JP2008288804A - Image reproducing device, and control method and control program of image reproducing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reproduce images by automatically impart various kinds of effects corresponding to image contents of the respective images when reproducing the plurality of images as moving images, still images or their combination. <P>SOLUTION: When reproducing a plurality of reproduction object images including the moving images, the still images or their combination, an image reproducing device 10 respectively extracts the feature quantities of the respective images included in the reproduction object images, assumes an n-dimensional image feature quantity space indicated by n kinds (n is an integer ≥2) of feature quantities, defines the extracted feature quantities of the respective images as attribute vectors or space coordinates in the n-dimensional image feature quantity space, and controls the reproduction of the plurality of reproduction object images on the basis of the attribute vectors or the space coordinates. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image playback device, a control method of the image playback device, and a control program. In particular, when a plurality of images as a moving image, a still image, or a combination thereof are played back, an image effect and BGM are automatically provided. And technology for playback.

Conventionally, when reproducing a plurality of images as a moving image, a still image, or a combination thereof, it is necessary to manually set what kind of effect is to be reproduced for each image. When playing back using images, it takes a lot of time and is a very complicated operation for users who are not used to it.
For example, when creating a slide show using a plurality of still images, if there is no image effect or BGM of the image, or if it is monotonous such as always using the same image effect or BGM, there is no production effect, boring And the playback effect is lost.
Furthermore, if these image effects and BGM are performed regardless of the image being reproduced, the viewer cannot feel the story from the image.
JP 2006-186505 A

In order to solve this, the technique described in Patent Document 1 detects the silent section so that the playback time of the image and the BGM coincide with each other. However, the content of the image is not considered, There is a problem that it is not possible to reproduce a BGM that is more suitable for an image or to impart a suitable image effect to an image to be reproduced.
In recent years, with the spread of digital cameras, digital videos, camera-equipped mobile phones, etc., content such as images stored by individuals is increasing. Digital image data, which is representative content, is often added with metadata including information indicating the shooting date and time by the function of the digital camera. If the information contained in this metadata is used, for example, digital image data taken on a specific day can be retrieved, which helps to manage the digital image data. However, the metadata is only used for management and has not been used for playback control.
Therefore, an object of the present invention is to reproduce images by automatically applying various effects according to the image contents of each image when reproducing a plurality of images as a moving image, a still image, or a combination thereof. It is an object to provide an image reproducing apparatus, a control method for the image reproducing apparatus, and a control program.

In order to solve the above-described problem, a first aspect of the present invention is an image reproduction device that reproduces a plurality of reproduction target images including moving images, still images, or a combination thereof, and each image included in the reproduction target images. Assuming a feature quantity extraction unit that extracts each of the feature quantities and an n-dimensional image feature quantity space represented by n types (n is an integer of 2 or more), the feature quantity extraction unit extracts the feature quantity. A reproduction control unit that performs the reproduction control of the plurality of reproduction target images based on the attribute vector or the spatial coordinates, and the feature quantity of each of the images as an attribute vector or a spatial coordinate in the n-dimensional image feature quantity space; It is characterized by having.
According to the above configuration, the feature amount extraction unit extracts the feature amount of each image included in the reproduction target image.
Accordingly, the reproduction control unit sets the feature amount of each image extracted by the feature amount extraction unit as an attribute vector or a spatial coordinate in the n-dimensional image feature amount space, and based on the attribute vector or the spatial coordinate, Perform playback control of the playback target image.
Therefore, when reproducing a plurality of images as a moving image, a still image, or a combination thereof, image reproduction can be performed by automatically applying various effects according to the feature amount of each image.

