WO2010073905A1

WO2010073905A1 - Moving image viewing apparatus

Info

Publication number: WO2010073905A1
Application number: PCT/JP2009/070566
Authority: WO
Inventors: 貴則山田
Original assignee: シャープ株式会社
Priority date: 2008-12-25
Filing date: 2009-12-08
Publication date: 2010-07-01

Abstract

Provided is a moving image viewing apparatus capable of efficiently searching for and retrieving a desired scene from a plurality of moving images delivered from a plurality of image pickup apparatuses. If no still image meta information for use in search is existent in a still image meta information storing unit (71), such still image meta information is acquired from among still images delivered over a network or from a still image storing unit (80) in the moving image viewing apparatus, and then image characteristic information is extracted by use of a characteristic extracting unit (60) and then stored into the still image meta information storing unit (71). Characteristic information is extracted, by use of the characteristic extracting unit (60), from the frame data of moving images delivered from the image pickup apparatuses and then stored into a moving image frame meta information storing unit (72). An image search control unit (90) executes a moving image searching process using entered keywords, the still image meta information stored in the still image meta information storing unit (71), and the moving image frame meta information stored in the moving image frame meta information storing unit (72).

Description

Video viewing device

The present invention relates to a moving image viewing apparatus that searches and displays moving images, which are shot by a plurality of moving shooting devices and distributed in real time, using feature information calculated from still images.

Due to the improvement of transmission speed in mobile communications, the application will be used for the transmission and reception of high-quality moving images, and everyone will be able to deliver video in real time at any time. When a photographing device such as an in-vehicle camera or a mobile phone is taking a picture while moving from place to place, if the image is delivered in real time, the content of the scene or delivered video changes from moment to moment. As the number of terminals that distribute such videos increases, viewers need to search and select videos that they want to view efficiently.

In Patent Literature 1, in order for a viewer to select a moving image distributed from a plurality of photographing devices, a method of presenting and selecting a snap video acquired from each photographing device, or a position of the photographing device on a map is displayed. And suggesting a way to choose from.

In Patent Document 2, it is possible to perform a search using an attribute list composed of text information directly attached to an image to be searched or feature information extracted from the image.

JP 2008-154192 A JP 2001-167095 A

However, although the method disclosed in Patent Document 1 provides a means for intuitively selecting a scene to be viewed, when the number of photographing devices increases, how to search for a target moving image from among them is increased. There is no mention of what to do. Further, the method disclosed in Patent Document 2 has a problem of how to perform accurate attribute extraction and feature extraction in the case of a video that is updated in real time and whose contents change from moment to moment.

It is an object of the present invention to provide a moving image viewing apparatus that can efficiently search a desired scene from a plurality of moving images distributed from a plurality of photographing devices.

The present invention is a moving image viewing apparatus that enables a moving image captured by a moving image capturing apparatus and distributed via a network to be searched and viewed using a still image,
A communication unit that communicates by connecting to a network, an input interface unit that receives input from a user, a display unit that presents search results to the user, and a moving image distributed from the network via the communication unit A moving image frame acquisition unit that acquires frame data, a feature extraction unit that extracts feature information indicating image characteristics from the still image used for search and the frame data acquired by the moving image frame acquisition unit, and the feature extraction A meta information storage unit for storing the meta information of the still image and the frame data including the feature information extracted from the unit, and an image search control unit for controlling moving image search,
The image search control unit searches the meta information of the still image of the meta information storage unit based on the search instruction input from the input interface unit, and the meta information of the still image corresponding to the search instruction exists , Calculating the similarity of the feature information of the frame data of the meta information storage unit with respect to the feature information of the still image, selecting the meta information of the frame data in descending order of similarity, and based on the selected frame data meta information The corresponding moving image is acquired via the communication unit and displayed on the display unit.

The image search control unit may select meta information of frame data having feature information whose similarity is a certain value or more, and display the moving image corresponding to the selected frame data on the display unit The user may select a target moving image by the input interface.

