JP2007293399A - Image exchange device, image exchange method, and image exchange program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image exchange device, image exchange method, and image exchange program, capable of exchanging images by efficiently selecting the images in a system designed to share a large number of images among a plurality of users by using a network. <P>SOLUTION: The image exchange device identifies whether a face of a person is included in each image of image data, on the basis of acquired image data and detects a face zone as an existing zone of the face of the person, if an image includes the face of the person (114), recognizes the person related to the face in the detected face zone (115), determines the person having such a face in the face zone (116), and distributes the image data to a designated distribution address (117), in response to the determined results. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

The present invention relates to an image exchanging apparatus, an image exchanging method, and an image exchanging processing program using a network. An image for enabling easy execution of exchanging images by distribution while sharing digital images among a plurality of users. The present invention relates to an exchange device, an image exchange method, and an image exchange processing program.

A method of enjoying digital images stored in a predetermined server on a network and shared and viewed among a plurality of users is becoming widespread. For this, there are known a method of sharing a digital image storage destination URL by e-mail notification, etc., and a network print system for ordering prints of image data to a lab store connected to an image server. ing.
In addition, a peer-to-peer network is used to build a system that allows images to be shared between limited users, and no upload procedure is required, and anyone can easily print the desired finished quality at the nearest lab store. Proposals have also been made to provide available departments.

In this proposal, as a user terminal device used in an image print system, image data is acquired by an image acquisition unit from another user terminal device connected by peer-to-peer, and the acquired image is browsed by an image display unit. On the other hand, image data is stored in the image storage unit, and at least one of the images stored in other user terminal devices and the images stored in the image storage unit is selected and printed. Has been disclosed (see Patent Document 1).

In addition, for example, a terminal such as a mobile phone or a PDA in which special display software for displaying an electronic file of the original file format of the electronic photograph information is not installed on the electronic photograph information taken by the digital camera There are also specific proposals for technologies that enable browsing in common on the device.
For example, electrophotographic information photographed by a digital camera is received via a network by a personal computer as a plurality of terminal devices, the electrophotographic information is converted into a JPEG format that can be displayed by a Web browser, and the converted electrophotography The information is stored at a predetermined URL address, and can be browsed by a normal PDA or the like on the Internet or an intranet (see Patent Document 2).
JP 2004-259249 A (paragraph 0035, FIG. 1) Japanese Patent Laying-Open No. 2003-141028 (paragraph 0022 to paragraph 0028, FIG. 1)

However, if the total amount of image data to be handled gradually increases, it is clear that visual confirmation work becomes impossible early and night, and it is requested that some measures be taken as a practical problem. ing.
In Patent Document 1 and Patent Document 2, there is no particular recognition about the above-mentioned realistic and important problems, and naturally there is no specific proposal for dealing with this problem.

The present invention has been made in view of the above situation, and in a system constructed so that a large amount of images can be shared among a plurality of users using a network, images can be efficiently selected. An object is to provide an image exchange apparatus, a method, and a program that can be exchanged.

In order to solve the above-described problems, the present application proposes the following techniques.
(1) Based on the acquired image data, it is determined whether or not each image based on the image data includes a person's face, and for an image including the person's face, an area in which the person's face exists A face detection unit that detects a face area, a face recognition unit that recognizes a person related to a face in the face area detected by the face detection unit, and a predetermined distribution destination for each person recognized by the face recognition unit An image exchange apparatus comprising: a distribution control unit that distributes image data of an image including a face corresponding to the above.

In the image exchange apparatus according to (1), the face detection unit identifies whether or not a person's face is included in each image based on the acquired image data based on the acquired image data, and the person's face is included. For an image, a face area is detected as an area where the person's face exists. The face recognition unit recognizes a person related to the face of the detected face area. The distribution control unit distributes image data of an image including a face corresponding to a predetermined distribution destination for each recognized person.
Since distribution destinations according to the person's face are automatically sorted and distributed, a large amount of images including the face can be efficiently distributed to an appropriate distribution destination.

