JP2009232022A - Image searching apparatus and method, image reproducer and reproduction method, image searching program, and image reproduction program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide image searching device and method which can search an image including many figures efficiently from an image group, image reproducer and reproduction method for reproducing an image searched by the image searching device and method, and an image searching program and an image reproduction program being run on a computer. <P>SOLUTION: An image searching device includes a section 22 for specifying an individual by detecting the face image of a figure included in each of a plurality of images from the plurality of images, and a section 25 for searching an image from the plurality of images such that the number of images is minimized while including all figures included in the plurality of images based on the specification results of an individual from the section 22 for specifying an individual. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image search device, an image search method, an image playback device, and an image playback method, and more particularly, to an image search device and an image search method for searching for one or more representative images from a plurality of captured images. The present invention relates to an image playback apparatus, an image playback method, an image search program to be executed by a computer, and an image playback program.

  In recent years, the capacity of recording media has been increasing due to technological innovation of digital cameras. Compared to cameras in the film era, the continuous shooting function is also greatly improved, so that camera users can enjoy taking pictures easily and obtain a large amount of images.

  By the way, there are opportunities for participants to view images taken at an event or a trip at the same time. In such a case, it is more enjoyable to view images containing as many people as possible. However, if an image containing as many people as possible is searched for and viewed in a random manner from a large number of images, it takes time, and there is a risk that a pleasant atmosphere will be created. Also, after an event or a trip, I sometimes miss a person who was participating. Even in such a case, it takes time to search a large number of images randomly.

As a search for a plurality of image objects, Patent Document 1 proposes an image display method for displaying an image showing a desired object from among a plurality of images. That is, in the image display method described in Patent Document 1, the type and number of objects such as a person shown in the image are recorded as tag information for each image, and the type and number of objects are input at the time of search. By doing so, a desired image can be quickly searched.
JP 2006-261914 A

  In the image display method disclosed in Patent Document 1, it is necessary to input the number of objects as a search condition, and it is difficult to find an image in which as many people as possible are captured. In other words, in the images taken in the past, the number of people is often not stored accurately. It takes a long time to search.

  The present invention has been made in view of such circumstances, and an image search apparatus, an image search method, an image search apparatus, and an image search capable of efficiently searching an image including a large number of persons from a plurality of images. It is an object of the present invention to provide an image reproduction apparatus that reproduces an image searched by a method, an image reproduction method, an image search program to be executed by a computer, and an image reproduction program.

  In order to achieve the above object, an image search apparatus according to a first aspect of the present invention is a personal identification unit that detects a face image of a person included in each of a plurality of images and identifies a person from the plurality of images. And an image search for searching for an image from the plurality of images so as to include all persons included in the plurality of images and to minimize the number of images based on a result of personal identification by the individual identification unit. A section.

  An image search and display device according to a second invention is the image search and display device according to the first invention, wherein the image search device further includes an image classification unit for classifying images according to the type of image, and the plurality of images are the image classification Belongs to one of the image groups classified by division.

  According to a third aspect of the present invention, there is provided an image search method for detecting a face image of a person included in each of the plurality of images from among a plurality of images, identifying the individual, And searching for images from the plurality of images so as to include all persons included in the plurality of images and to minimize the number of images.

  An image reproduction apparatus according to a fourth invention is an image reproduction apparatus for reproducing an image searched by the image search apparatus according to the second invention, a display unit for displaying information representing the type of the image, and the display A selection unit that selects desired information from information representing the type of image displayed on the unit, and an image display unit that displays the image searched by the image search unit in response to an operation of selecting the selection unit. .

  An image reproduction method according to a fifth invention is an image reproduction method for reproducing an image searched for by the image search device according to the second invention, the step of displaying information representing the type of the image, A step of selecting desired information from the information indicating the type, and a step of displaying the searched image in response to an operation of selecting the desired information.

