camera model) 364 expressing the camera model having taken the photo, an item (Taken date/time) 366 expressing the date and time when the photo was taken, an item (GPS information) 368 accurately expressing the location information of the place where the photo was taken, with information obtained by a global positioning system (GPS), an item (Image width) 370 expressing the width information in the size of the photo, an item (Image height) 372 expressing the height information in the size of the photo, an item (Flash on/off) 374 expressing whether or not a camera flash was used to take the photo, an item (Brightness) 376 expressing the brightness information of the photo, an item (Contrast) 378 expressing the contrast information of the photo, and an item (Sharpness) 379
  • the information 380 expressing a content-based feature value extracted from a photo may include an item (Visual descriptor) 382 expressing feature values of color, texture, and shape extracted by using MPEG-7 Visual Descriptor, and an item (Audio descriptor) 384 expressing a feature value of voice extracted by using MPEG-7 Audio Descriptor.
  • Visual descriptor expressing feature values of color, texture, and shape extracted by using MPEG-7 Visual Descriptor
  • Audio descriptor expressing a feature value of voice extracted by using MPEG-7 Audio Descriptor.
  • FTG. 4 illustrates a description scheme expressing parameters for person-based photo clustering in a process for person-based photo clustering by using the photo information, e.g., as described in FTG. 3.
  • parameters 40 for photo clustering by person may include an item (person-based group list) 400 describing an already indexed person-based group list, and person-based clustering hints 450.
  • FTG. 5 is a block diagram showing a description scheme expressing an item
  • (person-based group list) 400 describing a person-based group list among parameters for person-based photo clustering, e.g., as described in FTG. 4.
  • the item 400 describing a person-based group list may be an item to express lists of groups and persons stored after photos are person-based clustered and indexed.
  • the item 400 may be formed with a group information item (Group info) 500 and a person information item (Person info) 550.
  • the group information item 500 may be an item expressing group information of plurality of persons, and can include information of one or more groups.
  • the group information item 500 may include an item (Group ID) 502, expressing a unique ID of a group, an item (Group name) 504, expressing a unique name of the group, an item (Group importance value) 506, expressing the degree of importance of the group, an item (Inter-group relationship table) 508, expressing the relationship with other groups in the form of a table, and an item (Parent group IDs) 509, expressing information on a parent group (a higher-level group) to which the group belongs, with an ID.
  • the item 506, expressing the degree of importance of the group can be automatically measured by using information recording how frequently the user watches photos of the group or persons belonging to the group, or can be set with an arbitrary value directly by the user, for example.
  • Methods for determining the degree of importance of a group include the above examples, but are not limited to the same.
  • the item 508, expressing the relationship with other groups, in the form of a table is an item expressing the relationship of two groups, storing the frequency of photos in which people in the group and a person in another arbitrary group are photographed together, and numerically describing the frequency.
  • Methods of expressing the relationship with other groups, in the form of a table include the above examples, but are not limited to the same.
  • the person information item 550 is an item expressing information of people belonging to a group, and can include information of one or more people.
  • the person information item 550 may include an item (Person ID) 552, expressing the unique ID of the person, an item (Person name) 554, expressing the unique name of the person, an item (Person importance value) 556, expressing the degree of importance of the person, an item (Inter-person relationship table) 558, expressing any relationships with other people, in the form of a table, and an item (Parent group IDs) 559, expressing information of a parent group (a higher-level group) to which the person belongs, with an ID.
  • the item 556 expressing the degree of importance of the person, can be automatically measured by using information recording how frequently the user watches photos to which that person belongs, or can be set with an arbitrary value directly by the user, for example.
  • Methods for determining the degree of importance of a person include the above examples, but are not limited to the same.
  • the item 558 expressing the relationship with other people, in the form of a table, may be an item expressing the relationship of two people, by storing the frequency of photos in which a person and another arbitrary person are photographed together, and numerically describing the frequency.
