KR101629832B1 - Chronological multimedia database construction device and method using a face image - Google Patents

Abstract

Disclosed is an apparatus for chronologically constructing a multimedia database by using a face image. The apparatus for chronologically constructing a multimedia database comprises: an input unit which receives a plurality of images including face images; a control unit which chronologically constructs a multimedia database based on the received images; and a database which chronologically classifies and stores multimedia data. The control unit comprises: a face detection unit which detects faces from the received images; a face classification unit which groups images including the same faces as the detected faces and tags the grouped images; a chronological alignment unit which chronologically aligns the tagged images; and a database construction unit which receives multimedia data corresponding to a chronicle of the aligned images from the database and groups the received multimedia data.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multimedia database construction device and method using a face image,

The present invention relates to an apparatus and method for constructing a multimedia database, and more particularly, to an apparatus and method for constructing a multimedia database for each generation using face images.

2. Description of the Related Art In recent years, interest in multimedia contents including large-sized moving pictures and voice has been increasing due to expansion of computer capacity, activation of Internet use, and spread of high-speed communication networks.

Accordingly, search service providers who have provided search services only to portions of conventional text have recently begun to provide search services for multimedia contents.

In particular, as the amount of multimedia data increases, it is required to construct a database for multimedia data and to provide a search service suitable for a user by using the constructed database.

Korean Patent Registration No. 10-0754157 relates to a method for constructing a database for multimedia contents in which multimedia contents are semantically categorized and stored in a predetermined database.

Korean Patent Registration No. 10-0518724 relates to a method and system for constructing a multimedia information database by indexing multimedia information and a method and system for providing a search service using the multimedia information database.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an apparatus and method for constructing a multimedia database for each age grouping images of faces identical to detected faces in a received image and using face images for tagging grouped images, respectively.

It is another object of the present invention to provide an apparatus and method for constructing a multimedia database for each age using a face image for sorting face images by grouping images.

It is another object of the present invention to provide an apparatus and method for constructing a multimedia database for each age using a face image for building a database for each generation by linking multimedia data of the same age to face images arranged by dates.

In order to achieve the above object,

An apparatus for constructing a multimedia database for each age using a face image according to the present invention includes an input unit for receiving a plurality of images including a face image, a controller for constructing a multimedia database for each generation based on the received images, Wherein the control unit includes a face detecting unit for detecting a face in the received images, a grouping unit for grouping the images including the same face as the detected face, and a DB constructing unit for receiving the multimedia data corresponding to the age of the sorted images from the database and grouping the received multimedia data.

According to another aspect of the present invention, there is provided a method of constructing a multimedia database for each age using a face image, the method comprising: receiving a plurality of images including a face image; detecting a face from the received images; The method includes the steps of grouping images, tagging the grouped images, sorting the tagged images by age, and extracting multimedia data corresponding to the age of the sorted images, And receiving and grouping them from the classified and stored database.

An apparatus and method for constructing a multimedia database for each age group using facial images according to the present invention can group images of the same face as the detected face in the received image and tag each of the grouped images.

Also, the grouped images can be sorted by age.

It is also possible to construct a database for each age by linking the multimedia data of the same age as the face image sorted by age.

1 is a block diagram for explaining a database building apparatus according to an embodiment of the present invention.
2 is a block diagram illustrating a control unit according to an embodiment of the present invention.
3 is an exemplary view for explaining a control unit according to an embodiment of the present invention.
4 is an exemplary view for explaining a face detecting unit according to an embodiment of the present invention.
5 is a diagram illustrating an example of a face classifying unit according to an embodiment of the present invention.
FIG. 6 is an exemplary view for explaining a face classifying unit according to another embodiment of the present invention.
FIG. 7 is an exemplary diagram illustrating an arrangement unit according to an embodiment of the present invention. Referring to FIG.
8 is a flowchart for explaining a database building method according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals as used in the appended drawings denote like elements, unless indicated otherwise. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather obvious or understandable to those skilled in the art.

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a personalized database using a face image, and more particularly, to an apparatus and a method for constructing a personalized multimedia database for each age by using an image including a face image.

1 is a block diagram for explaining a database building apparatus according to an embodiment of the present invention.

Referring to FIG. 1, the database construction apparatus 1 groups images of the same face as the detected face in the received image, and performs tagging of each of the grouped images. The database construction apparatus 1 constructs a database for each age by sorting the grouped images by age and then adding multimedia data of the same age as the sorted face images. Thus, the database construction apparatus 1 can build a personalized database. The database construction apparatus 1 includes an input unit 110, a control unit 120, an output unit 130, and a database 140.

The input unit 110 receives a plurality of images including a face image. The input unit 110 may receive a plurality of images at predetermined intervals or receive them at the same time. The plurality of images may be BMP, JPG (JPEG), GIF, EPS, TIF (TIFF), PSD, PIC, DXF. PP3, PICT, PNG, Desktop Color Separations (DCS), FlashPix, PCX, PDF, PIXAR, Raw, Scitex CT (sct), Targa. In particular, the plurality of images may include time information, location information, description information, and the like.