  An image search program according to a sixth aspect of the present invention includes a step of detecting a face image of a person included in each of a plurality of images and specifying an individual out of the plurality of images, and the result of specifying the individual Based on the above, the computer is caused to search for an image from the plurality of images so as to include all persons included in the plurality of images and to minimize the number of images.

According to a seventh aspect of the invention, there is provided a program for reproducing an image, the step of detecting a face image of a person included in each of the plurality of images from a plurality of images and specifying an individual, and the result of specifying the individual A step of searching for an image from the plurality of images so as to include all persons included in the plurality of images and minimizing the number of images, and a step of displaying the searched images; Is executed on the computer.
According to an eighth aspect of the invention, in the seventh aspect of the invention, when searching for the image, the image reproduction program classifies the plurality of images and searches only for images belonging to the classification.

  An image retrieval program according to a ninth aspect of the present invention is a personal identification step of identifying a person by detecting a face image of a person included in each of the plurality of images, and the personal identification Based on the individual identification result of the step, the step of extracting the first image including the most persons among the images included in the plurality of images and the person not included in the first image are the most. And causing the computer to execute a step of extracting a second image that is included in a large amount.

  According to the present invention, an image search apparatus, an image search method, an image search apparatus, and an image playback that reproduces an image searched by the image search method can efficiently search for an image including a large number of people from a plurality of images. An apparatus, an image reproduction method, an image search program to be executed by a computer, and an image reproduction program can be provided.

  Hereinafter, a preferred embodiment will be described using a camera to which the present invention is applied according to the drawings. The camera according to the present embodiment records the number of characters, the scene of the image, and the type of event in the captured image so that the number of characters is the largest in the image group determined by the type of scene or event. Search for one or more images and play them.

  FIG. 1 is a block diagram showing the configuration of a camera according to an embodiment of the present invention. This camera is a digital camera, and this camera has an imaging unit 11, an optical system control unit 12, a clock unit 13, a recording unit 14, an output unit 15, a face detection unit 16, and a control unit 20. It is connected to the data bus 10.

  The imaging unit 11 includes an optical system for forming a subject image, an imaging element that photoelectrically converts the subject image, and the like, and outputs image data. The optical system control unit 12 performs control such as focusing of the optical system in the imaging unit 11. The clock unit 13 has a timekeeping function and a calendar function, and is used to acquire date and time information at the time of shooting.

  The recording unit 14 records the compressed image data output from the imaging unit 11 as a captured image. The date and time information obtained by the clock unit 13 is recorded as a file together with the photographed image when the photographed image is recorded in the recording unit 14. Furthermore, the determination results in the face detection unit 16, the scene classification unit 22, the individual identification unit 24, and the event classification unit 25, which will be described later, are recorded in the recording unit 14 as information files related to images. The information file is tabulated, and (1) who is reflected in each image and how many people are captured, (2) information on a group of images for each event, and (3) information such as the scene type of the event Is recorded. Details will be described later with reference to FIG.

  The output unit 15 outputs image data or the like to an external device such as a printer, a television, or a personal computer. An image searched by an image search unit 23 described later is also output from here. In the shooting standby state, the face detection unit 5 performs face pattern determination on image data for a through image (also referred to as a live view) output from the imaging unit 11 to detect a face part. Further, after photographing, the face detection information (number of detected persons) detected by the face detection unit 16 is recorded together with the photographed image in the information file.

  The control unit 20 performs overall control of the camera, and includes an image processing unit 21, an individual identification unit 22, a scene classification unit 23, an event classification unit 24, and an image search unit 25, except for the image processing unit 21. Performed by software. The image processing unit 21 performs image processing and compression processing of digital image data output from the imaging unit 11. The image data compressed here is recorded in the recording unit 14.

  Based on the face detection information detected by the face detection unit 16, the individual specifying unit 22 measures the feature amount of all the faces in the image and specifies the individual. Here, information about the number of people and who are in the image based on the specified information is recorded in the recording unit 14 as information in the information file. The specifying operation by the individual specifying unit 22 is executed when the camera is powered off.