(2) The distribution control unit stores copy data of each corresponding image data in a plurality of folders assigned for each person, and then stores the copy data stored in the folder for a predetermined distribution destination for each person. (1) The image exchange device according to (1),
In the image exchange device of (2) above, the distribution control unit in the image exchange device of (1)
Store copy data of each corresponding image data in multiple folders assigned for each person,
Next, since the copy data stored in the folder is distributed to a predetermined distribution destination for each person, many images can be efficiently transferred for each distribution destination.

(3) Based on the acquired image data, it is determined whether or not each image based on the image data includes a person's face, and for an image including the person's face, an area in which the person's face exists And detecting a face area as a person, recognizing a person related to the face of the detected face area, and distributing image data of an image including a face corresponding to a predetermined distribution destination for each recognized person. Image exchange method.

In the image exchanging method of (3) above, whether or not a human face is included in each image based on the acquired image data is identified, and the image including the human face is An image of an image including a face corresponding to a predetermined distribution destination for each recognized person by detecting a face area as an area where a human face exists, recognizing the person related to the face of the detected face area Since data is distributed, a large amount of images including a face can be efficiently distributed to an appropriate distribution destination.

(4) In the distribution, copy data of each corresponding image data is stored in a plurality of folders assigned for each person, and then the copy data stored in the folder is stored in a predetermined distribution destination for each person. (3) The image exchange method according to (3), wherein the image exchange method is distributed.
In the image exchanging method of (4), particularly in the image exchanging method of (3), at the time of distribution, copy data of each corresponding image data is stored in a plurality of folders assigned for each person, and then the folder Since the copy data stored in is distributed to a predetermined distribution destination for each person, many images can be efficiently transferred for each distribution destination.

(5) Based on the acquired image data, it is identified whether or not a person's face is included in each image based on the image data, and for an image that includes a person's face, an area in which the person's face exists A face detection processing step for detecting a face area, a face recognition processing step for recognizing a person related to a face in the face area detected in the face detection processing step, and a person recognized in the face recognition processing step An image exchange processing program, comprising: an instruction for causing an information processing apparatus to execute a distribution control processing step for distributing image data to a predetermined distribution destination for each person.

When the information processing apparatus is operated by the image exchange processing program of (5) above, it is determined whether or not a human face is included in each image based on the acquired image data based on the acquired image data. For an image including a face, a face detection processing step for detecting a face region as a region where the person's face exists, and a face for recognizing a person related to the face in the face region detected in the face detection processing step The processing of each processing step is executed, including a recognition processing step and a distribution control processing step for distributing image data including the person recognized in the face recognition processing step to a predetermined distribution destination for each person.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the drawings to be referred to below, for the sake of convenience, the main part that is the subject of the description is exaggerated as appropriate, and other than the main part is appropriately simplified or omitted.
FIG. 1 is a diagram illustrating a network including an image exchange apparatus according to an embodiment of the present invention.

In the network 100 of FIG. 1, an image exchange device 110 according to an embodiment of the present invention includes a storage device 121 for a user of the image exchange device 110 to store image data or other various data. Is connected.
The image exchange device 110 is also connected to a panel display device 122 that displays images stored or processed (or processed) by itself and that can also display other images. The image exchange device 110 further transmits the server 2 through the Internet 200.
11 and a web terminal 210 including a storage device 212. The storage device 212 stores personal data of each of a plurality of users who exchange images with each other, as will be described later.

The image exchange device 110 is further connected to a mobile phone 301 and a PDA (personal digital assistance) 302 owned by each user by a known wireless communication transmission path. Also, since it is well known, the details of the connection route are omitted, but LAN, WAN
The image data can be exchanged with the personal computer 303 via the Internet 200 and further with the digital camera 304 via the personal computer 303. According to the technical idea of the present invention, the mobile phone 301, PDA 302, personal computer 303, and digital camera 304 receive and store images received from the image exchange device 110, or receive images for each user who views them. It functions as the terminal 300.

