JP6164445B2 - Chapter setting device - Google Patents

Description

  The present invention relates to a chapter setting device that sets chapters in video data.

  Conventionally, as shown in Patent Document 1, caption data and video / audio data are extracted from a television signal, and the time when the caption data and the video / audio data corresponding to a preset keyword appear is specified as a chapter setting time. However, there is one that adds chapter information to video / audio data for writing.

  Further, as shown in Patent Document 2, a character registration unit in which a predetermined character is registered and an extraction unit that extracts a keyword from broadcast data are provided, and a predetermined character is extracted from the keyword extracted by the extraction unit. Is included, the broadcast data is divided into chapters at the time when the keyword is included.

  However, in the case where chapters are set only based on such preset keywords, for example, when the related keyword appears instead of the set keyword itself, the portion in which the user is interested is used. Even if it exists, there exists a problem that the chapter divided | segmented there cannot be set.

  In addition, in the case of setting the chapter start time based on a preset keyword, when the preset keyword appears at a long time interval, there are many very redundant chapters, that is, scenes of low user interest. There is a problem that the chapter setting is included.

Japanese Patent No. 4019085 JP 2011-223205 A

  Therefore, the present invention efficiently and accurately detects a scene with high user interest and sets a chapter, and efficiently and accurately detects a scene with low user interest and sets the end position of the chapter. The main objective is to set chapters that include high scenes and exclude scenes of low user interest.

That is, the chapter setting device according to the present invention is a chapter setting device that sets chapters in video data, and is a first for continuing to play back video from the time when the user keyword appears for the user keyword set by the user. Keyword information data including first effective time data indicating effective time, and keyword including second effective time data indicating second effective time for continuing to reproduce video from a point in time when a related keyword related to the user keyword appears. A keyword information data storage unit for storing information data; and the video data is analyzed, and a position where the user keyword or the related keyword appears after the first effective time and the second effective time has elapsed is set as a chapter start position And after the chapter start position After a lapse of one valid time and the second valid time, characterized in that it comprises, a chapter setting unit for setting a position where the user keyword and non-target keyword other than the related keyword appeared as a chapter end position.

If it is such, since the position where the user keyword set by the user or the related keyword related to the user keyword appears is set as the chapter start position, a scene with high user interest is detected efficiently and accurately. Chapters can be set. Further, since the position where the non-target keyword other than the user keyword and the related keyword appears is set as the chapter end position, the chapter end position is set by efficiently and accurately detecting the scene of low user interest. It is possible to set chapters including scenes of high interest and excluding scenes of low user interest. Further, the first effective time for continuing to play the video from the time when the user keyword appears and the second effective time for continuing to play the video from the time when the related keyword appears are set separately. Therefore, it is possible to set chapters according to the degree of interest of the user.

A chapter setting device for setting chapters in video data, wherein first valid time data indicating a first valid time for continuing to play back a video from a point in time when the user keyword appears for a user keyword set by a user Keyword information data storage unit for storing keyword information data including keyword information data including the second valid time data indicating the second valid time for continuing to reproduce the video from the time point when the related keyword related to the user keyword appears And analyzing the video data and setting a position where a non-target keyword other than the user keyword and the related keyword appears as the chapter start position after the first effective time and the second effective time have elapsed, wherein even after from the chapter start position user keyword It is desirable; and a chapter setting unit for setting a position where the related keyword appeared as a chapter end position.

In such a case, since the position where the non-target keyword other than the user keyword and the related keyword appears is set as the chapter start position, a chapter consisting of a scene with low user interest is set, and the video data A scene with low user interest can be clarified. In addition, since the position where the user keyword or the related keyword appears is set as the chapter end position, it is possible to efficiently and accurately detect a scene with high user interest and set the next chapter.

  As a specific configuration of the keyword information data storage unit, the keyword information data storage unit displays keyword data indicating a keyword that can be the user keyword, the related keyword, or the non-target keyword, and a video from the time when the keyword appears. It is desirable to store keyword information data including effective time data indicating effective time for continuing reproduction.

