JP6941950B2 - Image providing system, image providing method, and image providing program - Google Patents

Description

The present invention relates to an image providing system or the like that provides an image to a person concerned with a resident of a facility.

A technique of extracting a person's face from images such as still images and moving images taken, identifying who the person is, and using the information in various ways is widespread.

In the monitoring device of Patent Document 1, when a player is photographed at a game store and the photographed face image matches the face image of a person registered as a suspicious person, a suspicious person is detected. It is sent to the mobile terminal. In the information distribution system of Patent Document 2, the visitors and exiters of the apartment house are photographed by a surveillance camera, and when the photographed image matches the registered resident's face image, the image is notified and notified. Notify first.

Japanese Patent No. 5617627 Japanese Unexamined Patent Publication No. 2006-107187 Japanese Unexamined Patent Publication No. 2016-139865

By the way, in facilities such as nursery schools and nursing homes, residents (children and the elderly) live inside the facilities, away from their parents, children, relatives and other related parties. Since those concerned are concerned about the state of the residents spending time at the facility, it is preferable that the facility provides photos and videos of the residents so that they can grasp the living conditions of the residents. Although the above-mentioned image utilization technology may be applicable to such applications, the conventional technology has the following problems.

In the monitoring device of Patent Document 1, for the purpose of detecting a suspicious person, all the images determined to show the suspicious person are transmitted, so that even the images that should not be transmitted are transmitted. In addition, when a suspicious person is detected, the image is immediately transmitted to the staff member, so that the image is endlessly transmitted to the persons concerned.

The information distribution system of Patent Document 2 also has a similar problem. That is, since all the images determined to show the resident are transmitted to the notification contact, even the images that should not be transmitted are transmitted. In addition, when a person entering or leaving the building is detected, the image is immediately sent to the notification contact, so that the image is sent endlessly to the persons concerned.

The vehicle peripheral image transmitting device of Patent Document 3 continuously transmits the captured peripheral image to the monitoring device when the power is supplied from the power supply device, and the obstacle sensor causes an abnormality when the power is not supplied. When it is detected, the peripheral image is transmitted to the monitoring device. This peripheral image transmission device also has the same problems as the above-mentioned monitoring device and information distribution system, and there is a possibility that images that should not be transmitted may be transmitted, and the images are endlessly transmitted to the persons concerned. become. In addition, this peripheral image transmission device does not extract a person's face from an image, and even when setting conditions for image transmission, it is not suitable for transmitting the living conditions of residents.

The present invention is intended to solve the above-mentioned problems of the prior art, and an object of the present invention is to realize a system or the like capable of providing an appropriate image to a person concerned with a resident of a facility.

<Structure 1>
The feature configuration of the image providing system to achieve the above objectives is
A resident of the facility is identified from the captured image taken at the facility, and a candidate image that is a candidate for an image to be provided to the persons concerned of the resident is extracted based on the extraction conditions stored in the condition storage unit. Extractor and
Determining the similarity of the plurality of candidate images extracted by the extracting unit, a plurality of the candidate image having low similarity to each other, possess a providing unit for providing to the party,
The extraction unit does not extract the captured image that matches the exclusion condition stored in the condition storage unit as the candidate image .

<Structure 16 >
The feature configuration of the image providing method for achieving the above purpose is
An extraction step of identifying a resident of the facility from the captured image taken at the facility and extracting a candidate image which is a candidate of an image to be provided to the persons concerned of the resident based on the stored extraction conditions.
Determining the similarity of the plurality of the candidate images extracted by the extraction step, a plurality of the candidate image having low similarity to each other, possess a providing step of providing to the party,
The extraction step is that the captured image that matches the stored exclusion condition is not extracted as the candidate image .

<Structure 17 >
The feature structure of the image providing program to achieve the above purpose is
A resident of the facility is identified from the captured image taken at the facility, and a candidate image that is a candidate for an image to be provided to the persons concerned of the resident is extracted based on the extraction conditions stored in the condition storage unit. Extraction process procedure and
A provision processing procedure for determining the similarity of the plurality of candidate images extracted by the extraction processing procedure and providing a plurality of the candidate images having low similarity to each other to the related parties.
Let the computer run
The extraction processing procedure is that the captured image that matches the exclusion condition stored in the condition storage unit is not extracted as the candidate image .

According to the above feature configuration, since the candidate image is extracted from the captured image and provided to the related parties based on the extraction conditions, it is possible to provide an appropriate image to the related parties of the resident. Become. In the present invention, the person concerned is a person who is set in the image providing system to send the image of the resident, and is not limited to the relatives of the resident, but is limited to parents, guardians, children, and relatives. No.
Further, according to the above-mentioned feature configuration, since the captured image that matches the exclusion condition is not extracted as a candidate image, it is preferable to suppress the provision of an inappropriate image to the persons concerned.
In addition, when a plurality of images are selected from the candidate images and the images are provided, if the images having a lot in common in the shooting time condition, the shooting location condition, the facial expression condition, the number of people condition, and the volume condition are provided, the related parties will be asked. It is likely that only similar images will be provided. Therefore, when selecting an image to be provided, it is preferable to determine the similarity of a plurality of candidate images and provide as different images as possible. According to the above feature configuration, the providing unit determines the similarity of the plurality of candidate images and provides the plurality of candidate images having low similarity to each other to the related parties, so that various images having low similarity are provided. It is preferable that the related parties can grasp the living conditions of the residents from various images.

<Structure 2>
Another feature configuration of the image providing system according to the present invention is that the providing unit classifies the candidate images into a plurality of groups based on the information used for determining the extraction conditions, and is based on whether or not they belong to the same group. The point is to determine the similarity of the plurality of candidate images.
<Structure 3 >
Another characteristic configuration of the image providing system according to the present invention is that the extraction condition includes the following collation rate condition.
Collation rate condition: The collation rate between the face of the resident in the captured image and the registered face image of the resident is equal to or higher than a predetermined collation threshold value.

According to the above-mentioned feature configuration, an image having a collation rate equal to or higher than a predetermined collation threshold value can be provided to the related parties. Therefore, for example, it is possible to provide the related parties with an image that is reliably collated by increasing the collation threshold value. It is also possible to provide a large number of images to interested parties by lowering the collation threshold (although the collation may be uncertain). That is, according to the above-mentioned feature configuration, the degree of collation of the candidate images can be arbitrarily controlled, which is preferable.

<Structure 4 >
Another characteristic configuration of the image providing system according to the present invention is that the extraction conditions include the following shooting time conditions.
Shooting time condition: The shooting time of the shot image is included in a predetermined extraction target time zone.

According to the above feature configuration, it is possible to provide an image in which the shooting time is included in a predetermined extraction target time zone to the persons concerned, so that it is possible to provide an appropriate image corresponding to an event (meal, game, walk, etc.) at the facility. It is suitable. For example, it is possible to provide a person concerned with an image of a resident during a meal by setting the extraction target time zone as a meal time zone at a facility.