The moving image viewing apparatus of the present invention further includes a still image acquisition unit that acquires a still image from a network via the communication unit, and the image search control unit receives a search instruction input from the input interface unit. Based on the above, the still image acquisition unit searches and acquires a still image from the network.

The moving image viewing apparatus of the present invention further includes a still image storage unit that stores a still image, and the image search control unit is configured to store the still image storage unit based on a search instruction input from the input interface unit. It is characterized in that a still image is retrieved and acquired from the image.

Here, the image search control unit searches and acquires a still image when the meta information storage unit has a predetermined number or less of corresponding meta information.

Further, the meta information of the still image includes a keyword for performing a search, and the image search control unit searches based on the keyword input from the input interface unit.

Further, the meta information of the still image and the frame data includes position information that uniquely indicates a location of the image, and the image search control unit acquires an image based on the position information.

The moving image viewing apparatus of the present invention further includes a still image storage unit that stores a still image, and the position information of the still image is information that specifies a still image stored in the still image storage unit. It is characterized by that.

The moving image frame acquisition unit acquires an image at an arbitrarily specified time among the moving images distributed from the moving image capturing device. If the frame closest to the specified time is an intra frame, one moving image frame acquiring unit A frame is acquired, and when a frame closest to a specified time is a predicted frame, an intra frame and a frame including the predicted frame necessary for decoding the frame are acquired.

Further, the moving image frame acquisition unit acquires images from a plurality of the moving image photographing devices specified in advance according to a specified time.

In addition, the feature extraction unit extracts luminance information of each pixel from the image, and outputs the number of pixels having the same luminance value arranged in ascending order, descending order, or predetermined order as feature information. Features.

In addition, the feature extraction unit extracts color information of each pixel from the image, and outputs, as feature information, the number of pixels having the same color arranged in ascending order, descending order, or a predetermined order. And

The present invention uses meta information of still images for search to search for moving images transmitted with contents that change scenes and subjects from moment to moment, and therefore uses still images as keywords or keys. The scene you want to see can be searched efficiently and easily.

1 is a system diagram showing an embodiment of a moving image viewing system. It is a block diagram which shows the structure of the function with which the moving image viewing apparatus of a moving image viewing system is provided. It is a flowchart which shows the procedure from searching a still image on a network based on the input keyword, extracting the characteristic of the acquired still image, and storing in the still image meta-information storage part. It is a figure which shows the example of the still image meta information stored in a still image meta information storage part. It is a flowchart which shows the procedure from acquiring a still image from a still image preservation | save part based on the input keyword, extracting the characteristic of the acquired still image, and storing in a still image meta-information preservation | save part. It is a figure which shows the example of the moving image frame meta information stored in a moving image frame meta information storage part. It is a flowchart which shows the procedure until it calculates the similarity of the still image meta-information extracted based on the search keyword, and the feature information of moving image frame meta-information, and outputs it as a search result.

An example of an embodiment of the present invention will be described below with reference to FIGS.

FIG. 1 is a system diagram showing an embodiment of a moving image viewing system. Reference numerals 11 to 14 denote moving image photographing devices, which are equipped with a communication device and capable of transmitting the images in real time while photographing. For example, a mobile phone or a smartphone equipped with a camera, a digital camera or digital video camera having a communication function, a camera installed in a car, a mobile personal computer equipped with a web camera, or the like. Here, four devices are shown to indicate that a plurality of different devices are mixed, but the number of devices is not limited in actual use.

Reference numerals 15 and 16 denote moving image viewing apparatuses, such as a mobile phone, a personal computer, and a television.
Reference numeral 17 denotes a network in which the moving image capturing device and the moving image viewing device can communicate with each other, and includes, for example, communication using a circuit switching network, a mobile phone network, a wireless LAN, Bluetooth, or the like.

Reference numerals

18 and 19 denote servers connected to the network 17, which can store and distribute still images, moving images taken by the moving image shooting device, and the like. Of course, the moving image capturing apparatus may provide a moving image to the moving image viewing apparatus via the network 17 without using the server.