  As a specific configuration for setting the user keyword, the related keyword, and the non-target keyword, a keyword setting unit that sets the user keyword, the related keyword, and the non-target keyword is provided, and the keyword setting unit includes: The keyword selected by the user from the keywords indicated by the keyword data stored in the keyword information data storage unit is set as the user keyword, and the keyword indicated by the keyword data stored in the keyword information data storage unit A keyword related to the user keyword is extracted and set as the related keyword, and keywords other than the user keyword and the related keyword are previously selected from the keywords indicated by the keyword data stored in the keyword information data storage unit. It is desirable to set a non-target keyword.

  In order to accurately set the related keyword and the non-target keyword based on the user keyword selected by the user, in the keyword information data storage unit, the keyword data is classified for each category to which the keyword belongs. It is desirable to be stored.

  In order to enable the user to edit the keyword information data stored in the keyword information data storage unit and to detect a scene of high interest of the user efficiently and accurately, the keyword information data storage unit It is desirable that the stored keyword information data is configured to be editable by an input signal from the user.

  In order to enable the user to edit the valid time data among the keyword information data stored in the keyword information data storage unit and to set chapters including scenes of high user interest efficiently, the keyword information data Of these, it is desirable that at least the valid time data can be edited by an input signal from the user.

  In order to enable the user to newly store the keyword information data stored in the keyword information data storage unit and to detect a scene of high interest of the user efficiently and accurately, an input signal from the user Therefore, it is preferable that new keyword information data can be stored in the keyword information data storage unit.

  From the user's viewing history, keywords included in scenes of high user interest are generated, and the keyword information data in the keyword information data storage unit is enriched so that scenes of high user interest can be detected efficiently and accurately. In order to achieve this, a viewing history management unit is provided that acquires user viewing history information, generates keyword information data indicating a new keyword from the viewing history information, and stores the keyword information data in the keyword information data storage unit. Is desirable.

  Specifically, the user keyword, the related keyword, and the non-target keyword are compared with the appearance keyword data in the video data to set the chapter start position and the chapter end position. As a specific configuration for setting, a keyword is detected from at least one of image data, audio data, or caption data in the video data, and an appearance keyword data indicating the keyword and an appearance time of the keyword are set. An appearance keyword information data generation unit that generates appearance keyword information data including the appearance time data shown, and the chapter setting unit acquires the appearance keyword information data obtained by the appearance keyword information data generation unit, and the appearance keyword Appearance indicated by information data Keyword and the user keyword, the related keywords and are compared with the non-target keyword, it is desirable to set the chapter start position and the chapter end position.

  According to the present invention configured as described above, a scene in which the user is highly interested is efficiently and accurately set to set a chapter, and a scene in which the user is not interested in is efficiently and accurately detected to set the end position of the chapter. In addition, it is possible to set chapters including scenes with high user interest and excluding scenes with low user interest.

The system block diagram which shows the chapter setting apparatus in this embodiment. The figure which showed typically the chapter set to the program in this embodiment. The figure which showed typically the chapter set to the program in this embodiment. The flowchart figure of the chapter setting process in this embodiment. The figure which shows the hierarchical structure of the keyword information data storage part in this embodiment. The system block diagram which shows the chapter setting apparatus in deformation | transformation embodiment.

  An embodiment of the present invention will be described below with reference to the drawings.

  The chapter setting apparatus 100 according to the present embodiment sets chapters in the video data D1, and is a general purpose or dedicated computer including a CPU, a memory, an input / output interface, an AD converter, and the like. In accordance with a control program stored in a predetermined area, the CPU, peripheral devices, etc. cooperate to set chapters in the video data.

  Specifically, the chapter setting device 100 includes an appearance keyword information data generation unit 11, a keyword information data storage unit 12, a keyword setting unit 13, and a chapter setting unit 14, as shown in FIG.

  As shown in FIG. 1, the appearance keyword information data generation unit 11 includes an image detection unit 11A that detects a keyword from the image data of the video data D1, a voice detection unit 11B that detects a keyword from the audio data of the video data D1, And subtitle detecting means 11C for detecting a keyword from the subtitle data of the video data.

  And this appearance keyword information data generation part 11 acquires a keyword and its appearance time from the image data, audio | speech data, and caption data contained in the acquired video data D1, and contains the information of the said keyword and its appearance time Appearance keyword information data D2 is generated.