  • Methods of expressing the relationship with other people, in the form of a table may include the above example, but are not limited to the same.
  • the person-based clustering hit item (Person-based clustering hints) 450 may include an item (Situation hints) 610, expressing the result of situation-based clustering of photos, an item (Personal appearance hints) 620, expressing appearance feature information of a person, an item (Camera hints) 630, expressing camera information and photographing information stored in a photo file, an item (Group preference hint) 640, expressing information on a rough group list which is directly selected by the user and to which photos to be clustered belong, and an item (Audio hints) 650, expressing audio information stored together with a photo.
  • an item (Situation hints) 610 expressing the result of situation-based clustering of photos
  • an item (Personal appearance hints) 620 expressing appearance feature information of a person
  • an item (Camera hints) 630 expressing camera information and photographing information stored in a photo file
  • an item (Group preference hint) 640 expressing information on a
  • the item 610 expressing the result of situation-based clustering of photos, may include situation information 612 of a situation to which a photo belongs, as the result of situation-based clustering of photos, and may further include a unique ID.
  • the item 620 may include an item (Cloth info) 622, expressing information on clothes that the person wears, an item (Skin tone) 624, expressing information on the skin tone of the person, and an item (Hair style) 626, expressing the hair style of the person.
  • the item 620, expressing appearance feature information of a person may include the above examples, but is not limited to the same.
  • the item 630 expressing camera information and photographing information stored in the photo file, may include an item (Taken time) 631, expressing information on the time when the photo was taken, an item (Focal length) 632, expressing information on the focal length when the photo was taken, an item (Exposure time) 633, expressing information on the exposure time when the photo was taken, an item (Zoom info) 634, expressing information on the camera zoom when the photo was taken, an item (Rash info) 635 expressing information on the flash when the photo was taken, and an item (GPS info) 636, expressing GPS information when the photo was taken.
  • Taken time 631
  • Focal length expressing information on the focal length when the photo was taken
  • an item (Exposure time) 633 expressing information on the exposure time when the photo was taken
  • an item (Zoom info) 634 expressing information on the camera zoom when the photo was taken
  • an item (Rash info) 635 expressing information on the flash when the photo was taken
  • GPS info GPS
  • the item 630 expressing camera information and photographing information stored in the photo file, may include the above examples, but is not limited to the same.
  • the item 640 expressing information on a rough group list that is directly selected by the user, and to which photos to be clustered belong, may include an item (Group preference value) 642 numerically expressing groups selected by the user and the degrees of importance of the groups.
  • the item 650 may include an item (Speech info) 652, expressing speech information extracted from the audio information stored together with the photo data, and recognized keyword information.
  • the item 650, expressing audio information stored together with a photo may include the above example, but is not limited to the same.
  • FIG. 7 is a block diagram showing a description scheme (photo group description)
  • the photo group may include a photo group (Person-based photo group) 710 by person- based photo clustering, and each cluster group may include a lower-level group (Photo series) 730.
  • Each photo group can include a plurality of photos as photo IDs (Photo ID) 735.
  • Each person-based cluster may include a group ID (Group ID) 720 and is referred to by a person-based group list.
  • FIG. 8 is a block diagram of a photo information description scheme, according to an embodiment of the present invention, expressed in an XML schema.
  • the description scheme expressing camera information and photographing information stored in a photo file and content-based feature value information extracted from the content of a photo can be expressed in an XML format as follows:
  • FlG. 9 is a block diagram of a parameter description scheme for photo albuming, according to an embodiment of the present invention expressed as an XML schema.
  • the description scheme expressing parameters for person-based photo clustering can be expressed in an XML format, as follows:
  • FlG. 10 is a block diagram of a photo group description scheme, according to an embodiment of the present invention, expressed as an XML schema.