According to a second aspect of the present invention, there is provided a music vector in which a plurality of music data is allocated to a vector space having at least a season and a time axis as a space axis based on the characteristics of the music data in the image reproduction device of the first aspect. A music vector space storage unit that stores spatial information, and includes information related to a shooting date and time when the image was shot as the feature amount, and the reproduction control unit refers to the music vector space information and the shooting date and time is The music data is selected based on the season and time to which it belongs, and the selected music data is assigned as BGM of the image.
According to the above configuration, the music vector space storage unit stores music vector space information in which a plurality of music data is assigned to a vector space having at least the season and time as a space axis based on the characteristics of the music data. The playback control unit refers to the music vector space information, selects music data based on the season and time to which the shooting date belongs, and assigns it as the BGM of the image, so the music corresponding to the season and time to which the shooting date belongs Data is automatically played back as BGM.
According to a third aspect of the present invention, in the image reproduction device of the first aspect or the second aspect, a TF / IDF value of an important word contained in a document related to the image is used as the feature amount. Yes.
According to the above configuration, since the reproduction control can be performed based on the more important important words included in the document related to the image based on the TF / IDF value, the content of the image is improved for the viewer. It is possible to perform playback control without any sense of incongruity.

According to a fourth aspect of the present invention, in the image reproduction device according to any one of the first to third aspects, metadata previously given to the image is used as the feature amount.
According to the above configuration, it is possible to perform reproduction control without any sense of incongruity according to the image content.
According to a fifth aspect of the present invention, in the image reproducing device according to the fourth aspect, the metadata includes at least one of a shooting date and time or a shooting location.
According to the above configuration, it is possible to perform optimal image reproduction control in accordance with the image content in association with the shooting date and time or the shooting location.

According to a sixth aspect of the present invention, in the image reproduction device according to any one of the first to fifth aspects, information relating to a shooting location extracted from a map database and the vicinity of the shooting location is used as the feature amount. It is said.
According to the above configuration, the information regarding the shooting location and the vicinity of the shooting location extracted from the map database includes information related to the image to be played back, and optimal image playback control can be performed according to the image content.
Further, according to a seventh aspect of the present invention, in the image reproduction device according to the second aspect, the dimension of the attribute vector or the spatial coordinate when the dimension of the attribute vector or the spatial coordinate corresponding to the image exceeds n dimensions. It is characterized by having a dimension reduction unit that reduces
According to the above configuration, the dimension reduction unit reduces the dimension of the attribute vector or the spatial coordinate when the dimension of the attribute vector or the spatial coordinate corresponding to the image exceeds n dimensions. Image playback control can be performed.

Further, an eighth aspect of the present invention is a control method for an image playback device that plays back a plurality of playback target images including moving images, still images, or a combination thereof, and features of each image included in the playback target images Assuming a feature quantity extraction process for extracting quantities respectively, and an n-dimensional image feature quantity space represented by n types (n is an integer of 2 or more), each feature quantity extraction unit extracts each A reproduction control step of setting the feature quantity of the image as an attribute vector or a spatial coordinate in the n-dimensional image feature quantity space, and performing reproduction control of the plurality of reproduction target images based on the attribute vector or the spatial coordinate. It is characterized by that.
According to the above configuration, when reproducing a plurality of images as a moving image, a still image, or a combination thereof, image reproduction can be performed by automatically applying various effects according to the feature amount of each image. it can.

According to a ninth aspect of the present invention, there is provided a control program for controlling, by a computer, an image playback device that plays back a plurality of playback target images including moving images, still images, or combinations thereof. A feature amount of each image included is extracted, an n-dimensional image feature amount space represented by n types (n is an integer of 2 or more) of the feature amount is assumed, and the feature of each of the extracted images A quantity is an attribute vector or a spatial coordinate in the n-dimensional image feature quantity space, and reproduction control of the plurality of reproduction target images is performed based on the attribute vector or the spatial coordinate.
According to the above configuration, when reproducing a plurality of images as a moving image, a still image, or a combination thereof, image reproduction can be performed by automatically applying various effects according to the feature amount of each image. it can.
In this case, the control program may be recorded on a computer-readable recording medium.