The moving image viewing apparatuses 15 and 16 receive a search request from the moving image photographing apparatuses 11 to 14 that distribute moving images in real time (or via the servers 18 and 19) at regular intervals or from the viewer. At a certain point, a snapshot of the image being distributed, that is, frame data (still image) of the moving image is acquired. For the still image of the acquired frame data, the feature information is extracted and stored in the apparatus as meta information of the moving image frame together with the acquisition source information (URI). It is assumed that the moving image capturing device as the acquisition source is selected in advance or can be acquired from a portal site (servers 18 and 19) that collects such devices.
Here, the meta information includes image feature information, acquisition source information (URI), acquisition time, position information indicating the location of the image, search keywords, copyright information, and the like. It is embedded in still image data, managed as separate data such as text or XML, and managed as a database. The search keyword is a character string representing an object, a landscape, a person, a state, and the like.

Further, the moving image viewing devices 15 and 16 extract feature information from still images in the devices or still images that can be acquired from the communication network 17 (still images in the servers 18 and 19), and search keywords, It is stored in the apparatus as meta information of a still image together with acquisition source information (information indicating whether a file in the apparatus exists or URI). The search keyword includes a keyword included in metadata included in a still image in advance, a keyword used as a search word when acquiring a still image from the communication network 17, and a keyword directly input by a user from an input device of a terminal. Is used. Furthermore, the meta information of the still image acquired by the same method is stored in the apparatus in advance.

When the viewer inputs a keyword as a search key or specifies and inputs a still image for search, the moving image viewing device extracts feature information of the corresponding still image, and features information of the moving image frame meta information The similarity is calculated, and it is determined that the high similarity is the moving image desired by the viewer, and the search result is presented. Here, the feature information of the extracted still image may be one or plural, and in the case where there are plural pieces, the result obtained by calculating the similarity with one moving image frame meta information is also plural. Therefore, the one with the highest similarity is adopted.

FIG. 2 is a block diagram showing a functional configuration of the moving image viewing apparatus of the moving image viewing system. As shown here, the moving image viewing apparatus includes a communication unit 20, a moving image frame acquisition unit 30, a still image acquisition unit 40, an image decoding unit 50, a feature extraction unit 60, a meta information storage unit 70, A still image storage unit 80, an image search control unit 90, an input interface unit 100, and a display unit 110 are provided.

The communication unit 20 exchanges data with the

servers

18 and 19 and the moving image photographing apparatuses 11 to 14 via the communication network 17.
The moving image frame acquisition unit 30 is a snapshot (moving image frame) of a moving image that is distributed at a certain time interval or when a search request is received from a viewer via the communication unit 20, that is, Get a still image.
The still image acquisition unit 40 acquires a still image distributed to the communication network 17 via the communication unit 20.
The image decoding unit 50 has a function of decoding a moving image and a function of decoding a still image, and are called a moving image decoding unit 51 and a still image decoding unit 52, respectively.

The feature extraction unit 60 extracts feature information (details will be described later) from the decoded image.
The meta information storage unit 70 has a function capable of separately storing the feature information extracted from the still image and the feature information extracted from the moving image, and the storage destination is the still image meta information storage unit 71, This is called a moving image frame meta information storage unit 72.
The still image storage unit 80 stores still images.
The image search control unit 90 controls each unit when moving image search processing is performed.

The input interface unit 100 has a function of inputting a keyword for performing a search and a function of selecting a moving image to be viewed, and are referred to as a search key input unit 101 and a moving image designating unit 102, respectively. The input interface unit 100 includes hardware such as an input device such as a keyboard, a mouse, a touch panel, and a button, and a pointing device. The search key input unit 101 can input a scene that the viewer wants to see by directly inputting a keyword or the like, selecting from a keyword or genre registered in advance, or selecting an image close to the scene to be viewed.
The display unit 110 can present a search result obtained by inputting a search key to a viewer or display a moving image selected from the viewer, and these can be displayed as a search result display unit 111 and a moving image display, respectively. This is referred to as part 112.