  • the description scheme expressing photo group information after photos are clustered can be expressed in an XML format, as follows:
  • FlG. 11 is a block diagram of an entire description scheme for digital photo albuming using person-based photo clustering, according to an embodiment of the present invention, expressed as an XML schema.
  • the entire description scheme for digital photo albuming can be expressed in an XML format, as follows:
  • FlG. 12 is a flowchart of the operations performed by a person-based digital photo album generation method, according to an embodiment of the present invention.
  • the method of generating a person-based digital photo album, according to an embodiment of the present invention, and the operation of an apparatus therefore will now be further explained with reference to FlG. 12.
  • a digital photo album may be effectively generated with digital photo data by using the information described above. Accordingly, first, if a digital photo is input through a photo input unit 100, in operation 1200, photo description information describing the photo and including at least the photo ID may be generated, in operation 1210.
  • albuming tool description information including a predetermined parameter for digital photo clustering, may be generated by the albuming tool description information generation unit 120, in operation 1220. Then, the photo is person-based clustered by using the input photo, the photo description information, and the albuming tool description information, in operation 1230.
  • the person-based clustered result may be generated as predetermined photo group description information by the photo group information generation unit 140, in operation 1240.
  • Predetermined photo album information may be generated in the photo album information generation unit 150, by using the photo description information and the photo group description information, in operation 1250.
  • FlG. 13 is a detailed flowchart of the operation 1230. The operation of the person- based digital photo clustering apparatus for person-based clustering and a method thereof will now be further explained with reference to FlG. 13.
  • a face area may be detected from photo data by the face area detection unit 200, in operation 1300.
  • P race (i) is pixel information of the extracted face area and can be expressed in the following equation 1 : [359]
  • p face (i,m) denotes pixel information of an m-th face area extracted from an i- th photo. This face extraction process may be repeated in all the N input photos.
  • the face area normalization unit 210 may normalize the size of all face areas into a size with a width of A and a length of B, in operation 1310.
  • P norm (i) is pixel information of the normalized face area and may be expressed as the following equation 2: [363]
  • P norm CO iPnorm Q ⁇ XPnorm Q ⁇ XPnorm Q?> ⁇ -, Pnorm (hTMX-,P»orr a QM ⁇ ⁇ ), Pnorm QM)) (2)
  • p (i,m) denotes pixel information of the normalized face area of an m-th norm face area extracted from an i-th photo. This face normalization process may be repeated in all the N input photos.
  • the face feature value extraction unit 220 may extract content-based face feature value information from the normalized face areas, in operation 1320.
  • F face (i) is information on M content-based face feature values extracted from the i- th photo and may be expressed as the following equation 3: [367]
  • f (i,m) denotes content-based feature value information of an m-th face area face extracted from an i-th photo.
  • a Face Recognition descriptor of an MPEG-7 or Advanced Face Recognition descriptor may be used. Methods of extracting content-based face feature values include the above example, but are not limited to the same.
  • the clustering hint obtaining unit 230 may extract person-based photo clustering hint information that is a parameter to improve the performance of person- based photo clustering, in operation 1330.
  • the person-based photo clustering hint information may include a 'Personal appearance hint' indicating appearance feature information of a person, a 'Situation hint' indicating information on the clustered result based on situation information such as the place of a photo or the background, a 'Camera hint' indicating camera information, photographing information, and photo information included in a photo file, and an 'Audio hint' indicating audio information of the user stored together with a photo file.
  • H (i,m) is person-based photo clustering information of an m-th person extracted hint from an i-th photo and may be expressed as the following equation 4: [371]
  • H personal_appearance (i,m) denotes the appearance feature information hint of an m- th person extracted from an i-th photo, denotes the camera information hint of the i-th photo, H camera (i) denotes the camera information hint of the i-th photo, H situation (i) denotes the situation-based clustering hint information of the i-th photo, and H (i) audio denotes the audio hint information of the i-th photo.