Next, preferred embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic configuration block diagram of an image reproducing apparatus according to an embodiment.
The image reproduction device 10 is configured as a personal computer, and includes an image reproduction device main body 11 that performs image processing, an image input device 12 that performs various image inputs, and an input device 13 such as a keyboard and a mouse for performing various operations. A display device 14 such as a liquid crystal display that performs various displays; an output device 15 such as a printer that performs printing; and an external storage device 16 such as a hard disk device that stores various data including data related to an image reproduction method to be described later. And.
The image reproduction apparatus main body 11 includes a microprocessor unit (MPU) 21 that controls the entire image reproduction apparatus main body 11, a ROM 22 that stores various data including various control programs, a RAM 23 that functions as a work area of the MPU 21, And a communication interface unit 24 that performs an interface operation with an external communication network such as a LAN or the Internet.

FIG. 2 is a functional block diagram of the image reproducing apparatus.
The image reproduction device 10 is roughly classified into a video / image input unit 31, a text extraction unit 32, a morpheme analysis unit 33, an importance calculation unit 34, a dimension reduction unit 35, and an image reproduction method storage unit 36. It is equipped with.
The video / image input unit 31 inputs moving image data (corresponding to moving images) or still image data (corresponding to still images) via a broadcast radio wave, a communication network, or a data storage medium such as an IC memory card. Specifically, images of various image formats are extracted or received from files having various compression formats and file formats, and input devices such as scanners, digital cameras, and digital video cameras. In this case, compression formats include JPEG, MPEG-4, H.264. H.264 or the like.

The text extraction unit 32 receives text data other than tag information from hypertext data or map database described in XHTML (personal homepage, blog, etc.), sent / received e-mails, XML (music search metadata, etc.). Extract.
The morpheme analysis unit 33 performs morpheme analysis on text corresponding to text data such as a blog or text data speech-recognized by the speech recognition unit, extracts words (phrases), and performs stemming as necessary.

The importance calculation unit 34 extracts a word / phrase as metadata from a related content (denoted as content i) that is most closely related from a plurality of related content.
In this case, a phrase representing the related content, that is, a phrase having a high importance in the related content is regarded as a phrase that best expresses a moving image or a still image, and is expressed as an important phrase.
Here, the importance is calculated by TF / IDF (Term Frequency & Inverse Document Frequency) processing. Then, the importance calculation unit 34 obtains the importance tfidf (i, j) of each word / phrase j extracted from the content i by the morpheme analysis unit 33 by the following equation (1).

This will be specifically described below.
First, the appearance frequency of the phrase j in the content i is set to Freq (i, j), and tf (i, j) related to the appearance frequency in the related content is calculated.
tf (i, j) = log (Freq (i, j) +1) / log (total number of phrases of content i)
Here, tf (i, j) is an index representing the importance of a phrase based on the appearance frequency of related content in text data. In other words, phrases that frequently appear in the content are regarded as important phrases.

Next, idf (j) is calculated with df (j) as the number of contents in which the phrase j appears.
idf (j) = log (N / dt (j))
Here, idf (j) is calculated from the number of text data of the content in which the phrase t appears among all the searched contents as the appearance frequency of the phrase t.
Furthermore, the importance of the phrase j in the content i is
tfidf (i, j) = tf (i, j) × idf (j)
And Therefore, when selecting the search metadata, a predetermined number may be selected in descending order of the value of tfidf (i, j).

The dimension reduction unit 35 assumes an n-dimensional image feature amount space represented by n types (n is an integer of 2 or more) of image feature amounts, and extracts the feature amount of each extracted image as an n-dimensional image feature amount space. When the spatial coordinate or vector corresponding to the actual image reproduction method is lower, that is, when n> m, the dimension n of the image to be reproduced is compressed. Then, the dimensions are compressed so as to match the spatial coordinates or vector dimension m corresponding to the actual image reproduction method.
The image reproduction method storage unit 36 can be broadly divided into an image frame (image frame) and a color palette changing unit 41 that changes the color that decorates the image, an image switching, and a transformation of the image itself, thereby producing an effect ( An image effect providing unit 42 that generates an image effect) and a BGM adding unit 43 that reproduces music (BGM) in synchronization with the reproduction of the image.