  The event classification unit 24 determines which event the image belongs to. This event classification operation determines whether an image group captured during one activation is added to an existing event or set as a new event at power-off. The recording unit 14 records common information of a plurality of images in the information file for each event.

  The scene classification unit 23 detects the ratio of face detection, the deviation of the shooting date and time, and the light source for a plurality of images in units of events classified by the event classification unit 24, and based on the result, snap, trip, indoor event The scenes are classified into outdoor event types, and the classification information is recorded in the information file of the recording unit 14.

  The image search unit 25 searches for an image based on the information file recorded in the recording unit 14 so that all members appear from a plurality of images in units of events and the number of images is minimized. .

  The operation unit 32 is various operation members of the camera, such as a release button, a shooting mode or playback mode setting member, a cursor operation member or its determination operation member in the display unit, and is operated by a user. The operation determination unit 31 determines the operation state of the operation unit 32 by the user and outputs the determination result to the control unit 20. The control unit 20 controls the camera according to the operation state.

  The display unit 34 is composed of a display device such as a liquid crystal monitor disposed on the back surface of the camera, etc., receives an instruction from the control unit 20, performs reproduction display of the image data recorded in the recording unit 14, and optical Instead of the type finder, it also displays a through image (live view) for observing the subject image. In addition, based on the information recorded in the information file recorded in the recording unit 14, during playback, an image is displayed for each event (event number, shooting date and time, scene type). The display control unit 33 performs display control of the display unit 34.

  The operation of the thus configured camera will be described with reference to the flowchart shown in FIG. When the power switch is turned on and this flow is entered, it is first determined whether the photographer has set the shooting mode or the playback mode (S1). If the result of determination is that shooting mode has been set, face detection is performed by the face detector 16 (S3). Face detection is performed using image data of a subject for a through image acquired by the imaging unit 11.

  Next, the release button is operated, and it is determined whether or not to shoot (S5). If the result of determination in step S5 is that no shooting operation is to be carried out, processing returns to step S1 and the aforementioned operation is performed. In the present embodiment, when the shooting mode is selected and the shooting operation is not performed, the face detection is repeatedly performed using the subject image acquired by the imaging unit 11.

  If the result of determination in step S5 is that the release button has been operated, a shooting operation is executed (S7). In this step, imaging is performed by the imaging unit 11 and image data of a still image is acquired. When the image data of the captured image is acquired, the face detection information such as the number of faces included in the image detected by the face detection unit 16 in step S3 immediately before shooting is tabulated (S9), and the recording unit 14 records an information file including the captured image and face detection information (S11). In addition to the photographed image, various photographing information such as photographing date and time information output from the clock unit 13, white balance and use state of the auxiliary light source are also recorded in the recording unit 14.

  When the recording is finished, it is determined whether or not the power is off (S13). When the user performs a power-off operation, the main power supply of the camera is turned off, and the camera does not operate when an operation member other than the power switch is operated. However, the personal identification unit 22 and the like in the control unit 20 can operate even when the power is turned off. If the result of determination in step S13 is not power off, processing returns to step S1.

  On the other hand, if the result of determination in step S13 is power off, individual identification is performed (S15). In this step, the individual identification unit 22 identifies the person captured in the captured image recorded by the recording unit 14.

  When the individual identification is completed, event classification is performed (S17). In this step, the event classification unit 24 classifies into a group of images based on the shooting date / time information recorded together with the shot images. Details of the event classification will be described later with reference to FIG.

  When the event classification is completed, next, scene classification is performed (S19). The scene classification is to classify the scenes of the image group classified by the event classification. In the present embodiment, the scene classification is classified into four types: snap, travel, outdoor event, and indoor event. Details of the scene classification will be described later with reference to FIG.