<Structure 5 >
Another characteristic configuration of the image providing system according to the present invention is that the extraction conditions include the following shooting location conditions.
Shooting location condition: The shooting location of the captured image is included in the predetermined extraction target location.

According to the above feature configuration, it is possible to provide the concerned parties with an image in which the shooting location is included in the predetermined extraction target location, so that it is possible to provide an appropriate image corresponding to the location in the facility (nursery room, garden, terrace, etc.). It is possible and suitable. For example, it is possible to provide an image of a resident playing in the garden with the extraction target place as the garden.

<Structure 6 >
Another characteristic configuration of the image providing system according to the present invention is that the extraction conditions include the following facial expression conditions.
Facial expression condition: The facial expression of the resident's face in the photographed image is a predetermined facial expression.

According to the above-mentioned feature configuration, it is possible to provide an image in which the facial expression of the resident is a predetermined facial expression to the persons concerned, so that it is possible to provide an appropriate image according to the facial expression of the resident, which is preferable. For example, an image of a resident's smile can be provided to a person concerned with a specific facial expression as a “smile”.

<Structure 7 >
Another characteristic configuration of the image providing system according to the present invention is that the extraction condition includes the following number of people condition.
Number of people condition: The number of people in the photographed image is within the predetermined number of people range.

According to the above feature configuration, it is possible to provide an image in which the number of people in the photographed image is within a predetermined number range to the persons concerned, so that an appropriate image is provided according to the number of persons in the photographed image. Is possible and suitable. For example, if the number of people range is "1", it is possible to provide an image showing only the residents to the people concerned, and if the number of people range is "2 or more", an image showing multiple people is possible. Can be provided to interested parties.

<Structure 8 >
Another characteristic configuration of the image providing system according to the present invention is that the extraction conditions include the following personal identification conditions.
Personal identification condition: The person specified from the identifier shown in the photographed image matches the resident.

Here, the identifier refers to a resident's name tag, a mark attached to the resident's clothes, or the like, which can identify the resident by image recognition. According to the above feature configuration, the person identified from the identifier in the photographed image can provide the related person with an image that matches the resident. It can be provided and is suitable.

<Structure 9 >
Another characteristic configuration of the image providing system according to the present invention is that the extraction condition includes the following volume condition.
Volume condition: The volume when the captured image is captured is within a predetermined volume range.

The volume of the surroundings when the captured image is captured changes according to the situation at that time. For example, in a nursery school, the volume is low during the nap time and the volume is high during the play time. In particular, the volume can be even louder when something fun happens, and the images taken in such cases can be favorable to those involved. According to the above-mentioned feature configuration, it is possible to provide an image in which the volume of the captured image is within a predetermined volume range to the persons concerned, so that it is possible to provide an appropriate image according to the surrounding conditions at the time of shooting. Suitable.

<Structure 10>
Another characteristic configuration of the image providing system according to the present invention is that the exclusion condition includes the following reflection exclusion condition.
Reflection exclusion condition: The face of a predetermined person is shown in the captured image.

According to the above-mentioned feature configuration, it is possible to exclude an image showing the face of a predetermined person and provide the image to the persons concerned, so that it is preferable to suppress the provision of an inappropriate image to the persons concerned. For example, if a specific person is a "facility staff", an image showing the staff's face is not extracted as a candidate image, and it is preferable because the provision to related parties can be suppressed.

<Structure 11>
Another characteristic configuration of the image providing system according to the present invention is that the exclusion condition includes the following age exclusion condition.
Age exclusion condition: The photographed image shows the face of a person within a predetermined age range.

According to the above-mentioned feature configuration, it is possible to exclude an image showing the face of a person within a predetermined age range and provide the image to the persons concerned, so that it is preferable to suppress the provision of an inappropriate image to the persons concerned. .. For example, if the facility is a nursery school and the predetermined age range is "20 years old or older", images showing the faces of adults who are not children, such as staff and related persons, are not extracted as candidate images and are related. It is suitable because it can suppress the provision to the person.

<Structure 12>
Another characteristic configuration of the image providing system according to the present invention is that the exclusion condition includes the following absence exclusion condition.
Absence exclusion condition: At the time of shooting the captured image, the resident shown in the captured image is absent from the facility.

For example, in the case of a nursery school, there are days and hours when the child does not go to the kindergarten due to illness, etc., and in the case of a nursing home, there is a period of absence due to temporary return home. Residents who are absent from the facility at the time of shooting the captured image should not be shown in the captured image, so the system's judgment that the person is "shown in the captured image" is a false recognition of the face. Etc., and such captured images should not be provided to interested parties. According to the above feature configuration, when the resident is absent from the facility at the time of shooting the captured image, the image can be excluded and provided to the related parties, so that an inappropriate image can be provided to the related parties. It is suppressed and suitable.

<Structure 13>
Another feature configuration of the image providing system according to the present invention is that it has an input unit that accepts an input regarding whether or not to provide the candidate image extracted by the extraction unit to the related parties.

According to the above feature configuration, the providing unit provides the image to the related party after receiving the input regarding whether or not the input unit provides it to the related party, so that the image provided to the related party is more appropriate. It becomes a thing and is preferable.

<Structure 14>
Another characteristic configuration of the image providing system according to the present invention is that the condition storage unit stores the extraction conditions in correspondence with the resident, and the extraction unit corresponds to the resident. The point is to extract the candidate image related to the resident based on the extraction conditions.

According to the above-mentioned feature configuration, since the extraction conditions for extracting the candidate images can be set for each resident, it is possible to provide an image according to the request of the persons concerned, which is preferable.

<Structure 15>
Another characteristic configuration of the image providing system according to the present invention is that when the number of candidate images extracted by the extraction unit for a specific resident is less than a predetermined number threshold value, the resident is photographed. The point is that it has a shooting instruction unit that issues the instruction of.

According to the above feature configuration, when the number of candidate images is small, an instruction to shoot a resident is issued, so that the number of shot images showing the resident can be increased and more images can be provided to the persons concerned. preferable.

According to the present invention, it is possible to realize a system or the like capable of providing an appropriate image to a person concerned with a resident of a facility.

Explanatory drawing showing the outline of the image providing system Block diagram showing the configuration of the image providing system Flowchart showing the operation of the image providing system Diagram showing an example of the operation screen of the image providing system Diagram showing an example of the operation screen of the image providing system Diagram showing an example of the operation screen of the image providing system

Hereinafter, the image providing system, the image providing method, and the image providing program according to the present embodiment will be described with reference to the drawings. These systems and the like are applied to facilities where residents live (stay), such as nursery schools and nursing homes. Hereinafter, the case where the image providing system or the like is applied to the nursery school will be described as an example.