  • the person-based photo clustering information hint items include the above examples, but are not limited to the same.
  • the operation 1330 may include at least one of an operation
  • H (i,m) is a personal appearance feature information hint item that personal_appearance includes information items indicating clothes information, skin tone information and hair style of a person, for example, and may be expressed as the following equation 5: [376]
  • h (i,m) denotes clothes information of the person
  • h (i,m) denotes skin cloth skin_tone tone information of the person
  • h (i,m) denotes hair style information of the hair_style person.
  • Items forming the appearance feature information hint of a person may include the above examples, but are not limited to the same.
  • H (i) is the camera information and pho- camera tographing information hint item stored in an i-th photo file and may include 'Taken time' that is an item expressing information on the time when the photo is taken, 'Focal length' that is an item expressing information on the focal length when the photo is taken, 'Exposure time' that is an item expressing information on the exposure time when the photo is taken, 'Zoom info' that is an item expressing information on the camera zoom when the photo is taken, 'Flash info' that is an item expressing information on the flash when the photo is taken, and 'GPS info' that is an item expressing GPS information when the photo is taken, and may be expressed as the following equation 6: [379]
  • H ⁇ n ⁇ CO t ⁇ tf ⁇ i . tTM. CO, hfocalj ⁇ CO, K jTM ⁇ CO, Koor, CO, h fl ⁇ h CO, V CO , ⁇ ⁇ ⁇ ⁇ C ⁇
  • h taken_time ( V i) ' may J denotes the time when the p fhoto was taken,> h focal_length ( V i) ' may J denote information on the focal length when the photo was taken, h (i) may denote ex_time exposure time information, h (i) may denote camera zoom information, h (i) may zoom flash denote flash information, and h (i) may denote GPS information.
  • the camera in- gps formation and photographing information hint items stored in a photo file may include the above examples, but are not limited to the same.
  • H situation (i) is the situation-based clustering information hint item of an i-th photo, and may include 'Situation ID' that is an item expressing the unique ID of a situation to which the photo belongs, and may be expressed as the following equation 7: [382]
  • h situationJD (i) may denote the unique ID information of the situation to which the photo belongs.
  • the situation-based clustering hint items of a photo may include the above example, but is not limited to the same.
  • H (i) is the audio information hint item of an i-th photo
  • h speech (i) may denote the speech information extracted from the audio in- formation stored together with the photo data and recognized keyword information.
  • the audio information hint items of a photo include the above example, but are not limited to the same.
  • clustering may bee performed first based on the situation in which a photo is taken. Then, in each situation-based cluster, person- based clustering may be performed. In the next operation, each clustered person-based cluster can be compared with person lists in the database and indexed.
  • FlG. 15 is a flow diagram showing an example of a method by which photo data may be first situation-based clustered and then indexed. This is to enable use of a variety of additional information items as well as personal face information in the process for clustering photos by person.
  • a new feature value may be generated, by applying the person-based clustering hint to extracted content-based face feature value information.
  • F (i) combined may denote newly generated feature value information of an i-th photo in the following equation 9: [390]
  • function ⁇ ( • ) is a function generating a new feature value, by using together
  • H h i nt (i,m) that is the hint information of an m-th p ferson of an i-th p fhoto and F ftce (vi,m) that is the content-based feature value information of the m-th person of the i-th photo.
  • photo data can be clustered based on situations, and person-based clustering can be performed for photos belonging to each situation.
  • a new feature value information generated by using together appearance feature information of a person, camera and/or photographing and/or photo information included in a photo file, the situation-based cluster information of the photo, and audio information of the user stored together with the photo file may be expressed as the following equation 10. This is based on the fact that in photo data having identical situation information the appearance feature information items of an identical person may be similar to each other. If N input photos are clustered into K situations, function ⁇ ( • ) may be defined in the following equation 10:
  • ⁇ • ⁇ l may indicate clustering of photos first based on situation in- situa ⁇ on(k) formation such as the place of photographing or backgrounds, and situation(k) may denote a k-th situation cluster.