  When individual identification in step S15, event classification in step S17, and scene classification in step S19 are performed, the execution result is written in the information file in each step. FIG. 3 shows examples of captured images and information files. In the example of FIG. 3, it is a group of a plurality of images classified by event classification, the images of six frames shown in (a) to (f), information corresponding to each image, and event information shown in (g) Consists of

  In the example shown in FIG. 3A, the faces of three persons consisting of persons A, B, and C are detected as information of the captured image (0001.JPG), and the scene is outdoors. In the example shown in FIG. 3B, the faces of two people consisting of persons B and D are detected as information of the captured image (0002.JPG), and the scene is outdoors. In the example shown in FIG. 3C, no face is detected as information of the captured image (0003.JPG), and the scene is outdoors.

  In the example shown in FIG. 3D, only the person A is detected as the information of the captured image (0004.JPG), and the scene of the person is outdoors. In the example shown in FIG. 3E, the face is not detected as the information of the captured image (0005.JPG), and the scene is outdoors. In the example shown in FIG. 3F, the faces of two people consisting of persons E and F are detected as the information of the captured image (0006.JPG), and the scene is outdoors.

  FIG. 3G shows information related to a group of a plurality of images consisting of FIGS. That is, these groups of multiple images have an event number of 0001, images are composed of 0001 to 0006, characters are A, B, C, D, E, and F, and the scene classification is It is an outdoor event. These captured images and tabulated information are determined and recorded in the recording unit 14 at the time of power-off as described above.

  If the result of determination in step S <b> 1 is playback mode, it is next determined whether or not playback is normal (S <b> 21). In the present embodiment, two types of normal reproduction and group reproduction are prepared for reproduction display of the captured image. The normal reproduction is a known reproduction display, and is a mode in which a photographed image recorded in the recording unit 14 is selected and displayed.

  If the result of determination in step S21 is normal playback, a known playback display is performed (S25). On the other hand, if the result of determination in step S21 is not normal playback, group playback is performed (S23). Details of this group reproduction will be described later with reference to FIG.

  Next, the event classification in step S17 (FIG. 2) will be described using the flowchart shown in FIG. Event classification is performed based on shooting date / time information recorded together with the shot image. For the same event, the event classification is performed on the assumption that the shooting date and time of the previous shooting and the previous shooting are close. In this embodiment, this interval is set to 24 hours, but of course, other than this, for example, 12 hours, 18 hours, etc. may be set as appropriate design values.

  When the event classification subroutine is entered, it is first determined whether or not the current shooting is within 24 hours from the previous shooting (S31). If it is within 24 hours as a result of this determination, it is treated as the previous event, and this captured image is added to the previous event (S33). If the result of determination in step S31 is not within 24 hours, it is treated as a new event, and this captured image is set as a new event (S35). When the process of step S33 or step S35 ends, the process returns to the original flow.

  Next, scene classification in step S19 (FIG. 2) will be described using the flowchart shown in FIG. The scene classification is performed based on shooting information recorded attached to the shot image. That is, the information is classified into four types, that is, snap, travel, outdoor event, and indoor event, using the ratio of the face detection image, the photographing date / time bias, and the information on the light source.

  When the scene classification subroutine is entered, it is first determined whether or not it is within 24 hours from the previous shooting (S41). Since the scene classification is performed based on the image group of the same event after the event classification is determined, the scene classification is not performed when 24 hours have not elapsed since the previous shooting.

  If the result of determination in step S41 is that it is within 24 hours from the previous shooting, the event is undecided (S59), and this scene undecided is recorded in the recording unit 14 (S61). That is, in the example shown in FIG. 3, undecided is recorded in the scenes in the information files in FIGS. 3A to 3F, and undecided is recorded for the events in the information file in FIG. When this recording ends, the original flow is restored.

  If the result of determination in step S41 is that it has exceeded 24 hours since the previous shooting, the event is confirmed. That is, in the example shown in FIG. 3, the captured images in FIGS. 3A to 3F are determined to belong to the same event, and the event number in the information file in FIG. 3G is recorded.