In the following description, a case where the feature amount of an image is handled as a vector will be described.
First, generation of an attribute vector as a feature amount of each image will be described.
Examples of the attribute vector as the feature amount of each image include shooting date / time, shooting location, document, and metadata.
When shooting date / time is used as the attribute vector, if n is the number of days since the beginning of the year and Year = 365, θ is

と定義することができる。
撮影日時の属性ベクトルａは、

Can be defined as
The attribute vector a of the shooting date is

と表すことができる。
撮影時刻ｔまで考慮に入れる場合は、Ｄａｙ＝２４とすれば、φは

と定義することができる。
また、属性ベクトルａは、経過日数と合わせて、

It can be expressed as.
When taking into account shooting time t, if Day = 24, φ is

Can be defined as
In addition, the attribute vector a is combined with the elapsed days,

としてもよい。
ここで、撮影日時を周期関数で写像しているのは、年または日の区切りを遷移的にするためである。
さらに、経過日数、または時刻をそのまま属性ベクトルの成分とするようにしても良い。
属性ベクトルとして、撮影場所を用いる場合には、撮影場所の緯度をｘ、経度ｙとすれば、

It is good.
The reason why the shooting date and time is mapped by a periodic function is to make the year or day separator transition.
Furthermore, the elapsed days or time may be used as the component of the attribute vector as it is.
When the shooting location is used as the attribute vector, if the latitude of the shooting location is x and longitude y,

として表すことができる。
あるいは、撮影場所の平均位置の緯度をｘ₀、経度をｙ₀として、

Can be expressed as
Alternatively, the latitude of the average location of the shooting location is x ₀ and the longitude is y ₀ .

として表すことができる。
属性ベクトルとして、文書（あるいは文書内のテキスト）を用いる場合には、画像に関連する文書の特徴を使用することとなる。
ここで、画像に関連する文書とは、例えば、以下のようなものが挙げられる。
撮影日時や撮影場所、イベントに関連した記述が含まれる、ニュースサイトや観光地のガイド、撮影日時および前後に送受信した電子メール、更新されたブログの内容などである。

Can be expressed as
When a document (or text in the document) is used as the attribute vector, the document features related to the image are used.
Here, examples of the document related to the image include the following.
These include news sites and sightseeing spot guides, shooting date and time, e-mails sent and received before and after, and updated blog contents.

More specifically, in the document set, the appearance frequency of the document i, the word j, and the word j appearing in the document i is set to freq (i, j), and tf (i, j) related to the appearance frequency in the document. Is defined as follows.
tf (i, j) = log (freq (i, j) +1)
/ Log (number of types of words appearing in document i)
The number of documents in which the word j appears is represented by df (j), and the reciprocal number considering the number N of document sets is
idf (j) = log (N / df (j))
TF / IDF value w (i, j) of word j in document i is
w (i, j) = tf (i, j) × idf (j)
It becomes. The attribute vector of document i is w (i, j)

そして、文書集合のうち、対応する画像に最も関係が深いと思われる、文書の属性ベクトルを用いればよい。
属性ベクトルとして、ＴＶ番組や画像データベース等のように、あらかじめ画像に付与された検索用のメタデータがある場合には、付与されたメタデータとして用いられている語を、上述した文書に含まれている語と同様に扱えばよい。
以上の説明は、属性ベクトルを単独に用いる場合のものであったが、上述した手法により生成した複数の属性ベクトルを合成し、さらに高次元のベクトルを生成して画像再生方法と細かく対応させるようにしても良い。
例えば、生成した複数の属性ベクトルａ₁、ａ₂を次式のように合成し、新しい属性ベクトルａ₃を生成しても良い。

Then, a document attribute vector that is considered to be most closely related to the corresponding image in the document set may be used.
When there is search metadata previously assigned to an image, such as a TV program or an image database, as an attribute vector, the word used as the assigned metadata is included in the above-described document. It can be treated in the same way as
The above description is for the case where attribute vectors are used alone, but a plurality of attribute vectors generated by the above-described method are combined, and a higher-dimensional vector is generated so as to correspond closely to the image reproduction method. Anyway.
For example, a plurality of generated attribute vectors a ₁ and a ₂ may be synthesized as in the following equation to generate a new attribute vector a ₃ .