The image exchanging apparatus 110 sequentially applies a large number (or an enormous number) of images uploaded and accumulated from a plurality of users in a database constructed in the storage device 212 of the specific web terminal 210 as necessary. Capture images, extract images containing the face image of a predetermined user (scheduled (or registered)) from these images, and evaluate “images” one by one according to a predetermined standard. Each user's image corresponding to an image that satisfies the above evaluation criteria and that wants to be distributed (for example, an image of himself or herself or a person of interest as a subject) Delivered to the receiving terminal 300.

FIG. 2 is a block diagram conceptually showing the internal configuration of the image exchange apparatus 110 in FIG. The image exchange device 110 includes functions of an image input unit 111, an image storage unit 112, an image reading unit 113, a face detection unit 114, a face recognition unit 115, an image classification unit 116, a distribution control unit 117, and a display control unit 118. This unit is configured to be connected to a system controller 140 that comprehensively controls these functional units through a bus 130.

The image exchange apparatus 110 is configured as one ASIC (Application Specific Circuit), each of which is a single ASIC (Application Specific Circuit) or a combination of some of the above-described functional units. Can be configured to cover the required functions.
However, the configuration of the image exchange device 110 is not limited to this, and all the functions of the above-described functional units can be realized as an algorithm by the application program of the present invention in one computer.

The image input unit 111 has a function of capturing an image designated by the user. Image storage unit 112
Has a function to temporarily store an image designated by the user. The system constructed as shown in FIG. 1 can function to temporarily store an image in the storage device 121.
The image reading unit 113 functions to read the temporarily stored images one by one in order to perform each process after the next step.

The face detection unit 114 functions to detect a face area by analyzing an image. A face detector that collects a large number of face images in advance and identifies an image of a specific size as a face image and an image other than a face based on the collected face images is configured.
In order to identify a face, for example, an image feature amount generated from luminance or an edge is used. Also,
The face detector can adopt a known configuration to which a statistical technique such as a multilayer neural network, a support vector machine, or discriminant analysis is applied.

As a method other than the above, "" Detection of face independent of background and face direction and estimation of face direction "IEICE Technical Report, PRMU2001-217, pp. 87-94 (2002-01), Yuichi Araki, Shin Shimada The method described in the literature, “Rei, Yoshiaki Shirai” can be applied. Details of this method can be found by referring to the document itself, but the outline is to obtain the facial part with probability, detect the face using the relaxation method, and use the part detection result. Then, the direction of the face is detected. When applied to the apparatus of the present invention, only the face detection method is used, and face direction detection is not required.

The face recognition unit 115 recognizes who the face detected by the face detection unit 114 is. A face image is collected in advance, and an individual can be specified by a statistical technique such as discriminant analysis based on the collected face image. In addition, a classifier that identifies an individual can be realized by learning using a neural network. As documents disclosing methods other than the above, there are the following patent documents. That is, in Japanese Patent Application Laid-Open No. 9-251534, “Person Authentication Device and Method”, feature points such as eyes and nose are stably extracted from an input image without being affected by changes in luminance, etc. The section to do is described.

The image classification unit 116 classifies the image into folders corresponding to persons (users) included in the image in order to distribute the image to the user whose face is recognized by the face recognition unit 115. That is, by this process, the image is associated with the classification destination folder.
The distribution control unit 117 classifies (sorts) each corresponding folder by the image classification unit 116.
Each image classified and stored in a folder by processing is stored in the mobile phone 301, PDA 302, personal computer 303, or digital camera 30 corresponding to these folders.
It is distributed to the image receiving terminal 300 (FIG. 1) of each user such as 4.
The display control means 118 arranges the image layout so that it can be visually recognized by the user and displays it on the output device (in the case of the system of FIG. 1, the panel display device 122).