  Subsequently, it is determined whether the ratio of the face detection image is larger or smaller than a predetermined value in the group of images determined in step S43 (S45). If the result of this determination is that there are few face detections, it is assumed that there are many snap shots, and snaps are recorded in the information file (S53), and the flow returns to the original flow.

  If the result of determination in step S45 is that the ratio of face detection images is high, then it is also determined whether the shooting date and time are biased (S47). As a result of the determination, if the photographing date and time bias is dispersed, the trip is recorded in the information file (S55), and the process returns to the original flow.

If the result of determination in step S47 is that the deviation in shooting date and time is concentrated, it is next determined whether the light source is artificial light or sunlight (S49). This determination of the light source is performed based on information on white balance recorded with the photographed image, information on use of the auxiliary light source, and the like.

  If the result of determination in step S49 is that the light source is sunlight, an outdoor event is recorded in the information file (S51), and if the light source is artificial light, an indoor event is recorded in the information file ( S57). When these records are completed, the original flow is restored.

  FIG. 6 shows an example of scene classification. Event A is a drinking party, but it is determined as an indoor event because the ratio of face detection is large, the shooting dates and times are concentrated, and shooting is performed with artificial light. Event B is an example of a trip for 3 days and 2 nights, but it is determined to be a trip because the ratio of face detection is large and the uneven shooting date and time is dispersed.

  In addition, event C is an athletic meet case, but the ratio of face detection is large, the bias of shooting date and time is concentrated, and it is determined as an outdoor event because it is shot with sunlight. Event D is a wedding case, but it is determined as an indoor event because the ratio of face detection is large, the shooting dates and times are concentrated, and shooting is performed with artificial light. Event E is determined to be a snap because the ratio of face detection is small.

  Next, the group reproduction in step S23 (FIG. 2) will be described using the flowchart shown in FIG. Group playback is a playback mode in which everyone can be viewed with a minimum number of images at a specified event. When this group reproduction is selected, a group reproduction display 100 is displayed on the display unit 34 as shown in FIG.

  When the group reproduction subroutine is entered, event names are first displayed in a list (S71). As a list display of event names, for example, an event display 101 is displayed as shown in FIG. 10, and event displays 102 and 103 are sequentially displayed by operating a cursor operation member as described later. Here, in each event display 101-103, an event number, a photographing date, and a scene are displayed, respectively.

  Next, it is determined whether or not the user has selected an event that he / she wants to browse from a list of event names (S73). If no event is selected as a result of this determination, it is determined whether a cursor operation has been performed or a cancel operation has been performed (S75). As a result of the determination, if a cursor operation has been performed, the event display 101 to 103 is switched according to the cursor operation, that is, the cursor operation (S77), and the process returns to step S73. On the other hand, if the result of determination in step S75 is that a cancel operation has been made, processing returns to the original flow.

  If the result of determination in step S73 is that an event has been selected, extraction processing is performed (S79). This extraction process is a step of extracting one or a plurality of images from which everyone can be extracted with the minimum number of images. Details of this extraction processing subroutine will be described later with reference to FIG.

  When the extraction process is completed, the browsing candidate images extracted in the extraction process are displayed (S81). That is, as shown in FIG. 10, the browsing candidate images 111 to 114 are sequentially displayed every time the operation member is operated. Subsequently, it is determined whether any of the browsing candidate images has been selected or canceled (S83). If the result of this determination is cancellation, the process returns to step S71 and the event names are displayed in a list.

  If the result of determination in step S83 is that an image has been selected from among browse candidate images, this image is selected (S85), and this image is output from the output unit 15 to a printer or the like (S87). When the image output ends, the process returns to step S71, and when the cancel operation is performed in step S75, the process returns to the original flow.

  Next, the extraction processing subroutine of step 79 will be described with reference to the flowchart shown in FIG. In this flow, first, an image including the most people is extracted (S91). That is, an image including the largest number of persons is extracted from a plurality of images belonging to the currently selected event based on the number of face detection detections recorded in the information file.