ここで、Ｃ₁〜Ｃ_m、Ｃ_nはベクトルの成分の重み付け係数である。

Here, C _{1 to} C _m and C _n are weighting coefficients of vector components.

Next, a specific method for assigning the color used for the background of the image or the image frame using the document (or text in the document) as the attribute vector will be described.
FIG. 3 is an explanatory diagram of an example in which colors used for the background of an image or an image frame are allocated two-dimensionally as an image reproduction method.
FIG. 3 shows the color used by the color palette changing unit 41 of the image reproduction method storage unit 36 for the image background and image frame as a document (or text in the document) as an attribute vector. This is the case where selection is made in correspondence with.
Specifically, if the word contained in the document is “daughter, wedding, joy, loneliness”, the color used for the background color or image frame comes from loneliness in addition to “joy”. A color ranging from purple to blue-purple in consideration of “sadness” will be selected.
In addition, when the word included in the document is “family trip, hot spring, fun”, the color used for the background color or image frame ranges from yellow to yellow-red considering “fun”. Will be selected.

Next, a specific method for assigning an image effect as an image reproduction method to an image using a shooting date and time as an attribute vector will be described.
FIG. 4 is an explanatory diagram in the case of assigning slide-in and strip-in (crossfade), which are image effects of image switching, to a three-dimensional space.
FIG. 4 shows a case where the image effect imparting unit 42 of the image reproduction method storage unit 36 assigns slide-in and strip-in (cross fade), which are image effects of image switching, using the shooting date and time as an attribute vector. .
For example, in the XY plane, a slide-in from the right or a strip-in from the right is assigned to the first quadrant, and a slide-in from the top or a strip-in from the top is assigned to the second quadrant. A slide-in from the left or a strip-in from the left is assigned to the third quadrant, and a slide-in from the bottom or a strip-in from the bottom is assigned to the fourth quadrant.
The Z-axis is assigned image effect intensity (strong-weak) or speed (fast-slow).
Therefore, when the shooting date is “November 15 6:00 pm”, the image effect is given so that the slide-in from the left is weak (weakest).

Next, a specific method for assigning BGM to an image using a shooting date and time as an attribute vector will be described.
FIG. 5 is an explanatory diagram when BGM is assigned to a two-dimensional space.
When the BGM is assigned using the shooting date and time, as shown in FIG. 5, the BGM can be arranged on the vector space based on the human impression.
Specifically, March-May is a classic song reminiscent of spring, June-August is a classic song reminiscent of summer, September-November is a classic song reminiscent of autumn, December-February is Assign a classic song reminiscent of winter.
In this case, a shooting time as an attribute vector is further assigned in the vertical direction on the page, a song reminiscent of daytime is arranged on the front side of the page, and a song reminiscent of night is arranged on the back side of the page. If it is “April 5, 5:00 am”, assign a classic song reminiscent of the spring dawn, and if the shooting date is “October 15 1:00 pm”, remind it of the autumn night length Assign a classic song.
In this case, in principle, the music piece closest to the attribute vector of the image is selected. However, when the music selection becomes monotonous, it is the same as the attribute vector of the image so that a change occurs. You may make it select the other music of a quadrant. In addition, the information which allocated the music to the vector space is memorize | stored in the external storage device 16 (equivalent to a music vector space memory | storage part).

Further, for example, BGM may be automatically classified based on the music feature amount by a technique disclosed in Japanese Patent Laid-Open No. 2003-58147. Specifically, it can be classified into a three-dimensional space according to the tempo, key, and frequency characteristics of BGM, and any BGM may be assigned in accordance with the shooting date and time (including the shooting time).
The above description has been made on the assumption that the dimension number of the attribute vector is equal to or less than the color, image effect, or BGM dimension to be allocated. However, the attribute vector of the document or the higher-dimensional vector is allocated. If the number of dimensions is higher than the number of dimensions of the image reproduction method such as power color, image effect, BGM, etc., it cannot be handled as it is.
Therefore, in such a case, the dimension reduction unit 35 performs principal component analysis on a set of attribute vectors, reduces the number of dimensions according to the required number of dimensions, or mainly corresponds to attribute vectors to be considered. The axis is selected as the main axis.