FIG. 3 shows a procedure from when the moving image viewing apparatus searches for a still image on the network based on the input keyword, extracts the characteristics of the acquired still image, and stores them in the still image meta information storage unit 71. It is a flowchart which shows. This is a partial extraction of the processing from search to viewing according to the present invention.

Here, first, in step S11, based on the keyword input from the search key input unit 101 of the input interface unit 100, the image search control unit 90 captures a still image on the communication network 17 via the communication unit 20. Search for. The search method is not particularly limited. For example, in the Internet, when some keywords are input and transmitted to the Web server, there are some services that create a list of still images corresponding to the keywords from the database in the Web server. You can use this as well. Later, in order to further improve the accuracy of the search result of moving images, a plurality of images are acquired from the plurality of searched still images.

Next, in step S12, the image search control unit 90 confirms whether or not the acquired still image has already been acquired and the feature information has been extracted. , And remove it from the target of the subsequent processing. Whether or not it has been extracted can be determined by information indicating a still image acquisition source stored in the still image meta information storage unit 71, that is, by a URI. If the feature information has already been extracted from all of the acquired still images, the process shown in FIG. 3 ends here. If there is an unextracted still image, the process proceeds to the next process.

Next, in step S13, the image search control unit 90 causes the still image acquisition unit 40 to acquire a still image from which patent information has not been extracted, and sets the still image in a format that can extract feature information. Therefore, the still image decoding unit 52 performs decoding.

Next, in step S14, the feature extraction unit 60 extracts the feature information of the decoded still image. The feature information is information that can be used to search for an original image or an image similar to the original image using the information as a clue. The feature information extraction method is not limited in the present invention. For example, the method of recording the color distribution of the image, the method of extracting the feature point of the image, the texture in the image, the luminance information, or the color information There is a method of extracting statistical information, that is, a histogram.

In the texture data histogram, for example, for all the pixels in the image, compared to the brightness of the surrounding eight pixels, whether or not more than half of the surrounding pixels are brighter than the brightness of the pixel of interest. Is obtained, and the number of pixels is totaled for each luminance and arranged for each luminance.
The histogram of color information is obtained, for example, by extracting the degree of red, green, and blue colors for each pixel, and counting and arranging the number of pixels for each color.

Alignment may be in ascending order or descending order in terms of the number of pixels, or in ascending order or descending order in terms of luminance or color. Alternatively, a predetermined arrangement may be used.

Next, in step S15, the feature extraction unit 60 collects the keyword input in step S11, the information (URI) of the still image acquisition source, and the feature information extracted in step S14 into a still image meta information. Store in the storage unit 71.

Here, when there are a plurality of still images acquired in step S11, the processing in steps S12 to S15 may be repeated for each still image, or the processing from steps S12 to S15 is performed for all the images. You may go in order.

FIG. 4 is a diagram illustrating an example of still image meta information stored in the still image meta information storage unit 71 as a result of the series of processes in FIG. Still image meta information numbers 1 to 3 are still meta image information extracted by keywords inputted so far, and still image meta information number 4 is added with still image meta information obtained by newly inputted keywords. Indicates that In this way, if still image meta information corresponding to a keyword input in the past is accumulated, the search speed when the same keyword is input can be increased.

The image search control unit 90 does not make the determination in step S12, and once performs the processing up to step S14, compares the feature information, and does not newly store still image meta information if they match completely. It is also possible. In addition, if the feature information matches but the keyword does not match, only the keyword can be added to the keyword portion of the stored still image meta information. As a result, the same feature information has a different meaning key. Can also be used.

FIG. 5 is a flowchart showing a procedure from acquiring a still image from the still image storage unit 80 based on the input keyword, extracting features of the acquired still image, and storing the feature in the still image meta information storage unit 71. It is. This is a partial extraction of the processing from search to viewing according to the present invention.