The control unit 120 constructs a multimedia database for each generation based on the images received from the input unit 110. The control unit 120 detects a face in the received images, groups the images including the same face as the detected face, and performs tagging on the grouped images.

The tagging allows a keyword or tag, which is a search tag that can represent the content of the content, to be classified in the form of a topic or a category.

The controller 120 arranges the tagged images according to their age, groups additional multimedia data for each of the sorted images, and constructs a customized database.

The output unit 130 outputs the multimedia database to the user. The output unit 130 includes a monitor, a projector, a printer, a recording medium, and the like. In particular, when the output unit 130 is a recording medium, the multimedia database may be stored in the recording medium.

The storage unit 140 stores a plurality of images received from the input unit 110 and stores a multimedia database for each generation established in the control unit 120. [ As a result of performing the control unit 120, for example, the detected face, the tagged image, and the like are temporarily stored in the storage unit 140.

The database 150 classifies and stores the multimedia data by age. The database 150 may receive multimedia data from at least one of a broadcast medium, an Internet medium, and a social network (SNS). In particular, the database 150 receives the multimedia data every predetermined period and updates the latest data.

FIG. 2 is a block diagram for explaining a control unit according to an embodiment of the present invention, and FIG. 3 is an exemplary view for explaining a control unit according to an embodiment of the present invention.

2 and 3, the control unit 120 includes a face detecting unit 210, a face classifying unit 220, a sorting unit 230, and a DB building unit 240.

The face detection unit 210 detects a face from a plurality of images received by the input unit 110. The face detection unit 210 performs face detection using at least one algorithm such as Haar-like and Convolutional Neural Network (CNN). That is, the face detecting unit 210 may use an algorithm combining characteristics of the HARQ and the convolutional neural network. The convolutional neural network algorithm shows excellent performance in big data processing.

Accordingly, the face detection unit 210 can classify whether the image is a personal image or a group image through face detection.

The face classifying unit 220 groups the images including the same face as the face detected by the face detecting unit 210.

The face classifying unit 220 classifies the same face using commercial face recognition software. Accordingly, the face classifying unit 220 may detect an image including the same face among a plurality of images, classify the detected images, and classify them automatically. If the state of the image is not good and the face can not be detected, the face classifying unit 220 can manually classify the face according to the user's input.

The face classifying unit 220 performs tagging on the grouped images. The face classifying unit 220 can make the face classification clear by performing tagging for each group.

The age-based sorting unit 230 arranges the images tagged in the face classification unit 220 by age. The age-specific sorting unit 230 arranges the images on the basis of at least one of the texture, the wrinkles, and the information included in the images of the skin region. The information included in the image may be time information, location information, explanatory information, and the like.

The aged sorting unit 230 may divide ages according to a predetermined period. The predetermined period may be 5 to 10 years.

The DB constructing unit 240 receives the multimedia data corresponding to the age of the sorted images in the sorting unit 230 according to the ages from the database 150 and groups them. The DB building unit 240 can link the age of the sorted images and the age of the database 150 to group them. That is, when the grouping is performed, the DB construction unit 240 may group only the addresses to be linked in the database 150, thereby preventing overlapping data from being generated. Thus, the link to the address has the effect of reducing the capacity for redundant data.

4 is an exemplary view for explaining a face detecting unit according to an embodiment of the present invention.

Referring to FIG. 4, the face detection unit 210 automatically detects the same face in a plurality of images. The face detection unit 210 may use the HARLEQUE of the face detection algorithm.

The HARQ is an algorithm for expressing feature values using a simple sum image. This has the effect of having many shapes depending on the position, shape, and size.

The face detection unit 210 detects a face by setting a window for one face when the image includes one face. When the image includes a plurality of faces, the face detecting unit 210 detects a face by setting a window on each face. At this time, the face detection unit 210 may temporarily store the detected faces of the window in the storage unit 140. [

FIG. 5 is an exemplary view for explaining a face classifying unit according to an embodiment of the present invention, and FIG. 6 is an exemplary view for explaining a face classifying unit according to another exemplary embodiment of the present invention.

Referring to FIGS. 5 and 6, the face classifying unit 220 groups images including the same face among a plurality of images, and performs tagging on the grouped images.

The face classifier 220 classifies the same face based on Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA).

The principal component analysis is a so-called multivariate analysis that analyzes various variables, and is a method of extracting several principal components from many variables. The linear discriminant analysis is a method of maximizing the ratio of between-class scatter and within-class scatter as one of feature vector dimension reduction methods.

The face classifying unit 220 classifies the images including the same face using the face recognition SDK. The face classifying unit 220 classifies the same face using VeriLook.

The face classifying unit 220 performs tagging for a person classified as the same face. The face classifier 220 tags the manuals of the grouped images. The face classifying unit 220 can more easily search the grouped images later through tagging.

FIG. 7 is an exemplary diagram illustrating an arrangement unit according to an embodiment of the present invention. Referring to FIG.

Referring to FIG. 7, the age-based sorting unit 230 sorts the grouped images according to their age. The age-specific sorting unit 230 can arrange images in ascending or descending order.