  The keyword information data storage unit 12 includes various keywords related to a person name or a related term matched to a genre, and effective time data indicating an effective time for continuing to play a video from the time when the keyword appears. The keyword information data D3 is stored in a hierarchy for each genre.

  The keyword information data D3 stored in the keyword information data storage unit 12 is configured to be editable by an input signal from the user. Specifically, it is configured such that a new keyword can be stored and the valid time included in the valid time data corresponding to each keyword can be edited.

  As shown in FIG. 1, the keyword setting unit 13 receives the keyword selected by the user from the input signal receiving unit 21 that has received the input signal from the user, and sets the keyword selected by the user as the user keyword K1. Set. Here, the input signal from the user may select one keyword information data D3 from the keyword information data D3 stored in the keyword information data storage unit 12, or the keyword information data D3 The user may directly input a new keyword that is not stored.

  Further, the keyword setting unit 13 extracts a keyword related to the user keyword K1 from the keywords indicated by the keyword information data D3 stored in the keyword information data storage unit 12, and sets the keyword as a related keyword K2.

  Further, the keyword setting unit 13 extracts keywords other than the user keyword K1 and the related keyword K2 from the keywords indicated by the keyword information data D3 stored in the keyword information data storage unit 12, and sets them as non-target keywords K3.

  Then, the keyword setting unit 13 includes user keyword data DK1 including user keyword K1 and first effective time data T1 indicating effective time for continuing to reproduce the video from the time when the user keyword K1 appears, the related keyword K2, Generating related keyword data DK2 including second effective time data T2 indicating effective time for continuing to reproduce the video from the time when the related keyword K2 appears, and non-target keyword data DK3 including the non-target keyword K3 It is.

  The chapter setting unit 14 includes the appearance keyword information data D2 generated by the appearance keyword information data generation unit 11, and the user keyword data DK1, the related keyword data DK2, and the non-target keyword data DK3 generated by the keyword setting unit 13. The keyword indicated by the appearance keyword information data D2 is compared with the keyword indicated by the user keyword data DK1, the keyword indicated by the related keyword data DK2, and the keyword indicated by the non-target keyword data DK3, and the chapter start position Cs and the chapter are determined. An end position Cf is set.

  Then, the chapter setting unit 14 transmits information on the set chapter start position Cs and chapter end position Cf to the recorded program data storage unit 22. The recorded program data storage unit 22 stores chapter information including the set chapter start position Cs and chapter end position Cf, and stores appearance keyword information data D2 generated by the appearance keyword information data generation unit 11. Is. Since the appearance keyword information data D2 is stored in the recorded program data storage unit 22 in this way, when the user changes the keyword, the appearance keyword information data D2 is referred to, so that a new keyword based on the changed keyword is obtained. It is possible to easily reflect the priority of the keyword by chapter setting or user viewing, and the accuracy is improved. In other words, once the appearance time of a keyword is recorded (at the time of reception or first playback), when the user keyword is changed, the chapters of the changed user keyword and its related keywords without being played or viewed are displayed. Setting is possible.

  A specific flow of setting chapters in chapter setting device 100 configured as described above will be described with reference to FIGS. 2, 3, and 4.

  The chapter setting device 100 includes a chapter Cx1 start determination step S1 for starting the first chapter Cx1 including a scene with high user interest at the time when the user keyword K1 or the related keyword K2 first appears, and a scene with high user interest. A chapter Cxn end determination step S2 for determining whether to end a chapter Cxn including a chapter Cxn, and a chapter Cyn end determination step S3 for determining whether to end a chapter Cyn composed only of scenes of low user interest.

  First, the chapter Cx1 start determination step S1 will be described. At the same time as the program E included in the video data D1 is started, the first chapter, Cy0, is started. This chapter Cy0 ends when it is determined that the user keyword K1 or the related keyword K2 has appeared in the chapter Cx1 start determination step S1 for determining whether the user keyword K1 or the related keyword K2 has appeared. At the same time as the end of chapter Cy0, chapter Cx1, which is the next chapter, is started.

  Here, since the chapter Cy0 is a chapter that does not include the user keyword K1 or the related keyword K2, it can be said that the chapter Cy0 is a chapter composed of scenes of low user interest.