  • w (i,m) may denote a weight obtained personal_appearance from H (i > m) that is the appearance feature information of an m-th person personal_appearance of an i-th photo
  • w camera (i) may denote a weight obtained from H camera (i) that is the camera and/or photographing information of the i-th photo
  • w (i) may denote a audio weight obtained from H audio (i) that is the audio information of the i-th photo.
  • C(k,t) denotes a t-th person cluster belonging to the k-th situation interval situation(k).
  • Each person cluster may include one or more person photos.
  • a feature value, representing the t-th person cluster belonging to the k-th situation cluster may be generated.
  • a method by which a representative person is selected among person photo data belonging to the person cluster, and the feature value of the representative person is used as the feature value representing the t-th person cluster can be used.
  • F person (k,t) denotes a feature value representing the t-th person cluster belonging to the k-th situation cluster.
  • F I denotes the person-based feature
  • Methods of generating a feature value representing a t-th person cluster may include the above example, but are not limited to the same.
  • a person-based indexing operation 1350 if person-based clustering for all K situation clusters is finished, an indexing operation for each person- based cluster may be performed next. Indexing of the person-based clusters may include an operation for comparing each person-based cluster with personal photos stored (already indexed) in a database, and storing with a closest index.
  • ⁇ person ⁇ X ⁇ V ID ⁇ X h V narne ⁇ X )> V smpor tance (4 delation ⁇ X h V parentIDs ⁇ X ) ) Q- ⁇ )
  • v ID (x) may denote the unique ID of the x-th person
  • v name (x) may denote the unique name of the person
  • v importance (x) may denote the degree of importance of the person
  • v (x) may denote information on the relationship with other persons, relation expressed in the form of a table
  • v parentIDs (x) may denote the parent group
  • a y-th person group may be stored with person group list information 1650.
  • the person group list information may be expressed in the following equation 14:
  • V g fa ⁇ O ( V ID OX V namz OX ⁇ mpor tan,, OX 1 W 0n OX V- ⁇ O) ⁇ 0 4 )
  • v (y) may denote the unique ID of the y-th person group
  • v (y) may denote the unique ID of the y-th person group
  • ID name denote the unique name of the person group
  • v (y) may denote the degree of importance importance of the person group
  • v relation (y) may denote information on the relationship with other person groups, expressed in the form of a table
  • v (y) may denote parentIDs the parent group (higher-level group) to which the person group belongs.
  • FlG. 16 illustrates an embodiment of a person list and a person group list stored in a database.
  • the person group list may be formed with people of the given person list (e.g., 'Me', 'Dad', 'Mom', 'Tom', 'Nick', 'John', 'Mike', 'Patty', and 'School mate'), and each person list has an importance value ranging from 0.0 to 1.0.
  • FlG. 17 illustrates an embodiment of relations among people expressed in a table.
  • the degree of relationship with 'John' is the highest, and this means that the frequency of photos in which 'Me' and 'John' are photographed together is the highest.
  • R (s)l is the feature value information of the s-th person photo data and may be formed with face feature value information, person/group list information, and a person group preference hint specified by the user, and may be expressed as the following equation 15:
  • H (y) may denote the preference of a y-th person group by the user.
  • a similarity distance value may be measured, by comparing the feature value (F person (k,t)) of the t-th person cluster, belonging to the k-th situation cluster, with feature value (R
  • ⁇ ( • ) denotes a function measuring the similarity distance value between two feature values.
  • V I denotes a person index allocated to the t-th person cluster, belonging to the k-th situation cluster, and the person index may be in the form of a unique ID. If the person is not in the database, a new ID may be allocated.
  • Embodiments of the present invention can also be embodied as computer readable code in a medium, e.g., a computer readable recording medium.
  • a computer may includes all apparatuses having an information processing function, for example.
  • the medium may be any data storage device that can store/transfer data which can be thereafter read by a computer system. Examples of the computer readable recording medium/media may include read-only memory (ROM), random-access memory (RAM), CD-ROMs, magnetic tapes, floppy disks, and optical data storage devices.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Multimedia (AREA)
  • General Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Library & Information Science (AREA)
  • Human Computer Interaction (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Health & Medical Sciences (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Image Analysis (AREA)