FIG. 3 is a flowchart showing a flow of operations related to face recognition face image registration.
For the (face recognition face) image stored by the user, the face recognition face image registration process described below is performed. First, when there is an unprocessed image in the stored image (step S301: YES)
), The process proceeds to the next step S302. If there is no unprocessed image (step S301)
: NO), the registration process is terminated. In step S302, images are captured one by one from images stored by the user. For each image captured in step S302, the face detection unit 114 detects the face in the image (step S303). At this time, the detected face is positioned (x, y coordinates).
And the size including the outline of the face.

In response to the face detection result (position, size, etc.) in step S303, face parts such as eyes and nose are detected (step S304). By detecting the face part, the feature amount extraction accuracy in the next step S305 is improved. That is, in step S305, the face recognition unit 115 extracts a feature amount for specifying an individual. In addition, the color of a person's skin is expressed in a color space (such as an HSL color system), and a value is recorded. Next, the face recognition registration image (the feature amount extracted in step S305) is associated with the person information, and the information added when the person is specified is associated.

FIG. 4 is a flowchart showing a flow of operations relating to face detection and face recognition.
The face detection / face recognition processing described below is performed on the image stored (desired) by the user. If there is an unprocessed image in the stored image (step S401: YES), the process proceeds to the next step S402. If there is no unprocessed image (step S301: NO), the process ends. In step S402, images are captured one by one from images registered by the user.

Next, the face detection unit 114 detects a face in the image for each captured image (step S403). At this time, the detected face is recorded with the position (x, y coordinates), the size including the face outline, and the like.
As a result of executing the detection process in step S403, if a face is present in the image (step S404: YES), the process proceeds to the next step S405. If no face is present in the image (step S404: NO), the next image is read (return to step S401).

In response to the face detection result (position, size, etc.) in step S403, face parts such as eyes and nose are detected (step S405). By detecting the face part, the accuracy of the person identification process in the next step S406 is improved.
In step S406, it is recognized who the detected face is. That is, the face recognition unit 115 is based on the face images of users registered in advance as in the flowchart of FIG.
To identify the individual. Further, by using the feature points such as the face part detected in step S405, it is possible to perform highly accurate person authentication. If the identified person is a previously registered user face (step S406: YES), the process proceeds to the next step S407.

If it is not the user's face registered in advance (step S406: NO), the process proceeds to step S408. That is, for the image that could not be specified by the user registered in advance by the face recognition unit in the determination processing in step S406, a message is output in step S408 to notify the user of that fact. When the work is complete, load the next image (
(Return to step S401). In step S407, the specified person is recorded, and the next image is read (return to step S401).

FIG. 5 is a flowchart showing a flow of operations relating to image classification and distribution control.
In response to the face recognition result, the reflection determination / distribution control described below is performed. If there is an unprocessed image in the stored image (step S501: YES), the process proceeds to step S502. On the other hand, if there is no unprocessed image (step S501: NO), the process is terminated.
In step S502, the face detection result (position, size, etc.) in step S405 is read. Next, person data is added to the image based on the face recognition result in step S406 described above (step S503).

Further, image classification is executed for each person assigned in step S503. That is, the image is copied to the folder for each individual according to the folder name corresponding to the assigned person data (step S504).
Next, the content of the person data given in step S503, and accordingly step S504.
The distribution control to the user's terminal is performed in accordance with the sorting into the folder for each individual in (Step S).
505).

FIG. 6 is a diagram illustrating an example of a table related to face recognition image management. This table
For example, it corresponds to a table storing results obtained by the face recognition face image registration operation shown in the flowchart of FIG. “Person ID” is an identifier when an individual is specified, “Face recognition feature value” is a feature value used by the face recognition means, and “Folder name” is a folder name for storing the result of image classification (the user terminal of the distribution destination If it does not exist, there will be no folder name).