  Next, the chapter Cxn end determination step S2 will be described. In this chapter Cxn end determination step S2, the chapter Cxn for which it is determined whether or not to end the chapter, for example, the chapter Cx1 started in the chapter Cx1 start determination step S1, is started by the appearance of the user keyword K1 or the related keyword K2. It can be said that it is a chapter including a scene with high user interest. The chapter Cxn end determination step S2 has a configuration in which the video is continuously reproduced without dividing the chapter Cx1 while the user keyword K1 or the related keyword K2 that is an element constituting the scene of high user interest appears. It has become.

  Specifically, the chapter Cxn end determination step S2 is after the expiration of the effective time in the first effective time data T1 of the user keyword K1 or the effective time in the second effective time data T2 of the related keyword K2, and The chapter Cx1 is terminated only when a non-target keyword appears. Here, when the chapter Cx1 is finished, the next chapter, the chapter Cy1, is started at the same time.

  Next, the chapter Cyn end determination step S3 will be described. In this chapter Cyn end determination step S3, the chapter Cyn that is a target for determining whether to end the chapter, for example, the chapter Cy1 started in the chapter Cxn end determination step S2, is a keyword other than the user keyword K1 or the related keyword K2. Since it is a chapter started by the appearance of a certain non-target keyword K3, it can be said that it is a chapter consisting of a scene with low user interest. This chapter Cyn end determination step S3 is configured to continue to play the video without dividing the chapter Cy1 until the user keyword K1 or the related keyword K2 which is an element constituting the scene of high user interest appears.

  Specifically, the chapter Cy1 end determination step S3 ends the chapter Cy1 when it is determined that the user keyword K1 or the related keyword K2 has appeared. Here, at the same time as the end of chapter Cy1, chapter Cx2, which is the next chapter, is started. At the same time as the chapter Cx2 is started, the process returns to the chapter Cxn end determination step S2. In this way, the chapter Cxn end determination step S2 and the chapter Cyn end determination step S3 are repeatedly performed until the program E ends.

  In the present embodiment, since the program E ends during creation of the chapter Cy4, the chapter end position Cf of the chapter Cy4 is set as the end position of the program E.

  Next, the hierarchical structure of the keyword information data storage unit 12 will be described with reference to FIG.

  First, the keyword information data storage unit 12 stores the keywords indicated by the keyword information data D3 in hierarchical order in the large category L1, the medium category L2, and the small category L3.

  The keywords classified into the large category L1 are keywords indicating a large range, for example, “sports”, “foreign film”, “Japanese film”, “drama”, and the like.

  The keywords classified in the middle category L2, which is the hierarchy below the large category L1, are included in more specific keywords related to the keywords classified in the large category L1, for example, “sports” in the large category L1. The middle category L2 includes “general”, “tennis”, “basketball”, “rugby”, and the like. The keywords of the medium category L2 classified in the medium category L2 included in the “foreign film” of the large category L1 are “general”, “action”, “SF”, and the like.

  As keywords classified in the small category L3, which is a layer below the middle category L2, keywords such as person names or related terms, for example, “tennis” in the middle category L2 included in “sports” in the large category L1 "A player", "B player", "Smash" or "Break". The keywords of the small category L3 classified as the “basketball” of the middle category L2 included in the “sports” of the large category L1 are “C player”, “D player”, “E player”, “dunk shot”. Or “break”.

  Valid time data is set for each of the keywords classified in the small category L3. The effective time indicated by the effective time data set for each keyword is set to a fixed time, for example, 3 seconds uniformly in the initial setting. Furthermore, since it is configured to be editable by an input signal from the user, for example, it is possible to lengthen the effective time of a keyword that is highly interested by the user.

  Moreover, priority is set to the keywords classified into the small category L3, and “N” is set by default. This priority is configured to change when setting the chapters described above. Specifically, when a certain keyword is set as the user keyword K1, “A” is set, and the keyword set as the related keyword K2 is “B”.

  Therefore, the keyword set with the priority “A” is the user keyword K1, the keyword set with the priority “B” is the related keyword K2, and the keyword that remains at the priority “N” is the non-target keyword K3. Become.