Abstract

La présente invention a trait à un procédé, un support, et un appareil pour le groupage de photographies basées sur des personnes dans un album de photographies numériques, et procédé, support et appareil de réalisation d'album de photographies numériques basées sur des personnes. Le procédé de groupage peut comprendre la détection d'une zone de visage à partir de données photographiques, l'extraction d'information prédéterminée de valeur de caractéristique de visage basé sur le contenu à partir de la zone de visage détectée, l'obtention d'une information d'algorithme d'optimisation de groupage de photographies basées sur des personnes, et la réalisation d'un groupage basé sur des personnes par l'application de l'information d'algorithme d'optimisation de groupage à l'information de valeur de caractéristique de visage basée sur le contenu. Le procédé de réalisation de photographies numériques basées sur des personnes peut comprendre la génération d'information de description de photographies pour l'identification d'au moins une photographie, la génération d'information de description d'outils de réalisation d'album en support de groupage de photographies basées sur des personnes et l'inclusion d'au moins un paramètre pour le groupage de photographies basées sur des personnes, la réalisation de l'album de photographies comprenant le groupage basé sur des personnes à l'aide d'au moins une information de description d'outils de réalisation d'album, la génération d'information de description de groupes de photographies prédéterminée à partir des données mises en album, et la génération d'information d'album de photographies prédéterminée à l'aide d'information de description de photographies et d'information de description de groupes de photographies.
PCT/KR2005/003401 2004-10-12 2005-10-12 Procede, support, et appareil pour le groupage de photographies basees sur des personnes dans un album de photographies numeriques, et procede, support et appareil de realisation d'album de photographies numeriques basees sur des personnes WO2006080755A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP05856532A EP1805684A4 (fr) 2004-10-12 2005-10-12 Procede, support, et appareil pour le groupage de photographies basees sur des personnes dans un album de photographies numeriques, et procede, support et appareil de realisation d'album de photographies numeriques basees sur des personnes

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US61708604P 2004-10-12 2004-10-12
US60/617,086 2004-10-12
KR1020050048399A KR100601997B1 (ko) 2004-10-12 2005-06-07 인물기반 디지털 사진 클러스터링 방법 및 장치와 이를이용한 인물기반 디지털 사진 앨버밍 방법 및 장치
KR10-2005-0048399 2005-06-07

Publications (1)

Publication Number Publication Date
WO2006080755A1 true WO2006080755A1 (fr) 2006-08-03

Family

ID=36740719

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2005/003401 WO2006080755A1 (fr) 2004-10-12 2005-10-12 Procede, support, et appareil pour le groupage de photographies basees sur des personnes dans un album de photographies numeriques, et procede, support et appareil de realisation d'album de photographies numeriques basees sur des personnes

Country Status (2)

Country Link
EP (1) EP1805684A4 (fr)
WO (1) WO2006080755A1 (fr)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008039487A1 (fr) * 2006-09-26 2008-04-03 Hewlett-Packard Development Company L.P. Extraction de caractéristiques à partir de régions du visage et de régions d'identification auxiliaire d'images utilisée pour la reconnaissance de personnes et pour d'autres applications
WO2008076179A1 (fr) * 2006-11-14 2008-06-26 Eastman Kodak Company Interface utilisateur pour une reconnaissance faciale
WO2009048513A2 (fr) * 2007-10-05 2009-04-16 Eastman Kodak Company Vêtements comme marqueur d'événement
WO2010047843A1 (fr) 2008-10-26 2010-04-29 Hewlett-Packard Development Company, L.P. Agencement d’images dans des pages à l'aide d'un filtrage basé sur un contenu et d'un groupage basé sur un thème
EP2273784A1 (fr) * 2008-04-25 2011-01-12 FUJIFILM Corporation Dispositif et procédé d'enregistrement d'image
US20120002881A1 (en) * 2010-01-22 2012-01-05 Kazuhiko Maeda Image management device, image management method, program, recording medium, and integrated circuit
EP2397970A3 (fr) * 2010-06-21 2014-12-31 Lg Electronics Inc. Terminal mobile et son procédé de génération de groupes
EP2693404A4 (fr) * 2011-03-28 2016-02-17 Nec Corp Dispositif de suivi de personne, procédé de suivi de personne et support lisible par ordinateur non temporaire stockant un programme de suivi de personne