FIG. 6 is a flowchart of an outline process according to the embodiment.
For the image input through the image input device 12 that functions as the video / image input unit 31, the MPU 21 of the image playback device main body 11 uses the metadata ( In the case of the present embodiment, an attribute vector is generated based on the shooting date / time and shooting location) and the document (or text in the document) related to the input image (step S11).
In this case, the MPU 21 of the image reproduction device main body 11 can also generate an attribute vector based on a plurality of data.
Next, the MPU 21 of the image reproduction device 10 functions as the dimension reduction unit 35. When the number of dimensions of the attribute vector exceeds the number of dimensions of the attribute vector corresponding to the image reproduction method, the number of dimensions is reduced. (Step S12).
As a result, the MPU 21 of the image reproduction device 10 refers to the external storage device 16 that cooperates and functions as the image reproduction method storage unit 36, and determines the image based on the value of the attribute vector of the image generated in step S11. A reproduction method is selected (step S13).

Specifically, the MPU 21 and the external storage device 16 work together as a color palette changing unit 41, an image effect applying unit 42, and a BGM adding unit 43 to select an image reproduction method as described above. Become.
As a result, the MPU 21 reproduces an image according to the selected image reproduction method and displays it on the display device 14 (step S14).
As described above, according to the present embodiment, by using an attribute vector associated with an image, an image reproduction method corresponding to the image content can be automatically selected, and the original image can be naturally selected. It is possible to perform reproduction with an image effect and BGM in a smooth flow, and the viewer can view the reproduced image without a sense of incongruity.

In the above description, among the attribute vectors assigned to the image, which vector is used for selection of the image reproduction method has been described in advance, but the user selects according to the purpose. It is also possible to configure as described above.
For example, in the above description, the color palette is configured to be referred to based on words classified into emotions, but the user can also set the season and day and night as attribute vectors. .
In this case, the saturation is higher as it approaches the outer periphery of the circle shown in FIG. 3, and the saturation is lower as it is closer to the inner side, so that the base point of the shooting date, that is, θ = 0 is green. Then, when the shooting time is t, (cos θ, sin θ) may correspond to the circumferential coordinate, and (sin (φ) +1) / 2 may correspond to the radial coordinate. As a result, it is possible to determine colors corresponding to changes in the season and changes in the day, and to perform image reproduction (for example, a slide show) in which the correspondence with the season is easy to understand.
In the above description, the attribute vector is assigned to the image in association with the feature of the image. However, the coordinates (spatial coordinates) of the end point of the attribute vector are used instead of the attribute vector in association with the feature of the image. It is also possible to adopt a configuration for adding to an image.
In the above description, the shooting location as the feature amount has been described as being given to the image in advance as metadata. However, information about the shooting location and the vicinity of the shooting location extracted from the map database after shooting is used. It is also possible to configure.

In the above description, the image reproducing apparatus 10 is specifically applicable to a personal computer, a PDA, a mobile phone, etc., but a digital still camera, a digital video camera, a scanner, a projector, a television, a printer, and others. It is possible to mount or connect to an electronic device that performs the image reproduction process.
In the above description, the case where the control program for realizing the above functions is stored in the ROM 22 in advance has been described. However, the control program is recorded on a computer (CPU) readable recording medium. Also good. With such a configuration, when the program is read from the storage medium by the computer and the computer executes processing according to the read program, the same operations and effects as those in the above embodiments can be obtained.
Here, the storage medium is a semiconductor storage medium such as RAM or ROM, a magnetic storage type storage medium such as FD or HD, an optical reading type storage medium such as CD, CDV, LD, or DVD, or a magnetic storage type such as MO. / Optical reading type storage medium, and any storage medium can be used as long as it can be read by a computer regardless of electronic, magnetic, optical, etc. .
It is also possible to configure such that a control program is downloaded, installed and executed via a communication network such as the Internet or a LAN and the communication interface unit 24.