  Subsequently, it is determined whether all the members have been extracted from the image extracted in step S91 (S93). As a result of the determination, if all the members have been extracted, the process returns to the original flow, and the image extracted in step S81 is displayed.

  If all the members have not been extracted as a result of the determination in step S93, an image including the largest number of persons excluding the persons included in the extracted image is extracted from images other than the extracted image (S95). When this extraction is completed, the process returns to step S93, and steps S93 and S95 are repeatedly executed until all the members are extracted. If all the members are extracted as a result of the determination in step S93, the process returns to the original flow.

  Next, a more specific example of the extraction process shown in FIG. 8 will be described using the flowchart shown in FIG. When this extraction process subroutine is entered, first, 1 is set as the number n of images to be extracted from a plurality of images belonging to the same event (S101). Subsequently, in the set number n of extracted images, an OR set of characters, that is, a union of characters is obtained (S103).

  Subsequently, it is determined whether or not the result of the union of the characters covers all members (S105). Individual identification is recorded in the information file (see FIG. 3G), and in this step, it is determined whether all identified individuals have been extracted.

  If all the members are not covered as a result of the determination in step S105, it is determined whether or not all combinations that can be taken with the number n of extracted images have been determined (S111). As a result of the determination, if the determination is not performed for all combinations, the combinations are changed (S113), the process returns to step S103, and the union of the characters is obtained again.

  If it is determined in step S111 that all combinations have been determined, 1 is added to the number n of extracted images (S113), the process returns to step S103, and the union of the characters is obtained again.

  If all the members can be covered as a result of the determination in step S105 while repeating the above-described series of determinations, an image that can cover all the members is selected (S107). When the selection is completed, the process returns to the original flow, and browsing candidate images are displayed in step S81. In this subroutine, since the search is started from 1 as the number of extracted images n, it is possible to select an image that can cover everyone with the minimum number of extracted images.

  As described above, in the present embodiment, a face image of a person is detected from a plurality of images by the face detection unit 16, and an individual is specified based on the face image by the individual specification unit 22. Then, when searching for an image from a plurality of images, the image search unit 25 includes all the persons included in the plurality of images and minimizes the number of images based on the individual identification result by the individual identification unit. An image is selected from a plurality of images.

  For this reason, in this embodiment, it is possible to efficiently search for an image including many persons from the image group. Further, when the image is reproduced and displayed, it is possible to prevent only the same person from appearing, and the reproduction display is not monotonous. In addition, since the faces of all the subjects can be confirmed with a small number of images, it is possible to efficiently search for the subjects. For example, when you want to quickly find a picture of Mr. A, you can quickly list everyone's faces, find Mr. A's face from the images, move the cursor on the screen, or touch Mr. A's face on the touch panel. If instructed, the next display image can be applied as another image in which Mr. A appears. As described above, the camera according to the present embodiment can also be effective for a specific image search.

  In the present embodiment, a plurality of images are classified, and the minimum number of images that cover all characters in the same classification are obtained. For this reason, the time required for image search can be shortened, and when reproduced and displayed based on the result of this image search, since they are of the same classification, the viewer can enjoy easily.

  In the present embodiment, face detection is performed at the time of shooting standby, but this detection may be performed at a different timing, such as at power-off, without being performed at the time of shooting standby.

  Further, in the present embodiment, an image captured by the camera is subjected to group reproduction processing within the camera and is reproduced and displayed on the camera display device or an external display device. You may make it memorize | store in a personal computer, perform group reproduction | regeneration based on this memorize | stored image, and may be made to perform a reproduction | regeneration display.

  Furthermore, the event classification and the scene classification in this embodiment are merely examples, and it is only necessary to classify a plurality of images into an image group as appropriate. Of course, other classifications may be adopted.

  Furthermore, in the present embodiment, the digital camera is used as the photographing device. However, the camera may be a digital single-lens reflex camera or a compact digital camera, and may be a mobile phone or a personal digital assistant (PDA). Of course, a camera built in Personal Digital Assist) may be used.