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0476690A (ja) * 1990-07-13 1992-03-11 Hitachi Ltd 現金自動取引方式及び現金自動取引装置
KR950029985A (ko) * 1994-04-13 1995-11-24 배순훈 사람 얼굴의 화상처리를 이용한 개인 인식방법
EP0751473A1 (fr) * 1995-06-26 1997-01-02 Lucent Technologies Inc. Localisation de caractéristiques dans une image
US5835616A (en) * 1994-02-18 1998-11-10 University Of Central Florida Face detection using templates

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE69915566T2 (de) * 1998-11-25 2005-04-07 Eastman Kodak Co. Zusammenstellung und Änderung von Fotocollagen durch Bilderkennung

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0476690A (ja) * 1990-07-13 1992-03-11 Hitachi Ltd 現金自動取引方式及び現金自動取引装置
US5835616A (en) * 1994-02-18 1998-11-10 University Of Central Florida Face detection using templates
KR950029985A (ko) * 1994-04-13 1995-11-24 배순훈 사람 얼굴의 화상처리를 이용한 개인 인식방법
EP0751473A1 (fr) * 1995-06-26 1997-01-02 Lucent Technologies Inc. Localisation de caractéristiques dans une image

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP1805684A4 *

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7689011B2 (en) 2006-09-26 2010-03-30 Hewlett-Packard Development Company, L.P. Extracting features from face regions and auxiliary identification regions of images for person recognition and other applications
WO2008039487A1 (fr) * 2006-09-26 2008-04-03 Hewlett-Packard Development Company L.P. Extraction de caractéristiques à partir de régions du visage et de régions d'identification auxiliaire d'images utilisée pour la reconnaissance de personnes et pour d'autres applications
US8315463B2 (en) 2006-11-14 2012-11-20 Eastman Kodak Company User interface for face recognition
WO2008076179A1 (fr) * 2006-11-14 2008-06-26 Eastman Kodak Company Interface utilisateur pour une reconnaissance faciale
WO2009048513A3 (fr) * 2007-10-05 2009-07-23 Eastman Kodak Co Vêtements comme marqueur d'événement
WO2009048513A2 (fr) * 2007-10-05 2009-04-16 Eastman Kodak Company Vêtements comme marqueur d'événement
EP2273784A1 (fr) * 2008-04-25 2011-01-12 FUJIFILM Corporation Dispositif et procédé d'enregistrement d'image
US20110122291A1 (en) * 2008-04-25 2011-05-26 Fujifilm Corporation Image recording device and method
EP2273784A4 (fr) * 2008-04-25 2011-07-20 Fujifilm Corp Dispositif et procédé d'enregistrement d'image
US8780226B2 (en) 2008-04-25 2014-07-15 Fujifilm Corporation Image recording device and method which generates multi-image file based on classification
EP2351352A1 (fr) * 2008-10-26 2011-08-03 Hewlett-Packard Development Company, L.P. Agencement d images dans des pages à l'aide d'un filtrage basé sur un contenu et d'un groupage basé sur un thème
EP2351352A4 (fr) * 2008-10-26 2012-11-14 Hewlett Packard Development Co Agencement d images dans des pages à l'aide d'un filtrage basé sur un contenu et d'un groupage basé sur un thème
WO2010047843A1 (fr) 2008-10-26 2010-04-29 Hewlett-Packard Development Company, L.P. Agencement d’images dans des pages à l'aide d'un filtrage basé sur un contenu et d'un groupage basé sur un thème
US20120002881A1 (en) * 2010-01-22 2012-01-05 Kazuhiko Maeda Image management device, image management method, program, recording medium, and integrated circuit
EP2397970A3 (fr) * 2010-06-21 2014-12-31 Lg Electronics Inc. Terminal mobile et son procédé de génération de groupes
US9195882B2 (en) 2010-06-21 2015-11-24 Lg Electronics Inc. Mobile terminal and group generating method therein
EP2693404A4 (fr) * 2011-03-28 2016-02-17 Nec Corp Dispositif de suivi de personne, procédé de suivi de personne et support lisible par ordinateur non temporaire stockant un programme de suivi de personne