FIG. 1 shows an outline of the image providing system 1. Cameras (hall camera 11, garden camera 14, etc.) are placed inside the nursery school 2 or in the garden, and the residents 3 (children) of the facility are photographed. The captured image 7 (still image / moving image) captured by the camera is stored in the image storage unit 21b of the control device 20. Then, the candidate image 8 is extracted based on various extraction conditions and provided to the guardian 4 (related person). Various methods are possible to provide the image to the guardian 4. For example, the image can be attached to an e-mail and sent, or the image is saved in a server or the like on the Internet and viewed by the guardian 4. It is also possible.

As shown in FIG. 2, the image providing system 1 according to the present embodiment includes a photographing device 10 and a control device 20.

The photographing device 10 is a device that photographs the inside of the facility and generates a moving image or a still image. The generated moving image / still image is transmitted to the control device 20. Specifically, the photographing device 10 is a video camera or a digital camera. The photographing device 10 can be fixed to a wall surface or ceiling (fixed type), or can be carried by a person by hand (mobile type). It is also possible to apply a camera of a mobile communication terminal such as a smartphone or a mobile phone as the photographing device 10 of the image providing system 1. In the present specification, the term "image" includes both moving images and still images. The audio during shooting may be recorded together with the moving image. The still image may be recorded with the sound before and after the shooting.

In the present embodiment, the fixed imaging device 10 continuously captures moving images like a so-called surveillance camera. In the case of the fixed type, the photographing direction and the angle of view of the photographing device 10 may be fixed values or may be changed by the control of the control device 20. The mobile photographing device 10 shall capture a still image or a moving image according to the operation of the photographer.

In the present embodiment, the hall camera 11, the infant nursery room camera 12, the pediatric nursery room camera 13, the garden camera 14, and the play equipment camera 15 are used as the fixed imaging device 10, and the mobile camera 16 is used as the mobile imaging device 10. Is used.

The hall camera 11 is a photographing device 10 installed in the hall inside the building of the nursery school 2. The infant nursery room camera 12 is a photographing device 10 installed in the infant nursery room inside the building of the nursery school 2. The pediatric nursery room camera 13 is a photographing device 10 installed in the pediatric nursery room inside the building of the nursery school 2. Since these photographing devices 10 continuously photograph the room and generate an image, the resident 3 in each room is reflected in the generated image.

The garden camera 14 is a photographing device 10 installed in the garden of the nursery school 2. The playset camera 15 is a photographing device 10 installed in the garden of the nursery school 2 and aimed at the playset (swing, slide, etc.). Since these photographing devices 10 continuously photograph the garden and generate an image, the resident 3 in the garden is reflected in the generated image. In particular, in the case of the playset camera 15, the resident 3 who plays with the playset is reflected in the generated image.

Each of these fixed imaging devices 10 photographs a specific place in the nursery school 2 to generate an image. Therefore, the shooting location of the shot image can be determined based on which shooting device 10 was shot. When the photographing device 10 generates an image or transmits the generated image to the control device 20, information regarding the photographing location may be attached to the image. In that case, the information regarding the shooting location includes information regarding which shooting device 10 was used for shooting, information regarding the installation location of the shooting device 10 (for example, the name of the room), and the location where the shooting device is shooting (for example). Information related to the name of the room and the name of the play equipment) is possible.

The mobile camera 16 is a portable photographing device 10. The mobile camera 16 is carried by the staff 5 of the nursery school 2 and photographs the resident 3. Therefore, the shooting location of the image taken by the moving camera 16 changes each time. The image may be transmitted from the mobile camera 16 to the control device 20 by communication such as a wireless LAN, or by copying the image data via a storage medium such as a memory card. When the moving camera 16 generates an image, or when the generated image is transmitted to the control device 20, information about the photographing location may be attached to the image. In that case, the information regarding the shooting location may be position information acquired by GPS or the like, or information indicating that the image was taken by the moving camera 16. Information about the shooting location (for example, the name of the room, the name of the playset, etc.) may be added to the image by the staff of the nursery school 2 using a personal computer or the like.

The control device 20 controls the overall operation of the image providing system 1 and performs various processes. The control device 20 includes a storage unit 21, a control unit 22, a display input unit 23, and a communication unit 24. The control device 20 is, for example, a personal computer and is arranged in the office of the nursery school 2.

The storage unit 21 is composed of a storage device such as an HDD (Hard Disk Drive) or a non-volatile RAM (Random Access Memory). The storage unit 21 may be a storage device provided in the control device 20, or may be a storage area on a so-called cloud. In the present embodiment, the storage unit 21 includes a condition storage unit 21a and an image storage unit 21b.

Extraction conditions and exclusion conditions are stored in the condition storage unit 21a. The extraction condition is a condition for extracting the candidate image 8 from the captured image 7. The extraction conditions are, for example, a collation rate condition, a shooting time condition, a shooting location condition, a facial expression condition, a number of people condition, and the like. The exclusion condition is a condition in which the candidate image 8 is not extracted as a candidate image when the candidate image 8 is extracted from the captured image 7. The exclusion conditions are, for example, a reflection exclusion condition, an age exclusion condition, and the like. Details of these conditions will be described later.

The above-mentioned extraction conditions are stored in the condition storage unit 21a in a form corresponding to the resident. Specifically, the extraction conditions applied to each resident are stored. For example, it is stored in the condition storage unit 21a that the collation rate condition and the shooting time condition are applied to A-chan, a resident. For example, it is stored in the condition storage unit 21a that the collation rate condition, the shooting time condition, and the shooting location condition are applied to Mr. B, who is a resident. For example, for a certain resident, whether or not it is applied may be stored for each extraction condition.

The image storage unit 21b stores an image captured by the photographing device 10 and transmitted to the control device 20.

In addition to the above-mentioned data, the storage unit 21 contains a resident's face image, an employee's face image, an image providing method for each guardian (email address, etc.), resident's attendance data at a nursery school, and resident's identifier. (The name tag of the resident, the mark on the clothes of the resident, etc.) are memorized.

The control unit 22 includes an extraction unit 22a, an input unit 22b, a providing unit 22c, and a shooting instruction unit 22d. The control unit 22 is composed of, for example, a software program for realizing various processes, a CPU for executing the software program, various hardware controlled by the CPU, and the like. Software programs and data necessary for the operation of the control unit 22 are stored in the storage unit 21, but the storage destination is not particularly limited, and the software programs and data are stored in a storage device such as a disk or a flash memory provided separately. It may be present, or it may be stored in a communicablely connected server, storage device, or the like.

The extraction unit 22a identifies the residents of the nursery school from the images taken at the nursery school (facility), and provides the images to the guardians of the residents of the nursery school based on the extraction conditions stored in the condition storage unit 21a. Extract candidate images that are candidates. The extraction unit 22a identifies the resident by using appearance features such as a face and an identifier. The extraction unit 22a does not extract the captured image that matches the exclusion condition stored in the condition storage unit 21a as a candidate image. The extraction process of the candidate image by the extraction unit 22a will be described in detail later.