The age-specific sorting unit 230 can measure the age of an image by using image features such as texture and wrinkles of the skin region of the face, and can measure the age of the image using the information included in the image . The information included on the image may be time information, location information, explanatory information, and the like.

The age-specific arranging unit 230 may arrange the ages according to a predetermined period.

Life cycle Age group Life cycle Age group Infant and child 0 ~ 5 years old (before entering elementary school) Old age 30 to 49 years old (30/40) Childhood 6 ~ 12 years old (elementary school) Middle age 50 to 69 years old (50/60) Adolescence 13 ~ 19 years old (middle / high school) Old age 70 to 84 years adolescence 20 to 29 years old (20s) Super-ordained 85 years old or older

The sorting unit 230 according to the ages can be sorted into at least one age group such as infant, child, adolescence, adolescence, adulthood, middle age, old age, and first age as shown in Table 1 above.

8 is a flowchart for explaining a database building method according to an embodiment of the present invention.

Referring to FIG. 8, the database construction method constructs a multimedia database for each generation by using images including face images. The database construction method can build a multimedia database for each generation according to the following procedure.

In the first step, an image including a face image is received (S100). The first step is to receive a plurality of images including face images. In the first step, a plurality of images can be received at predetermined intervals or simultaneously.

In the second step, the face included in the image is detected (S110). The second step performs face detection using at least one algorithm of the HARQ and convolutional neural network in the images received in the first step.

Accordingly, the second step can classify whether the image is a personal image or a group image through face detection.

In the third step, the images including the same face as the detected face are classified and tagged (S120). In the third step, the same face is classified using the commercial face recognition software for the face detected in the second step. In the third step, after detecting an image including the same face among a plurality of images, the detected images may be grouped and automatically classified.

The third step is to tag the grouped images. In the third step, the classification of the face can be clarified by tagging the explanation to each group.

In the fourth step, the tagged images are sorted by age (S130). In the fourth step, images tagged in the third step are sorted by age. In a fourth step, the images may be sorted by age, based on at least one of texture, wrinkles, and information contained in the image of the skin area.

In the fifth step, a database is constructed for the sorted images for each generation (S140). In the fifth step, the sorted images stored in the database 150 are linked to the sorted images in the fourth step, and then the images are grouped to construct a database for each generation.

The database construction method can be used when information corresponding to a certain age is needed for an arbitrary photograph in the future. Particularly, since the database construction method stores the information of each generation in the form of multimedia data, the user can easily obtain information about the age at that time.

The present invention can also be embodied as computer-readable codes on a computer-readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer apparatus is stored. Examples of the computer-readable recording medium include a hard disk, a ROM, a RAM, a CD-ROM, a hard disk, a magnetic tape, a floppy disk, an optical data storage device, And the like.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is clearly understood that the same is by way of illustration and example only and is not to be taken by way of limitation in the embodiment in which said invention is directed. It will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the scope of the appended claims.

1: Database building device
110: input unit
120:
130:
140:
150: Database
210:
220: face classification unit
230:
240: DB construction unit

Claims (9)

An input unit for receiving a plurality of images including a face image;
A controller for constructing a multimedia database for each generation based on the received images; And
And a database in which multimedia data is categorized and stored by age,
Wherein,
A face detector for detecting a face in the received images;
A face classifying unit for grouping the images including the same face as the detected face and tagging the grouped images;
A sorting unit for sorting the images based on at least one of a texture, a wrinkle, and a time, position, and description information included in the image; And
And a DB constructing unit for receiving the multimedia data corresponding to the age of the sorted images from the database and grouping them and grouping only the addresses to be linked in the database so as to prevent duplicated data from being generated. An apparatus for constructing a multimedia database for each generation using video images.
The method according to claim 1,
The database includes:
Wherein the multimedia data is received from at least one of a broadcasting medium, an Internet medium, and a social network (SNS).
The method according to claim 1,
Wherein the face detection unit comprises:
Wherein the face detection is performed using at least one of Haar-like and Convolutional Neural Network (CNN).
The method according to claim 1,
Wherein,
Wherein the images are classified into a group including images of the same face using a face recognition SDK and the grouped images are generated as a group.
The method according to claim 1,
Wherein,
And tagging a manual for the grouped images.
delete The method according to claim 1,
Wherein the sorting-
Wherein the plurality of images are arranged in at least one of an age group of at least one of infant, child, child, adolescent, adolescent, adulthood, middle age, old age and first generation age.
The method comprising: receiving a plurality of images including an input image;
Detecting a face in the received images by a face detection unit in a control unit;
Grouping the images including the same face as the detected face in the control unit and tagging the grouped images;
Arranging the images by age according to at least one of a texture, a wrinkle and a time, position, and explanatory information included in the image; And
The DB constructing unit receives the multimedia data corresponding to the age of the sorted image from the database in which the multimedia data is sorted and stored for the respective ages and groups the addresses in the database so that no duplicated data is generated A method for constructing a multimedia database for each age using a face image including a face image.
A computer-readable recording medium storing a program for causing a computer to execute the method according to claim 8.

Images