1 is a schematic configuration block diagram of an image reproduction apparatus according to an embodiment. It is a functional block diagram of an image reproducing device. It is explanatory drawing of an example in the case of assigning the color used for the background or image frame of an image as an image reproduction method in two dimensions. It is explanatory drawing at the time of allocating the slide-in and strip-in (cross fade) which are the image effects of image switching to three-dimensional space. It is explanatory drawing at the time of allocating BGM to two-dimensional space. It is an outline processing flowchart of an embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Image reproduction apparatus, 11 ... Image reproduction apparatus main body, 11 ... Afternoon, 12 ... Image input apparatus, 13 ... Input apparatus, 14 ... Display apparatus, 15 ... Output apparatus, 16 ... External storage device, 21 ... MPU, 22 ... ROM, 23 ... RAM, 24 ... communication interface unit, 31 ... video / image input unit, 32 ... text extraction unit, 33 ... morpheme analysis unit, 34 ... importance calculation unit, 35 ... dimension reduction unit, 36 ... image reproduction method Storage unit 41... Color palette changing unit 42. Image effect applying unit 43.

Claims (9)

An image reproduction apparatus for reproducing a plurality of reproduction target images including a moving image, a still image, or a combination thereof,
A feature amount extraction unit that extracts the feature amount of each image included in the reproduction target image;
Assuming an n-dimensional image feature amount space represented by n types (n is an integer of 2 or more), the feature amount of each image extracted by the feature amount extraction unit is the n-dimensional image feature. A reproduction control unit configured to perform reproduction control of the plurality of reproduction target images based on the attribute vector or the spatial coordinate in the quantity space, and based on the attribute vector or the spatial coordinate;
An image reproducing apparatus comprising: The image reproducing apparatus according to claim 1,
A music vector space storage unit for storing music vector space information assigned to a vector space having at least a season and a time as a space axis based on the characteristics of the music data, and a plurality of music data,
As the feature amount, including information on the shooting date and time when the image was shot,
The playback control unit refers to the music vector space information, selects the music data based on the season and time to which the shooting date belongs, and assigns the selected music data as BGM of the image. An image reproducing apparatus characterized by that. The image reproduction apparatus according to claim 1 or 2,
An image reproduction apparatus characterized in that a TF / IDF value of an important word contained in a document related to the image is used as the feature amount. The image reproduction apparatus according to any one of claims 1 to 3,
An image reproducing apparatus using metadata previously given to the image as the feature amount. The image reproduction apparatus according to claim 4, wherein
The image reproduction apparatus, wherein the metadata includes at least one of a shooting date and a shooting location. The image reproduction apparatus according to any one of claims 1 to 5,
An image reproduction apparatus characterized in that information relating to a shooting location extracted from a map database and the vicinity of the shooting location is used as the feature amount. The image reproducing apparatus according to claim 2, wherein
An image reproducing apparatus comprising: a dimension reduction unit that reduces the dimension of the attribute vector or the spatial coordinate when the dimension of the attribute vector or the spatial coordinate corresponding to the image exceeds n dimensions. A control method for an image playback apparatus for playing back a plurality of playback target images including moving images, still images, or a combination thereof,
A feature amount extraction process for extracting each feature amount of each image included in the reproduction target image;
Assuming an n-dimensional image feature amount space represented by n types (n is an integer of 2 or more), the feature amount of each image extracted by the feature amount extraction unit is the n-dimensional image feature. A reproduction control process for performing reproduction control of the plurality of reproduction target images based on the attribute vector or the spatial coordinates in the quantity space, and based on the attribute vector or the spatial coordinates;
An image reproducing apparatus control method comprising: A control program for controlling, by a computer, an image playback device that plays back a plurality of playback target images including moving images, still images, or combinations thereof,
Extracting each feature amount of each image included in the reproduction target image;
Assuming an n-dimensional image feature amount space represented by n types (n is an integer of 2 or more), the extracted feature values of the images are attribute vectors in the n-dimensional image feature amount space. Alternatively, spatial coordinates are used, and reproduction control of the plurality of reproduction target images is performed based on the attribute vector or the spatial coordinates.
A control program characterized by that.

Images