  Furthermore, since the effective time indicated by the effective time data of the keyword set with the priority “A” is extended, the user keyword K1 set with the priority “A” The effective time indicated by is extended to 5 seconds. Data indicating the effective time of the keyword set with the priority “A” is the first effective time data T1. Further, the data indicating the effective time of the keyword set with the priority “B” is the second effective time data T2.

  As shown in FIG. 5, when “dunk” in the small category L3 is set as the user keyword, this “dunk” is set as the priority “A” and is effective. The effective time indicated by the time data is 5 seconds. Further, the keyword of the other small category L3 included in the middle category “basketball” including “dunk shot” is set as the related keyword K2, that is, the priority “B”.

Of all the keywords, keywords other than “basketball” in the middle category L2 are all set as the priority “N”, that is, the non-target keyword K3.
<Effect of this embodiment>

According to the chapter setting device 100 configured as described above, in the chapter Cx1 start determination step S1 and the chapter Cyn end determination step S3, the position where the user keyword K1 or the related keyword K2 set by the user appears is set as the start position of the chapter Cxn. Since the setting is performed, it is possible to efficiently and accurately detect a scene with high user interest and set a chapter.

  Further, in the chapter Cxn end determination step S2, the position where the non-target keyword K3 other than the user keyword K1 and the related keyword K2 appears is set as the chapter Cyn end position, so that a scene with low user interest is efficiently and accurately detected. Thus, the end position of the chapter Cxn can be set, and the chapter Cxn can be set excluding the scene that includes the scene with high user interest and excludes the user with low interest.

  Further, the first effective time data T1 for continuing to reproduce the video from the time when the user keyword K1 appears and the second effective time data T2 for continuing to reproduce the video from the time when the related keyword K2 appears are separately provided. Since it is set, it is possible to set chapters according to the degree of interest of the user.

  In addition, for example, only a specific sports scene can be viewed on a sports program, or only news of a specific genre can be viewed on a news program, and conversely, a specific sport or player can be viewed. Become. In addition, in a movie or a drama, the appearance scene of a specific performer can be extracted by storing the related data between the performer and the role name in the keyword information data storage unit 12.

<Other modified embodiments>
The present invention is not limited to the above embodiment. For example, as shown in FIG. 6, user viewing history data is acquired, keyword information data D3 indicating a new keyword is generated from keywords included in the viewing history data, and stored in the keyword information data storage unit 12. The viewing history management unit 15 may be provided.

  The keyword valid time data stored in the keyword information data storage unit 12 can be edited by the viewing history management unit 15, and the viewing history management unit 15 is included in a scene having a high appearance frequency in the user viewing history data. The effective time data of the keyword may be extended. In this case, even if the user does not edit the valid time data himself, chapters including scenes of high interest of the user can be set efficiently.

  In addition, it goes without saying that the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention.

DESCRIPTION OF SYMBOLS 100 ... Chapter setting apparatus 11 ... Appearance keyword information data generation part 11A ... Image detection means 11B ... Audio | voice detection means 11C ... Subtitle detection means 12 ... Keyword information data storage part 13 ... -Keyword setting unit 14-Chapter setting unit 15-Viewing history management unit 21-Input signal receiving unit 22-Recorded program data storage unit Cs-Chapter start position Cf-Chapter end position D1 ... Video data D2 ... Appearance keyword information data D3 ... Keyword information data DK1 ... User keyword data DK2 ... Related keyword data DK3 ... Non-target keyword data L1 ... Large category L2・ ・ ・ Medium category L3 ・ ・ ・ Small category S1 ・ ・ ・ Chapter Cx1 start determination step S2 Chapter Cxn end decision step S3 ··· chapter Cyn end decision step E ··· program

Claims (12)