Here, first, in step S21, the image search control unit 90 includes a still image that includes the keyword input from the search key input unit 101 as a part of the file name or has as file metadata. It is confirmed whether it exists in the preservation | save part 80, and when it exists, it is acquired.

Next, in step S <b> 22, the image search control unit 90 confirms whether any of the still images stored in the still image storage unit 80 has already been acquired and feature information has been extracted. However, those that have already been extracted are excluded from the target of subsequent processing. Whether or not it has been extracted can be determined by information (see FIG. 4) indicating a still image acquisition source stored in the still image meta information storage unit 71, that is, a file name. If the feature information has already been extracted from all of the stored still images, the processing shown in FIG. 5 ends here. If there is an unextracted still image, the process proceeds to the next process.

Next, in step S23, the image search control unit 90 sets the still image from which the patent information stored in the still image storage unit 80 has not been extracted into a format in which the feature information can be extracted. Decoding is performed by the image decoding unit 52.
Next, in step S24, the feature extraction unit 60 extracts the feature information of the decoded still image.

In step S25, the feature extraction unit 60 stores the keyword input in step S21, the still image file path, and the feature information extracted in S24 together in the still image meta information storage unit 71.

The saved information is saved as in the example shown in FIG.

Next, a procedure for acquiring information for searching from a moving image to be searched will be described.

An apparatus for viewing a moving image (moving image viewing apparatus) needs to have information on a moving image acquisition source, that is, a moving image photographing apparatus, in advance. For example, several photographing devices are registered in advance, information on each photographing device can be registered by a viewer, or information is acquired from a portal site that collects information on moving image photographing devices. There are methods.

The moving image frame acquisition unit 30 of the moving image viewing apparatus sequentially takes a snapshot of the moving image being distributed at a certain time interval or at the time when a search request is received from the viewer. (Moving image frame) is acquired. For example, an image at an arbitrarily specified time is acquired from the moving images distributed from the moving image capturing device, and if the frame closest to the specified time is an intra frame, one intra frame is acquired and specified. If the frame closest to the predetermined time is a predicted frame, a frame including an intra frame and a predicted frame necessary for decoding the frame is acquired.

The image acquired by the moving image frame acquisition unit 30 is decoded by the still image decoding unit 52 of the image decoding unit 50, and the feature extraction unit 60 extracts features by the same method as the feature extraction of the still image.

The feature extraction unit 60 stores the extracted feature information in the moving image frame meta information storage unit 72 together with information (URI) indicating the acquisition source. At this time, if moving image frame meta information having the same information indicating the acquisition source already exists in the moving image frame meta information storage unit 72, the feature extraction unit 60 overwrites and stores the information. FIG. 6 is an example of moving picture frame meta information stored in the moving picture frame meta information storage unit.

Next, the flow from when the viewer searches to view a moving image will be described with reference to FIG.

In step S31, first, the viewer is presented with a screen for searching for a moving image on the display unit 110. By inputting some information here, the moving image to be viewed can be searched and selected. For example, one or a plurality of keywords are input from the search key input unit 101 of the input interface unit 100, selected from a keyword list, or a list of images that represent scenes to be viewed is displayed and selected from there. And so on.

Subsequently, in step S32, it is determined whether or not a still image used in the search is newly acquired from the network. The determination condition is whether or not the still image meta information corresponding to the keyword input in step S31 exists in the still image meta information storage unit 71. Normally, if this does not exist, a moving image cannot be searched unless a new still image is acquired. However, if it exists, it is not necessary to newly acquire it. However, since the search accuracy of moving images is improved by using meta information of a plurality of still images, for example, as a result of the search, the still image meta information storage unit 71 If only a certain number or less of still image meta information can be secured, a still image for search is acquired from the network. If it is expected in step S34 that the still image in the apparatus can be used for search purposes, or if it is not necessary to find one as a search result, there is no need to obtain it.