The input unit 22b accepts input of extraction conditions used when the extraction unit 22a extracts a candidate image. The input unit 22b stores the input extraction conditions in the storage unit 21.

Further, the input unit 22b accepts an input regarding whether or not to provide the candidate image extracted by the extraction unit 22a to the guardian. For example, the input unit 22b causes the display input unit 23, which will be described later, to display the operation screen of the image providing system 1, display a transmission button to the guardian for each candidate image on the operation screen, and operate the image providing system 1. Accepts input from people. When the operator clicks the send button, the candidate image is sent to the guardian.

The providing unit 22c provides the candidate image extracted by the extracting unit 22a to the guardian. The image is provided to the guardian, for example, by sending it as an attachment to an e-mail or by storing the provided image in a server or the like on the Internet. In the former case, the providing unit 22c functions as a mail transmission server, and in the latter case, the providing unit 22c transmits data to a server or the like that provides images.

When the number of candidate images extracted by the extraction unit 22a for a specific resident is less than a predetermined number threshold value, the shooting instruction unit 22d issues an instruction to photograph the resident. For example, when the number of candidate images extracted by the extraction unit 22a for the resident A is less than the number threshold value of 5, the shooting instruction unit 22d controls the display input unit 23 to perform A-chan three times or more. Display the instruction to shoot. For example, a staff member who sees this instruction shoots A-chan three times or more with the mobile camera 16.
Further, even if the number of candidate images extracted by the extraction unit 22a for a specific resident is larger than the predetermined number threshold value, the shooting instruction unit 22d may have many candidate images with high unclearness, as will be described later. When there are many candidate images with high similarity and many images with low collation rate, an instruction to photograph the resident may be issued.

The display input unit 23 is, for example, a liquid crystal monitor, a keyboard, and a mouse. The display input unit 23 is a device that displays an operation screen of the image providing system 1 and a shooting instruction by the shooting instruction unit 22d and receives an input from an operator of the image providing system 1.

The communication unit 24 is an interface unit for performing data communication via a wireless LAN, a wired LAN, or the like, and communicates with the photographing device 10 and the above-mentioned image providing server.

<Extraction conditions>
Hereinafter, an example of extraction conditions will be described with reference to FIGS. 5 and 6. 5 and 6 are examples of an input screen for the extraction conditions and the like of the image providing system 1 described above (hereinafter, may be referred to as a “condition input screen”). In this example, for each target person (resident) extracted by the extraction unit 22a, from the left of the table, the target person names such as residents and staff, attendance data, personal identification information, attributes, providers, extraction conditions, and Exclusion conditions are displayed, and you can add the target person and set each item. In this example, options are displayed for the extraction condition and the exclusion condition, "●" is displayed for the selected option, and "○" is displayed for the unselected option. The condition input screens shown in FIGS. 5 and 6 are continuous screens, which are switched and displayed by scrolling.

Personal identification information is an appearance feature that can be used for personal identification such as a face image or an identifier. It is possible to store a plurality of personal information for each target person. Specifically, a plurality of face images can be stored, and images of a plurality of identifiers such as name tags and marks can be stored. When an image of personal identification information is selected on the condition input screen, it is desirable that a plurality of stored images of faces and identifiers are displayed. The identifier may be a symbol or a number as well as an image. By providing the extraction unit 22a with a character recognition function, it is possible to collate character by character. The identifiers in FIGS. 5 and 6 are examples of storing images of name tags. Further, the feature extracted from the face image may be stored instead of the face image.

The collation rate condition is a condition that the collation rate between the face shown in the captured image and the face of the registered resident is equal to or higher than a predetermined collation threshold value. The collation rate is calculated by the extraction unit 22a by collating the face shown in the captured image with the face of the registered resident. For example, if the collation threshold is set to 80% and the collation rate between the face shown in the captured image and the face of the registered resident A-chan is 85%, the collation rate condition for A-chan is set. Match. The collation rate condition may be a condition that the collation rate between the face in the captured image and the face of the resident related to the guardian who transmits the captured image is equal to or higher than a predetermined collation threshold value. It is preferable to set the collation threshold value high because it is possible to reduce the possibility that an image of a resident who is not related to the guardian is provided due to misrecognition. If the collation threshold is set low, the possibility of the above-mentioned erroneous recognition / erroneous transmission increases, but there is an advantage that the number of extracted images increases. The collation rate condition (collation threshold) may be set according to the request (preference) of the guardian who is provided with the image.

The condition input screen of FIG. 5 shows an example in which options of "high", "slightly high", "normal", "slightly low", and "low" are displayed corresponding to the collation threshold value in order to facilitate the understanding of the operator. There is. For example, it is advisable to set the matching threshold value corresponding to the option so that the matching threshold value is set to 80% when "normal" is selected and the matching threshold value is set to 85% when "slightly high" is selected.

The shooting time condition is a condition that the shooting time of the shot image is included in the predetermined extraction target time zone. For example, for A-chan, when the extraction target time zone is set to "9:00 to 13:00", the image whose shooting time is 10:15 matches the shooting time condition. The image whose shooting time is 14:30 does not meet the shooting time condition. If the extraction target time zone is set long, the number of images to be extracted increases, and if the extraction target time zone is set short, the number of extracted images decreases. For example, if the extraction target time zone is set in the morning, images of residents from the time they enter the nursery school to the time of lunch are extracted. If the extraction target time zone is set to the evening, images of residents before leaving the park will be extracted. The shooting time condition (extraction target time zone) may be set according to the request (preference) of the guardian who is provided with the image. The condition input screen of FIG. 5 shows an example in which an arbitrary time is set and options of "all day", "am", and "afternoon" are displayed.

The shooting location condition is a condition that the shooting location of the captured image is included in the predetermined extraction target location. For example, for A-chan, when the extraction target location is set to "playset", the image whose shooting location is the playset (for example, the image taken by the playset camera 15) matches the shooting location condition. An image whose shooting location is a hall (for example, an image taken by the hall camera 11) does not meet the shooting location condition. If the number of extraction target locations is large (wide), the number of images to be extracted increases, and if the number of extraction target locations is small (narrow), the number of extracted images is small. The shooting location condition (extraction target location) may be set according to the request (preference) of the guardian who is provided with the image. The condition input screen of FIG. 5 shows an example in which a plurality of shooting locations can be selected.

The facial expression condition is a condition that the facial expression of the resident's face in the photographed image is a predetermined facial expression. The facial expression of the resident in the captured image is determined by the extraction unit 22a by analyzing the captured image. For example, when the predetermined facial expression of A-chan is set to "smile" and the face of A-chan in the captured image is a smile, the captured image matches the facial expression condition. For example, the extraction unit 22a may be configured to quantify the degree of smile, and a predetermined facial expression may be defined within the numerical range of the degree of smile. The facial expression condition (predetermined facial expression) may be set according to the request (preference) of the guardian who is provided with the image. The condition input screen of FIG. 5 shows an example in which a plurality of facial expressions can be selected and the degree of smile can be arbitrarily input when "smile" is selected.