A chapter setting device for setting chapters in video data,
Keyword information data including first effective time data indicating the first effective time for continuing to reproduce video from the time when the user keyword appears for the user keyword set by the user, and related keywords related to the user keyword A keyword information data storage unit for storing keyword information data including second effective time data indicating a second effective time for continuing to reproduce the video from the point of time when
Analyzing the video data, setting the position where the user keyword or the related keyword appears after the first effective time and the second effective time as a chapter start position, and after the chapter start position, A chapter setting device comprising: a chapter setting unit configured to set, as a chapter end position, a position where a non-target keyword other than the user keyword and the related keyword appears after the first effective time and the second effective time have elapsed. A chapter setting device for setting chapters in video data,
Keyword information data including first effective time data indicating the first effective time for continuing to reproduce video from the time when the user keyword appears for the user keyword set by the user, and related keywords related to the user keyword A keyword information data storage unit for storing keyword information data including second effective time data indicating a second effective time for continuing to reproduce the video from the point of time when
The video data is analyzed, and a position where a non-target keyword other than the user keyword and the related keyword appears after the first effective time and the second effective time has elapsed is set as a chapter start position, and the chapter A chapter setting device comprising: a chapter setting unit that sets a position where the user keyword or the related keyword appears after the start position as a chapter end position.   The keyword information data storage unit includes keyword data indicating a keyword that can be the user keyword, the related keyword, or the non-target keyword, and effective time data indicating an effective time for continuing to play a video from the time when the keyword appears. The chapter setting device according to claim 1 or 2, wherein the keyword setting data is stored. A keyword setting unit for setting the user keyword, the related keyword, and the non-target keyword;
The keyword setting unit sets a keyword selected by the user from the keywords indicated by the keyword data stored in the keyword information data storage unit as the user keyword, and the keyword data stored in the keyword information data storage unit A keyword related to the user keyword is extracted from the keywords shown and set as the related keyword, and keywords other than the user keyword and the related keyword are selected from the keywords indicated by the keyword data stored in the keyword information data storage unit. The chapter setting device according to claim 3, wherein a keyword is set as the non-target keyword.   5. The chapter setting device according to claim 3, wherein the keyword data is classified and stored for each category to which the keyword belongs in the keyword information data storage unit.   6. The chapter setting device according to claim 3, wherein the keyword information data stored in the keyword information data storage unit is configured to be editable by an input signal from a user.   The chapter setting device according to claim 6, wherein at least valid time data among the keyword information data is configured to be editable by an input signal from a user.   The chapter setting device according to claim 3, wherein new keyword information data can be stored in the keyword information data storage unit in accordance with an input signal from a user.   9. A viewing history management unit that obtains user viewing history information, generates keyword information data indicating a new keyword from the viewing history information, and stores the keyword information data in the keyword information data storage unit. The chapter setting device according to the above. Appearance that detects a keyword from at least one of image data, audio data, or caption data in the video data and generates appearance keyword information data including appearance keyword data indicating the keyword and appearance time data indicating the appearance time of the keyword A keyword information data generation unit,
The chapter setting unit acquires the appearance keyword information data obtained by the appearance keyword information data generation unit, and compares the appearance keyword indicated by the appearance keyword information data with the user keyword, the related keyword, and the non-target keyword. The chapter setting device according to claim 1, wherein the chapter start position and the chapter end position are set. A chapter setting program for setting chapters in video data,
Keyword information data including first effective time data indicating the first effective time for continuing to reproduce video from the time when the user keyword appears for the user keyword set by the user, and related keywords related to the user keyword A keyword information data storage unit for storing keyword information data including second effective time data indicating a second effective time for continuing to reproduce the video from the point of time when
Analyzing the video data, setting the position where the user keyword or the related keyword appears after the first effective time and the second effective time as a chapter start position, and after the chapter start position, The computer is caused to exhibit a function as a chapter setting unit that sets a position where a non-target keyword other than the user keyword and the related keyword appears after the first effective time and the second effective time have elapsed as a chapter end position. Program. A chapter setting program for setting chapters in video data,
Keyword information data including first effective time data indicating the first effective time for continuing to reproduce video from the time when the user keyword appears for the user keyword set by the user, and related keywords related to the user keyword A keyword information data storage unit for storing keyword information data including second effective time data indicating a second effective time for continuing to reproduce the video from the point of time when
The video data is analyzed, and a position where a non-target keyword other than the user keyword and the related keyword appears after the first effective time and the second effective time has elapsed is set as a chapter start position, and the chapter A program that causes a computer to exhibit a function as a chapter setting unit that sets a position at which the user keyword or the related keyword appears after a start position as a chapter end position.