If yes, go to step S33, otherwise go to step S34. In step S33, the series of processing shown in FIG. 3 is performed, and the meta information including the acquired still image feature information is stored in the still image meta information storage unit 71.

In step S34, the image search control unit 90 determines whether to use a still image in the apparatus as a still image used in the search. The determination condition is whether or not the still image meta information corresponding to the keyword input in step S31 exists in the still image meta information storage unit 71. Normally, if it exists, it is not necessary to use a still image in the apparatus as a search image, but it is desirable to use it in order to improve the accuracy of moving image search.

If it is used, the process proceeds to step S35. If it is not used, the process proceeds to step S36. In step S35, the series of processing shown in FIG. 5 is performed, and the feature information is stored in the still image meta information storage unit.

Here, the determination conditions in steps S32 and S34 may be settings unique to the viewing device, or may be set freely by the viewer.

In step S36, the image search control unit 90 inputs the keyword, the still image meta information stored in the still image meta information storage unit 71, and the moving image frame meta information stored in the moving image frame meta information storage unit 72. Then, the moving image search process is performed.

First, from the still image meta information stored in the still image meta information storage unit 71, still image meta information whose keyword matches part or all of the input keyword is extracted. Here, it does not matter how much the keywords match, but if the number of extracted still image meta information is large, an appropriate number of still image meta information is determined based on the degree of matching of the keywords, etc. It is good to extract.

Next, the similarity between the extracted feature information of the still image meta information and the feature information of the moving image frame meta information stored in the moving image frame meta information storage unit 72 is calculated. The similarity calculation method differs depending on the feature information format, and a plurality of calculation methods can be considered even if the feature feature format is the same. Therefore, although not particularly limited in the present invention, for example, as described above, the brightness of the image When the information histogram is used as the feature information, the distance between the values for each luminance in the feature information is calculated, and the sum of the calculated distances can be used as a scale representing the similarity. In this method, the similar value is smaller, and the value is larger if they are not similar. Also, with this method, even if images are taken of the same subject, the similarity will be different if the resolutions are different, so resizing so that the image size will be the same number of pixels before performing feature extraction. It is preferable to calculate the similarity after performing the feature extraction or by making the scales of the histograms equal before calculating the similarity, that is, by making the sum of all the values in the histogram the same.

After calculating the similarity between all the extracted still image meta information and all the moving image frame meta information in the same procedure, arrange the moving image frame meta information in descending order of the similarity, and then extract the URI from that and search As a result.

The search result is displayed on the search result display unit 111 in step S37. The search results may be displayed in order of URIs, or the corresponding snapshot images acquired earlier may be displayed instead of the URIs. If possible, a moving image may be received from each photographing apparatus and displayed.

Next, the viewer designates a moving image desired to be viewed from the search results displayed on the search result display unit 111 by the moving image designating unit 102. The image search control unit 90 issues a viewing start request to the URI specified via the communication unit 20. When a moving image is received from the moving image capturing device, the moving image decoding unit 51 decodes the moving image and the moving image display unit 112 displays the moving image.

In this way, the viewer simply enters the search keyword and selects the desired moving image from the search results, and the content of the scene or video to be distributed changes from moment to moment. However, it is possible to easily and efficiently search for a target moving image and view a moving image of a scene to be viewed.
Note that the present invention assumes a case where a moving image is distributed via a communication network. However, the present invention is also applicable to moving image distribution by broadcasting. In that case, the acquisition source information held as metadata is channel information or frequency information. The search keyword can also be used when broadcast information includes character information such as program information.
Furthermore, although the present invention assumes a moving image photographed by a photographing apparatus, it can also be applied to an animation moving image.