The number of people condition is a condition that the number of people in the photographed image is within a predetermined number of people range. The number of people in the captured image is determined by the extraction unit 22a by analyzing the image. For example, for A-chan, when the number of people range is set to "1 person", the image in which A-chan is shown by one person is extracted, and the image in which two or more people are shown is not extracted. The number of people condition (number of people range) may be set according to the request (preference) of the guardian who is provided with the image. On the condition input screen of FIG. 5, the number of people range can be selected from "1 person", "multiple people", and "arbitrary", the number of people range can be input in "arbitrary", and only the upper limit value or the lower limit value can be input. be. Furthermore, "including a designated person" can be selected, and the image can be narrowed down to the image that is shown together with the person selected in the designated person column.

The personal identification condition selects the personal identification information used for personal identification, and in addition to the above-mentioned face matching condition, a condition that the person identified from the identifier shown in the photographed image matches the resident. Is. An identifier is a name tag of a resident, a mark attached to a resident's clothes, or the like, which can identify the resident by image recognition. For example, when the person specified from the identifier shown in the captured image is A-chan, who is a resident, the captured image is extracted as a candidate for an image to be provided to A-chan's guardian. The condition input screen of FIG. 6 shows an example in which "face only", "identifier only", "face or identifier", and "face and identifier" can be selected. For example, if you want to reduce the probability of false recognition, you can select "face and identifier" and narrow down by two conditions. If you want to recognize the face even if it does not appear, select "face or identifier" and use the two conditions together, and the subject should attach a name tag to the outside of the sleeve of the jacket or the back.

The volume condition is a condition that the volume when the captured image is captured is within a predetermined volume range. For example, for A-chan, when the volume range is set to "60db or more", the image whose volume is 70db when the captured image is captured matches the volume condition. An image having a volume of 40 db does not meet the volume condition. When the volume range is increased (volume is increased), the image when the resident is actively active tends to be extracted. Lowering the volume range (lowering the volume) tends to extract images when the resident is quietly spending time. The volume condition (volume range) may be set according to the request (preference) of the guardian who is provided with the image. The condition input screen of FIG. 6 shows an example in which a plurality of volumes can be selected from "loud", "normal", and "quiet". It is advisable to set an appropriate volume range for each option "loud", "normal", and "quiet".

<Exclusion conditions>
An example of the exclusion condition will be described below.

The reflection exclusion condition is a condition that the face of a predetermined person is shown in the photographed image. For example, when a predetermined person is set as "staff", if the photographed image shows the face of the employee, the photographed image is not extracted as a candidate image. This setting is suitable when the privacy of facility staff is important.

For example, when a predetermined person is set as <a certain resident>, if the photographed image shows the face of the resident, the photographed image is not extracted as a candidate image. This setting is suitable when parents place importance on the privacy of their children (residents). If a guardian has set a reflection exclusion condition for a resident (own child), the captured image in which the resident's face is reflected is extracted as a candidate image for other guardians. Not done. In this case, it is preferable to configure the extraction unit 22a so that the captured image is not extracted as a candidate image for other guardians, but is extracted as a candidate image for the guardian.

The reflection exclusion condition may be applied / not applied to each resident in the same manner as the extraction condition. For example, it may be set to apply the reflection exclusion condition to the candidate image for the guardian of A-chan. In this case, the captured image showing the face of a predetermined person is not extracted as a candidate image. The predetermined person may be a "staff", a "other kindergarten child", or a "person other than A-chan".

The condition input screen of FIG. 6 shows an example in which the reflection exclusion condition can be divided into the reflection exclusion condition for another person and the reflection exclusion condition for another person and set for each resident.
For example, in order to prevent the above-mentioned "staff" from being reflected, if the staff member E is memorized as the target person and "all" is selected as the condition for excluding the reflection on others, the staff member E will be included in the captured image. If so, the captured image is not extracted as a candidate image for any guardian. In addition, if "other than the designated person" is selected as the condition for excluding the reflection on another person and the designated person is selected, it is possible to exclude the image from being included in the candidate image other than the designated person designated by the target person. This is the same as allowing it to be included in the candidate image for the designated person.
On the other hand, if the condition for excluding the reflection of another person is selected, the reflection of another person can be excluded. For example, when "staff" or "another kindergarten child" is selected, a photographed image in which a subject whose attribute is an employee or another kindergarten is reflected is not extracted as a candidate image. By selecting "other than the designated person" as the condition for excluding the reflection of another person and selecting the designated person, it is possible to exclude the target person other than the designated person from being included in the candidate image. This is the same as allowing the target person who is the designated person to be included in the candidate image.

The age exclusion condition is a condition that the photographed image shows the face of a person within a predetermined age range. For example, when a predetermined age range is set to "20 years old or older" and a 23-year-old employee is shown in the captured image, the captured image is not extracted as a candidate image. This is a useful exclusion condition when the age group of residents and the age group of staff / guardians are separated. In addition to memorizing the age of the person in the captured image, the age can be estimated by analyzing the face, identifying the person from the face, and specifying the age of the person from the list, etc. Is possible.

The absence exclusion condition is a condition that the resident shown in the photographed image is absent from the facility at the time of photographing the photographed image. For example, if A-chan's face is recognized in a certain photographed image and A-chan is absent (absent) by referring to the nursery school attendance data, the photographed image is not extracted as a candidate image. Even if a resident is misrecognized in the captured image, the provision to the guardian can be suppressed, which is useful because the reliability of the image providing system 1 can be improved.

Hereinafter, the operation (image providing method) of the image providing system 1 will be described with reference to the flowchart of FIG. At the same time, the processing of the extraction unit 22a will be described.

In the photographing step (# 101), the image is photographed by the photographing device 10. The captured image is transmitted to the control device 20 and stored in the image storage unit 21b. When the image is a moving image, transmission to the control device and storage in the image storage unit 21b may be continuously performed in real time, or the moving image may be divided into predetermined time intervals.

In the extraction step (# 102), the resident is identified from the captured image stored in the image storage unit 21b and provided to the guardian of the resident of the facility based on the extraction condition and the exclusion condition stored in the condition storage unit. Candidate images that are candidates for the image to be used are extracted. Specifically, the extraction step is performed as follows.

The extraction unit 22a analyzes the captured image stored in the image storage unit 21b, refers to the information (shooting time, etc.) associated with the captured image, and acquires information for use in determining the extraction condition / exclusion condition. do. Then, it is determined whether or not the captured image meets the extraction condition / exclusion condition, and the candidate image is extracted.

In the following description, first, a case where the image is a still image will be described.

The extraction unit 22a analyzes the captured image and extracts the face of the person reflected in the image. Then, the extracted face is collated with the registered resident's face to identify the person related to the extracted face. When the person is a resident, it is determined whether or not the captured image matches the extraction conditions (and exclusion conditions) corresponding to the resident (hereinafter referred to as "specific resident"). If the photographed image does not show the face of a person, or if the photographed person is not a resident, the photographed image is not extracted as a candidate image.