Also Published As

Publication number Publication date
EP1805684A4 (fr) 2008-10-22
EP1805684A1 (fr) 2007-07-11

Similar Documents

Publication Publication Date Title
US7756334B2 (en) Method, medium, and apparatus for person-based photo clustering in digital photo album, and person-based digital photo albuming method, medium, and apparatus
US20050280719A1 (en) Method, medium, and apparatus for detecting situation change of digital photo and method, medium, and apparatus for situation-based photo clustering in digital photo album
Cooper et al. Temporal event clustering for digital photo collections
US20060153460A1 (en) Method and apparatus for clustering digital photos based on situation and system and method for albuming using the same
US8150098B2 (en) Grouping images by location
Huang Color-spatial image indexing and applications
US8611677B2 (en) Method for event-based semantic classification
US20060074771A1 (en) Method and apparatus for category-based photo clustering in digital photo album
WO2006080755A1 (fr) Procede, support, et appareil pour le groupage de photographies basees sur des personnes dans un album de photographies numeriques, et procede, support et appareil de realisation d'album de photographies numeriques basees sur des personnes
EP2618289A2 (fr) Procédé et appareil pour incorporer une reconnaissance faciale automatique dans des collections d'images numériques
EP1854033A1 (fr) Groupage additif d'images depourvues d'informations temporelles
WO2007029916A1 (fr) Procédé et dispositif de codage/lecture de contenus multimédia
JP2014092955A (ja) 類似コンテンツ検索処理装置、類似コンテンツ検索処理方法、およびプログラム
Anguera et al. Multimodal photo annotation and retrieval on a mobile phone
KR100790865B1 (ko) 상황기반 디지털 사진 클러스터링 방법 및 장치와 이를이용한 디지털 사진 앨범 시스템 및 방법
Choi et al. Face annotation for personal photos using context-assisted face recognition
CN110879944A (zh) 基于人脸相似度的主播推荐方法、存储介质、设备及系统
KR100624465B1 (ko) 디지털 사진의 상황변환 검출 장치 및 방법 및 이를 이용한디지털 사진 앨범의 상황 기반 클러스터링 방법 및 장치
Li et al. Image content clustering and summarization for photo collections
Kutics et al. Use of adaptive still image descriptors for annotation of video frames
Kim et al. User‐Friendly Personal Photo Browsing for Mobile Devices
Ardizzone et al. Clustering techniques for personal photo album management
Chu et al. Travelmedia: An intelligent management system for media captured in travel
Souvannavong Indexation et recherche de plans vidéo par le contenu sémantique
Ryu et al. A priority queue-based hierarchical photo clustering method using photo timestamps

Legal Events

Date Code Title Description
DPE2 Request for preliminary examination filed before expiration of 19th month from priority date (pct application filed from 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 2005856532

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

WWP Wipo information: published in national office

Ref document number: 2005856532

Country of ref document: EP