11 to 14 Moving image capturing devices 15 and 16 Moving image viewing device 17

Network

18 and 19 Server 20 Communication unit 30 Moving image frame acquisition unit 40 Still image acquisition unit 50 Image decoding unit 51 Moving image decoding unit 52 Still image decoding unit 60 Features Extraction unit 70 Meta information storage unit 71 Still image meta information storage unit 72 Moving image frame meta information storage unit 80 Still image storage unit 90 Image search control unit 100 Input interface unit 101 Search key input unit 102 Moving image designation unit 110 Display unit 111 Search result display unit 112 Moving image display unit

Claims

A moving image viewing device that enables a moving image captured by a moving image capturing device and distributed via a network to be searched and viewed using a still image,
A communication unit that communicates by connecting to a network;
An input interface unit that receives input from the user;
A display for presenting search results to the user;
A moving image frame acquisition unit that acquires frame data of a moving image distributed from the network via the communication unit;
A feature extraction unit that extracts feature information indicating image features from the still image used for search and the frame data acquired by the moving image frame acquisition unit;
A meta information storage unit that stores the meta information of the still image and the frame data including the feature information extracted by the feature extraction unit;
An image search control unit for controlling moving image search;
With
The image search control unit
Based on the search instruction input from the input interface unit, search the meta information of the still image of the meta information storage unit,
When there is still image meta information corresponding to the search instruction, the similarity of the frame data feature information of the meta information storage unit with respect to the still image feature information is calculated, and the frame data meta information in descending order of similarity is calculated. Select information,
A moving image viewing / listening device, wherein a corresponding moving image is acquired via the communication unit based on meta information of the selected frame data and displayed on the display unit.
2. The moving image viewing apparatus according to claim 1, wherein the image search control unit selects meta information of frame data having feature information having a degree of similarity equal to or greater than a predetermined value.
The image search control unit displays the moving image corresponding to the selected frame data on the display unit, and allows the user to select a target moving image through the input interface. The moving image viewing device according to 1.
A still image acquisition unit that acquires a still image from the network via the communication unit;
The image search control unit causes the still image acquisition unit to search for and acquire a still image from a network based on a search instruction input from the input interface unit. The moving image viewing device according to 1.
A still image storage unit for storing still images;
5. The image search control unit according to claim 1, wherein the image search control unit searches for and acquires a still image from the still image storage unit based on a search instruction input from the input interface unit. Video viewing device.
The moving image viewing apparatus according to claim 4 or 5, wherein the image search control unit searches and acquires a still image when the meta information storage unit has a predetermined number or less of corresponding meta information. .
The meta information of the still image includes a keyword for performing a search,
The moving image viewing apparatus according to claim 1, wherein the image search control unit searches based on a keyword input from the input interface unit.
The meta information of the still image and the frame data includes position information that uniquely indicates the location of the image,
The moving image viewing apparatus according to claim 1, wherein the image search control unit acquires an image based on the position information.
The moving image viewing / listening device according to claim 8, wherein the position information is information indicating a position of an image on a network that can be acquired via the communication unit.
A still image storage unit for storing still images;
9. The moving image viewing apparatus according to claim 8, wherein the position information of the still image is information for specifying a still image stored in the still image storage unit.
The moving image frame acquisition unit acquires an image at an arbitrarily specified time among the moving images distributed from the moving image capturing device. If the frame closest to the specified time is an intra frame, one moving image frame acquiring unit 11. A frame is acquired, and when a frame closest to a specified time is a predicted frame, an intra frame and a frame including the predicted frame necessary for decoding the frame are acquired. The moving image viewing device according to any one of the above.
The moving image viewing apparatus according to claim 1, wherein the moving image frame acquisition unit acquires images according to a specified time from a plurality of the moving image capturing devices specified in advance. .
The feature extraction unit extracts brightness information of each pixel from an image, and outputs the number of pixels having the same brightness value as feature information arranged in ascending order, descending order, or a predetermined order. The moving image viewing apparatus according to claim 1.
The feature extraction unit extracts color information of each pixel from an image, and outputs the feature information obtained by arranging the number of pixels having the same color in ascending order, descending order, or a predetermined order. The moving image viewing apparatus according to claim 1.