The extraction of the candidate image by the extraction unit 22a is performed based on the extraction condition / exclusion condition related to the specific resident stored in the condition storage unit 21a. For example, when it is stored that the collation rate condition, the shooting time condition, and the shooting location condition are applied to the specific resident, the shot image is extracted as a candidate image when these three extraction conditions are satisfied. The remaining extraction conditions are not determined. Hereinafter, it will be described that all the extraction conditions / exclusion conditions are applied to the specific residents.

(Collation rate condition) The extraction unit 22a confirms the collation rate between the face of the specific resident in the captured image and the face of the registered specific resident. When the collation rate is equal to or higher than a predetermined collation threshold value, it is determined that the captured image matches the collation rate condition. When the collation rate is less than a predetermined collation threshold value, it is determined that the captured image does not meet the collation rate condition, and the captured image is not extracted as a candidate image.

(Shooting time condition) The extraction unit 22a confirms the shooting time of the shot image. When the shooting time is included in the predetermined extraction time zone, it is determined that the shot image matches the shooting time condition. If the shooting time is not included in the predetermined extraction time zone, it is determined that the shooting image does not match the shooting time condition, and the shooting image is not extracted as a candidate image.

(Facial expression condition) The extraction unit 22a confirms the facial expression of the specific resident in the photographed image. When the facial expression of the specific resident has a predetermined facial expression, it is determined that the captured image matches the facial expression condition. If the shooting time does not match the predetermined facial expression, it is determined that the shot image does not match the facial expression conditions, and the shot image is not extracted as a candidate image.

(Number of people condition) The extraction unit 22a confirms the number of people in the captured image. For example, the extraction unit 22a analyzes the captured image, extracts the faces of the person reflected in the image, and sets the number of the faces as the number of persons reflected in the captured image. When the number of people in the captured image is within the predetermined number range, it is determined that the captured image matches the number of people condition. If the number of people is not within the predetermined number range, it is determined that the captured image does not meet the number of people condition, and the captured image is not extracted as a candidate image.

(Personal identification condition) The extraction unit 22a analyzes the captured image and extracts the identifier (name tag of the resident, mark attached to the clothes of the resident, etc.) reflected in the image. Then, the extracted identifier is collated with the registered resident identifier to identify the person related to the extracted identifier. If the person matches the specific resident, it is determined that the person meets the personal identification conditions. If the person does not match the specific resident, it is determined that the person does not match the personal identification condition, and the captured image is not extracted as a candidate image. Alternatively, it is determined that the personal identification condition is satisfied based on the result of face matching and the result of identifier matching according to the setting contents.

(Volume condition) The extraction unit 22a confirms the volume when the captured image is captured. When the volume is within a predetermined volume range, it is determined that the volume condition is met. When the volume is not within the predetermined volume range, it is determined that the volume condition is not met, and the captured image is not extracted as a candidate image. The volume (volume when the captured image is captured) that serves as a reference for the judgment may be the volume at the shooting time of the captured image or the volume during a predetermined period including the shooting time (for example, 5 seconds before and after the shooting time). good.

(Conditions for Excluding Reflection) The extraction unit 22a confirms whether or not the set "predetermined person" is reflected in the captured image. Specifically, it is confirmed whether or not the person (described above) in the photographed image identified by analyzing the photographed image corresponds to a predetermined person. When a predetermined person appears in the captured image, it is determined that the captured image meets the reflection exclusion condition, and the captured image is not extracted as a candidate image. If a predetermined person is not shown in the captured image, it is determined that the reflection exclusion condition is not met. It is preferable to configure the extraction unit 22a so that the determination of the reflection exclusion condition is not performed for the specific resident shown in the captured image.

(Age exclusion condition) The extraction unit 22a confirms whether or not the face of a person within the set "predetermined age range" is reflected in the captured image. For example, the face shown in the captured image is analyzed to estimate the age, and it is confirmed whether or not the face is within a predetermined age range. When the photographed image shows the face of a person within a predetermined age range, it is determined that the photographed image matches the age exemption condition, and the photographed image is not extracted as a candidate image. If the face of a person within a predetermined age range is not shown, it is determined that the reflection exclusion condition is not met.

(Absence Exclusion Condition) The extraction unit 22a confirms whether or not the specific resident shown in the photographed image is absent from the facility at the time of photographing the photographed image. Specifically, the extraction unit 22a confirms whether or not the specific resident is attending the kindergarten by referring to the resident's attendance data to the nursery school stored in the storage unit 21. If the specific resident is absent from the facility at the time of shooting the captured image, it is determined that the absentee exclusion condition is met, and the captured image is not extracted as a candidate image. If the specific resident is not absent from the facility, it is judged that the reflection exclusion condition is not met.

Next, a case where the captured image is a moving image will be described. The extraction unit 22a sequentially analyzes the still images constituting the moving image. When the face of the person reflected in the image is extracted, the extracted face is compared with the face of the registered resident to identify the person related to the extracted face. If the person is a resident, the determination of whether or not the captured image matches the extraction conditions (and exclusion conditions) corresponding to the resident (hereinafter referred to as "specific resident") is as described above. Will be done.

The determination can take various aspects. For example, based on the still image in which a specific resident is detected, a predetermined period including the still image (for example, 5 seconds before and after the reference still image) is cut out from the captured image and used as the judgment target video, and the judgment target video is used. Extraction conditions / exclusion conditions may be determined for all the included still images, and if the determination is passed, the determined moving image may be used as a candidate image.

For example, the extraction condition / exclusion condition may be determined for the still image in which the specific resident is detected. When the still image passes the determination, a predetermined period including the still image (for example, 5 seconds before and after the reference still image) may be cut out from the captured image and used as a candidate image.

For example, the extraction condition may be determined for the still image in which the specific resident is detected. When the still image passes the determination of the extraction conditions, a predetermined period including the still image (for example, 5 seconds before and after the reference still image) is cut out from the captured image and used as the determination target moving image. Then, the exclusion condition may be determined for all the still images included in the determination target moving image, and if the determination is passed, the determination target moving image may be used as a candidate image.

The extraction of the candidate image (extraction step # 102) by the extraction unit 22a described above may be performed in real time each time the captured image (still image / moving image) is stored in the image storage unit 21b, or may be performed at a predetermined time interval (extraction step # 102). For example, it may be performed every hour) or at a predetermined time (for example, 5 pm).

In the input step (# 103), the input unit 22b receives an input regarding whether or not to provide the candidate image extracted by the extraction unit 22a to the guardian. For example, the input unit 22b causes the display input unit 23 to display the operation screen of the image providing system 1, displays a transmission button to the guardian for each candidate image on the operation screen, and causes the operator of the image providing system 1 to display the operation screen. Accepts input.