  The present invention is not limited to the above-described embodiments as they are, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, you may delete some components of all the components shown by embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

It is a block diagram which shows the structure of the camera concerning one Embodiment of this invention. It is a flowchart which shows operation | movement of the camera concerning one Embodiment of this invention. In one Embodiment of this invention, it is a figure which shows the structure of a picked-up image and an information file, (a)-(f) shows the structure of each picked-up image and an information file, (g) is the information file which shows event content FIG. It is a flowchart which shows the operation | movement of the event classification | category in the camera concerning one Embodiment of this invention. It is a flowchart which shows the operation | movement of the scene classification | category in the camera concerning one Embodiment of this invention. In one Embodiment of this invention, it is a figure which shows the example of a scene classification | category. It is a flowchart which shows the operation | movement of group reproduction | regeneration in the camera concerning one Embodiment of this invention. It is a flowchart which shows the operation | movement of the extraction process in the camera concerning one Embodiment of this invention. It is a flowchart which shows the more concrete operation | movement of the extraction process in the camera concerning one Embodiment of this invention. FIG. 10 is a diagram showing a display during group reproduction in an embodiment of the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Data bus, 11 ... Imaging part, 12 ... Optical system control part, 13 ... Clock part, 14 ... Recording part, 15 ... Output part, 16 ... Face detection , 20 ... control unit, 21 ... image processing unit, 22 ... personal identification unit, 23 ... scene classification unit, 24 ... event classification unit, 25 ... image search unit, 31 ... Operation determination unit, 32 ... Operation unit, 33 ... Display control unit, 34 ... Display unit, 100 ... Group playback display, 101 ... Event display, 102 ... Event display 103 ... Event display 111 ... Image 112 ... Image 113 ... Image 114 ... Image

Claims (9)

A person identifying unit for detecting a person's face image included in each of the plurality of images and identifying an individual from the plurality of images;
An image search unit for searching for an image from the plurality of images so as to include all persons included in the plurality of images and to minimize the number of images based on a result of personal identification by the individual identification unit; ,
An image search apparatus comprising:   2. The image search apparatus according to claim 1, further comprising an image classification unit that classifies images according to image types, wherein the plurality of images belong to one of image groups classified by the image classification unit. The image search device described in 1. Detecting a face image of a person included in each of the plurality of images from among a plurality of images and identifying an individual;
Searching for an image from the plurality of images so as to include all persons included in the plurality of images based on the individual identification result and to minimize the number of images;
An image search method characterized by comprising:   Responding to a display unit for displaying information representing the type of image, a selection unit for selecting desired information from the information representing the type of image displayed on the display unit, and an operation for selecting the selection unit. An image display device for playing back an image searched for by the image search device according to claim 2, further comprising: an image display unit that displays an image searched by the image search unit.   A step of displaying information representing the type of image, a step of selecting desired information from the information representing the type of image, and displaying the searched image in response to an operation of selecting the desired information. An image reproducing method for reproducing an image retrieved by the image retrieval apparatus according to claim 2, further comprising: Detecting a face image of a person included in each of the plurality of images and identifying an individual from the plurality of images;
Searching for an image from the plurality of images so as to include all persons included in the plurality of images based on the individual identification result and to minimize the number of images;
Image search program for causing a computer to execute the program. Detecting a face image of a person included in each of the plurality of images and identifying an individual from the plurality of images;
Searching for an image from the plurality of images so as to include all persons included in the plurality of images based on the individual identification result and to minimize the number of images;
Displaying the searched image, and
Replay program for causing a computer to execute   8. The image reproduction program according to claim 7, wherein when searching for the image, the plurality of images are classified, and only the images belonging to the classification are searched. A person identifying step of detecting a person's face image included in each of the plurality of images and identifying an individual from the plurality of images;
Extracting a first image including the largest number of persons from among the images included in the plurality of images based on the individual identification result in the individual identification step;
Extracting a second image containing the largest number of persons not included in the first image;
Image search program for causing a computer to execute the program.