FIG. 7 is a diagram illustrating an example of a table related to image management. This table corresponds to, for example, a table for storing results obtained by image classification and distribution control in the flowchart of FIG. “Image ID” is an identifier assigned to an image stored (organized) by the user,
“Person ID” is an identifier of an individual specified by receiving the face recognition result, and a “destination control” column is recorded with a copy destination folder name obtained by the image classification unit 116 of FIG.

FIG. 8 is a diagram illustrating an image when images are assigned to folders for each individual. Here, it is assumed that the face recognition unit 115 has determined that the image ID 0001 includes three persons IDs A0001, A0002, and A0005 as subjects. Therefore, in order to distribute to users A0001 and A0002, images are copied to folders A and B corresponding to these users, respectively.

In addition, it is assumed that the face recognition unit 115 determines that two persons, person IDs A0002 and A0004, are captured as subjects in the image ID 0002. Therefore, user A0
Folder B corresponding to each of these users in order to distribute to 002 and A0004
And copy the image to folder D.

FIG. 9 is a diagram illustrating an example of a view screen during image browsing. That is, it represents the display status when the result of image classification is displayed on the monitor of the image exchange apparatus. Images are sorted into folders for each person based on the face recognition results. After confirming whether or not to save the result of classification and arrangement in each user's terminal (FIG. 1: 300), sort into folders for each person. Delivered image set.

It is a figure showing the network containing the image exchange apparatus as one embodiment of this invention. FIG. 2 is a block diagram conceptually showing the internal structure of the image exchange device in FIG. 1. It is a flowchart which shows the flow of operation regarding face image registration for face recognition. It is a flowchart which shows the flow of operation regarding face detection and face recognition. It is a flowchart which shows the flow of operation regarding image classification and delivery control. It is a figure which shows an example of the table regarding the image management for face recognition. It is a figure which shows an example of the table regarding image management. It is a figure showing the image at the time of distributing an image to the folder for every individual. It is a figure which shows an example of the view screen at the time of image browsing.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Network 110 ... Image exchange apparatus 111 ... Image input part 112 ... Image storage part 113 ... Image reading part 114 ... Face detection part 115 ... Face recognition part 116 ... Image classification part
117 ... Distribution control unit 118 ... Display control unit 140 ... System controller 122 ... Panel display device 200 ... Internet 210 ... Web terminal 211 ...
Server 212 ... Storage device 300 ... Image receiving terminal 301 ... Mobile phone 302 ... PD
A 303 ... Personal computer 304 ... Digital camera

Claims (5)

Based on the acquired image data, each image based on the image data identifies whether or not a person's face is included, and for an image that includes a person's face, the face as an area where the person's face exists Corresponds to a face detection unit for detecting an area, a face recognition unit for recognizing a person related to the face in the face area detected by the face detection unit, and a predetermined distribution destination for each person recognized by the face recognition unit An image exchange device comprising: a distribution control unit that distributes image data of an image including a face. The distribution control unit stores copy data of each corresponding image data in a plurality of folders allocated for each person, and then distributes the copy data stored in the folder to a predetermined distribution destination for each person. The image exchange device according to claim 1, wherein the image exchange device is configured as described above. Based on the acquired image data, each image based on the image data identifies whether or not a person's face is included, and for an image that includes a person's face, the face as an area where the person's face exists An image characterized by detecting an area, recognizing a person related to the face in the detected face area, and distributing image data of an image including a face corresponding to a predetermined distribution destination for each recognized person method of exchange. In the distribution, copy data of each corresponding image data is stored in a plurality of folders assigned to each person, and then the copy data stored in the folder is distributed to a predetermined distribution destination for each person. The image exchange method according to claim 3. Based on the acquired image data, each image based on the image data identifies whether or not a person's face is included, and for an image that includes a person's face, the face as an area where the person's face exists A face detection processing step for detecting a region, a face recognition processing step for recognizing a person related to the face in the face region detected in the face detection processing step, and image data including the person recognized in the face recognition processing step. An image exchange processing program, comprising: an instruction for causing an information processing device to execute a distribution control processing step for distributing to a predetermined distribution destination for each person.

Images