FIG. 4 is an example of the operation screen of the image providing system 1 described above. In this example, for each candidate image, the provision preparation time, the resident, the provision destination, the extraction condition, the attached file, the send button, the delete button, and the save button are displayed from the left of the table.

The “preparation time for provision” is the shooting time or the time at which the candidate image was extracted. "Resident" is the name of a specific resident related to the candidate image. The "providing destination" is information indicating the providing destination of the candidate image, and in this example, the e-mail address of the guardian of the specific resident is displayed.

In the "extraction condition" column, the result of the determination regarding the candidate image and whether it is applied / not applied to the specific resident is displayed for each extraction condition. Diagonal lines are displayed in the columns for extraction conditions that are not applicable to specific residents. “○” is displayed in the column for the extraction conditions for which the candidate images have passed the judgment. In this example, "verification rate", "shooting time", "shooting location", "facial expression" and "number of people" are displayed as extraction conditions, but all extraction conditions (and exclusion conditions) are displayed on the screen. Alternatively, some extraction conditions (and exclusion conditions) may be displayed on the screen as in this example.

In the "Attachment" column, the candidate images extracted in the extraction step are displayed.

The "send button" is an input button regarding whether or not to provide the candidate image to the guardian. When the operator of the image providing system 1 operates the display input unit 23 and presses this button, the input unit 22b accepts an input to provide the candidate image to the guardian. For example, the image providing system 1 may be configured to immediately provide the candidate image to the guardian (providing step # 104) when the input button is pressed. For example, if the send button for each candidate image is pressed and then the batch send button (not shown) is pressed, the candidate image for which the send button is pressed is provided to the guardian (providing step # 104). , The image providing system 1 may be configured.

The "delete button" is an input button for whether or not to delete the candidate image from the image storage unit 21b. When the operator of the image providing system 1 operates the display input unit 23 and presses this button, the candidate image is deleted from the image storage unit 21b.

The "save button" is an input button for suspending the above-mentioned transmission / deletion decision for the candidate image and keeping the candidate image saved in the image storage unit 21b. When the operator of the image providing system 1 operates the display input unit 23 and presses this button, the candidate image remains saved in the image storage unit 21b, and the next time this operation screen is displayed, this candidate is displayed. The image is displayed again.

The operation screen display / input reception (input step # 103) by the input unit 22b described above may be performed each time the candidate image is extracted by the extraction unit 22a (extraction step # 102), or may be performed at a predetermined time interval (input step # 102). For example, it may be performed every hour) or at a predetermined time (for example, 5 pm).

In the providing step (# 104), the candidate image extracted by the extraction unit 22a in the extraction step # 102 is provided to the guardian. In the present embodiment, the candidate image for which the input to be provided to the guardian is received in the input step # 103 is provided to the guardian. Provision to parents is performed by sending an e-mail to the e-mail address of the provider or uploading a candidate image to a server or the like on the Internet so that the parents can view it.

The provisioning unit 22c described above may provide the candidate image to the guardian (providing step # 104) each time an input indicating that the candidate image is to be transmitted is received in the input step # 103, or may be performed at a predetermined time interval (providing step # 104). For example, it may be performed every hour) or at a predetermined time (for example, 5 pm).

As described above, in the image providing system 1 according to the present embodiment, the image providing method is executed. The method is an extraction step of extracting candidate images that are candidates for images to be provided to the guardians of residents of the facility based on the stored extraction conditions from the captured images taken at the facility, and an extraction step. It has a providing process of providing the candidate image to the guardian. The above-mentioned extraction step # 102 corresponds to the extraction step, and the providing step # 104 corresponds to the providing step.

(Other embodiments)
<1> In the above-described embodiment, a fixed camera, a mobile camera, and a camera of a personal digital assistant are exemplified as the photographing device 10. The photographing device 10 is not limited to these, and may be, for example, a camera mounted on a robot. For example, if the robot communicates with the resident by voice or gesture, the resident can be photographed from the front when communicating with the robot, so that the photographed image appropriately conveys the living situation of the resident. preferable.

<2> The image providing system 1 may include an image processing unit that processes a captured image or a candidate image. For example, the image processing unit processes a portion in which another face appears (for example, trimming, blurring, mosaic processing, etc.) so that only the face of a specific resident appears in the captured image. For example, the image analysis unit processes (for example, trimming, blurring, mosaic processing, etc.) the portion in which the face is reflected so that the face of the person corresponding to the reflection exclusion condition or the age exclusion condition does not appear in the captured image.

<3> As an extraction condition, a condition (voice condition) that the voice when the photographed image is photographed is the voice of a person (specific resident) in the photographed image can also be applied. The image taken while the resident is speaking is likely to be the image when the resident is speaking or actively working, and it is preferable to provide it to the guardian. For the determination of the voice condition, for example, the voice when the captured image is taken is analyzed and collated with the memorized voice of the resident (voiceprint, etc.) to identify which resident's voice is. , You may do it by confirming whether it matches the specific resident.

<4> As the extraction exclusion condition, a condition that the blurryness of the captured image is larger than a predetermined blurring threshold value (blurring exclusion condition) can also be applied. The blurryness is calculated from the blurred image, blurring, darkness, etc. of the photographed image by analyzing the photographed image. Blurred, blurred or dark images are unlikely to be useful for understanding the living conditions of residents even if provided to parents, and in some cases may cause discomfort to parents who see the images. .. It is preferable not to provide such blurry images to parents.

<5> When the number of extracted candidate images exceeds the number threshold value, it is desirable to provide images with as different conditions as possible from the images provided to the guardian. When multiple images are selected from the candidate images and the images are provided, if the images that have much in common in the shooting time condition, the shooting location condition, the facial expression condition, the number of people condition, and the volume condition are provided, the parents will do the same. There is a high possibility that only various images will be provided. Therefore, when selecting the image to be provided, it is preferable to judge the similarity of the plurality of candidate images and provide as different images as possible. Specifically, the providing unit 22c classifies the candidate images into a plurality of groups based on the information acquired by the extraction unit 22a for use in determining each extraction condition regardless of whether or not the extraction condition is set, and each of them. Select the image to be provided from the group. For example, it is preferable to classify them into groups according to the shooting location, shooting time, and number of people. In this way, the similarity of the candidate images of the same group is high, and the similarity of the candidate images of different groups is low. Therefore, if the candidate images provided from different groups are selected, the similarity between the provided images is low. Become. Alternatively, the providing unit 22c may collate the candidate images with each other, classify them into groups of similar candidate images, and select and provide the images to be provided from each group. The method by which the providing unit 22c determines the similarity of the candidate images is not limited to these above-mentioned methods as long as the similarity between the selected candidate images is low.

Regarding the provision of images having low similarity as described above, it is also possible to configure the provision unit as follows. When there are a plurality of candidate images similar to each other, the providing unit 22c performs a provision discontinuation process for discontinuing the provision of at least one candidate image among the candidate images. By discontinuing the provision of similar candidate images, the images provided to the guardian are suitable because they have low similarity and are different from each other. Judgment as to whether or not they are similar to each other can be performed by various methods. For example, the candidate images may be classified into a plurality of groups based on the information acquired by the extraction unit 22a for determination of each extraction condition, and similar determination may be made based on whether or not they belong to the same group. .. For example, the candidate image may be subjected to image analysis to acquire information such as parameters such as color tone and lightness and darkness of the image and the shape of a moving object, and a similar determination may be made based on these.

<6> The image providing system described in the above-described embodiment can also be configured as follows. The image providing system includes an extraction unit and a providing unit. The extraction unit identifies the residents of the facility from the captured images taken at the facility. The identified resident is called the provider resident. The extraction unit extracts candidate images that are candidates for images to be provided to the guardian of a specific resident (hereinafter referred to as “provider guardian”) based on the extraction conditions stored in the condition storage unit. In other words, the candidate image shows the resident of the provider. Then, the providing department provides the candidate image to the providing guardian. Further, the extraction unit does not extract the captured image that matches the exclusion condition as a candidate image.

The extraction conditions may be stored in the condition storage unit by setting the presence / absence of application, determination parameters, and the like for each resident (or each guardian). Extraction conditions that apply to all residents, parents, and captured images may exist. Exclusion conditions may be stored in the condition storage unit by setting application status, determination parameters, and the like for each resident (or each guardian). Exclusion conditions that apply to all residents, guardians, and captured images may exist.

The extraction unit may be configured to extract candidate images as candidates for images to be provided to the specific guardian based on the extraction conditions / exclusion conditions corresponding to the specific guardian.

The extraction unit is configured to extract candidate images as candidates for images to be provided to the specific guardian based on the extraction conditions / exclusion conditions corresponding to the specific resident (the resident corresponding to the specific guardian). You may.

The configurations disclosed in the above-described embodiments (including other embodiments, the same shall apply hereinafter) can be applied in combination with the configurations disclosed in the other embodiments as long as there is no contradiction. , The embodiments disclosed in the present specification are examples, and the embodiments of the present invention are not limited thereto, and can be appropriately modified without departing from the object of the present invention.

1: Image provision system 2: Nursery school (facility)
3: Residents 4: Parents (related persons)
5: Staff 7: Photographed image 8: Candidate image 10: Photographing device 11: Hall camera 12: Infant nursery room camera 13: Pediatric nursery room camera 14: Garden camera 15: Play equipment camera 16: Mobile camera 20: Control device 21: Storage unit 21a: Condition storage unit 21b: Image storage unit 22: Control unit 22a: Extraction unit 22b: Input unit 22c: Providing unit 22d: Shooting instruction unit 23: Display input unit 24: Communication unit

Claims (17)

A resident of the facility is identified from the captured image taken at the facility, and a candidate image that is a candidate for an image to be provided to the persons concerned of the resident is extracted based on the extraction conditions stored in the condition storage unit. Extractor and
Determining the similarity of the plurality of candidate images extracted by the extracting unit, a plurality of the candidate image having low similarity to each other, possess a providing unit for providing to the party,
The extraction unit does not extract the captured image that matches the exclusion condition stored in the condition storage unit as the candidate image.
An image providing system characterized by this. The providing unit classifies the candidate images into a plurality of groups based on the information used for determining the extraction conditions, and determines the similarity of the plurality of candidate images based on whether or not they belong to the same group. Item 1. The image providing system according to item 1. The image providing system according to claim 1 or 2 , wherein the extraction condition includes the following collation rate condition.
Collation rate condition: The collation rate between the face of the resident in the captured image and the registered face image of the resident is equal to or higher than a predetermined collation threshold value. The image providing system according to any one of claims 1 to 3, wherein the extraction condition includes the following shooting time condition.
Shooting time condition: The shooting time of the shot image is included in a predetermined extraction target time zone. The image providing system according to any one of claims 1 to 4 , wherein the extraction condition includes the following shooting location condition.
Shooting location condition: The shooting location of the captured image is included in the predetermined extraction target location. The image providing system according to any one of claims 1 to 5 , wherein the extraction condition includes the following facial expression conditions.
Facial expression condition: The facial expression of the resident's face in the photographed image is a predetermined facial expression. The image providing system according to any one of claims 1 to 6 , wherein the extraction condition includes the following number of people condition.
Number of people condition: The number of people in the photographed image is within the predetermined number of people range. The image providing system according to any one of claims 1 to 7 , wherein the extraction condition includes the following personal identification condition.
Personal identification condition: The person specified from the identifier shown in the photographed image matches the resident. The image providing system according to any one of claims 1 to 8 , wherein the extraction condition includes the following volume condition.
Volume condition: The volume when the captured image is captured is within a predetermined volume range. The image providing system according to any one of claims 1 to 9, wherein the exclusion condition includes the following reflection exclusion condition.
Reflection exclusion condition: The face of a predetermined person is shown in the captured image. The image providing system according to any one of claims 1 to 10, wherein the exclusion condition includes the following age exclusion condition.
Age exclusion condition: The photographed image shows the face of a person within a predetermined age range. The image providing system according to any one of claims 1 to 11, wherein the exclusion condition includes the following absence exclusion condition.
Absence exclusion condition: At the time of shooting the captured image, the resident shown in the captured image is absent from the facility. The image providing system according to any one of claims 1 to 12, further comprising an input unit that accepts an input regarding whether or not to provide the candidate image extracted by the extraction unit to the related parties. The condition storage unit stores the extraction conditions in correspondence with the resident.
The image providing system according to any one of claims 1 to 13, wherein the extraction unit extracts the candidate image relating to the resident based on the extraction conditions corresponding to the resident. Claims 1 to 14 having a photographing instruction unit that issues an instruction to photograph the resident when the number of the candidate images extracted by the extraction unit is less than a predetermined number threshold value for the specific resident. The image providing system according to any one of the above. An extraction step of identifying a resident of the facility from the captured image taken at the facility and extracting a candidate image which is a candidate of an image to be provided to the persons concerned of the resident based on the stored extraction conditions.
Determining the similarity of the plurality of the candidate images extracted by the extraction step, a plurality of the candidate image having low similarity to each other, possess a providing step of providing to the party,
In the extraction step, the captured image that matches the stored exclusion condition is not extracted as the candidate image.
An image providing method characterized by that. A resident of the facility is identified from the captured image taken at the facility, and a candidate image that is a candidate for an image to be provided to the persons concerned of the resident is extracted based on the extraction conditions stored in the condition storage unit. Extraction process procedure and
A provision processing procedure for determining the similarity of the plurality of candidate images extracted by the extraction processing procedure and providing a plurality of the candidate images having low similarity to each other to the related parties.
Let the computer run
The extraction processing procedure does not extract the captured image that matches the exclusion condition stored in the condition storage unit as the candidate image.
An image providing program characterized by this.

Images