JP2004336606A - Caption production system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a caption production system which reduces the burden of write starting work and produces caption data from input data including non-speech information. <P>SOLUTION: A production unit extracting part 11 in collaboration with a morpheme analyzing part 40 divides text included in the input data 100 to generate processing unit caption data 110. A display unit captioning part 12 divides the text included in the processing unit caption data 110 to generate display unit caption data 120. In such a case, newly needed additional timing information is generated by a proportional distribution method and added to a page including spare timing information. As to a page including no spare timing information, a display unit caption is notified to a synchronization detecting part 30 to make a time code at the time of a start/end to be outputted. A timing attaching part 13 generates additional timing information on the basis of the time code from the synchronization detecting part 30, adds the additional timing information to the display unit caption data, and outputs it as caption data for broadcasting. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a caption production system, and more particularly, to a caption production system capable of efficiently producing caption data including non-speech information.
[0002]
[Prior art]
In order to spread subtitle broadcasting, the Ministry of Posts and Telecommunications (at that time) has issued a guideline to provide subtitles to all programs to which subtitles can be added by 2007.
[0003]
However, while subtitle broadcasting is currently being implemented in Japan at a relatively high rate of 73.4% for NHK's general television, the average among the five commercial broadcasters in Tokyo is only 16.1%. This implementation rate is remarkably lower than that in Europe and the United States (for example, 70% in the United States), and there is a strong demand for expansion of subtitle broadcasting from various fields including subtitle broadcasting users.
[0004]
The biggest reason why the implementation rate of subtitle broadcasting does not improve is that many parts in the production of subtitle programs are performed manually. That is, creation of a subtitle program requires a lot of manpower, time, and cost. In particular, the creation of caption data including the transcript of speech greatly depends on the knowledge and ability of the worker, and therefore, it is extremely difficult to automate and make it more efficient.
[0005]
More specifically, in the transcript operation, a video tape on which a television material program is recorded is reproduced by a VTR, and sounds to be displayed as subtitles from the reproduced sounds (normal human voice, hereinafter referred to as speech). This is done by selectively listening to At that time, the start and end timings of the speech are specified based on the time code added to the reproduced video, and timing information indicating the timing is added to the written text data. In order to perform such operations, complex VTR operations must be performed, such as cueing the speech portion, adjusting the playback speed to facilitate transcription, repeating playback, or confirming the end point of the speech portion. No. Therefore, those who perform the transcript work are required to have not only high listening ability but also abundant vocabulary and high comprehension, and dexterity that can perform complicated VTR operations at the same time.
[0006]
From the above, in order to improve the production efficiency of a subtitle program, it is essential to reduce the burden on the person who performs the transcript work.
[0007]
2. Description of the Related Art As a conventional caption creation system that can reduce the burden of the transcripting operation, there is a system that automatically adds timing information to text data in which a speech is transcribed (for example, see Patent Document 1). ).
[0008]
In this subtitle production system, when digitized subtitle text data is provided, a text (character string) indicated by the data is a character string suitable for screen display, that is, a unit subtitle text which is a display unit on a television screen. , A line feed and a page break are automatically performed. On the other hand, a videotape containing a TV program that is the material of the subtitle text data is played back, and the timing at which the divided subtitle text should be displayed is determined using the timing information added to the video tape. Obtain timing information to perform. Then, the obtained timing information is added to the unit subtitle sentence data to generate subtitle data for subtitle broadcasting.
[0009]
With this subtitle creation system, a transcription operator can obtain subtitle data to which timing information is appropriately added only by transcribing a speech to text data.
[0010]
In addition, although it is not a caption production system, the text included in the caption data obtained by receiving the caption broadcast is divided according to the specification from the user, and the timing information corresponding to each of the divided caption sentences is added, and the TV screen is added. There is also a system that can change the display format described above (for example, see Patent Document 2).
[0011]
If this system is used for writing and working, it is possible to add timing information appropriately (for example, for each speaker or for each sentence) without considering the character string displayed on the screen. After that, it can be changed to subtitle data that enables appropriate screen display.
[0012]
[Patent Document 1]
JP-A-2002-342311
[0013]
[Patent Document 2]
JP 2000-350117 A
[0014]
[Problems to be solved by the invention]
In ordinary television programs, various non-speech information such as introduction of characters and scene descriptions is provided by subtitles. The provision of such non-speech information should be provided also in teletext.
[0015]
In closed caption broadcasting, in a scene where a plurality of characters alternate or speak simultaneously (for example, in a dialogue or a discussion), not only the contents of the statement are displayed in text but also information for identifying the speaker. For example, the speaker name needs to be displayed in characters.
[0016]
However, in the conventional subtitle production system, since the timing information is obtained by synchronizing the unit subtitle sentence data with the audio reproduced from the material program, the subtitle text and the content of the reproduced audio are obtained. Must be matched, and if the non-speech information as described above exists, the timing information cannot be automatically given.
[0017]
Further, when a conventional system for changing the display format is used for transcript, the existence of non-speech information does not cause a problem, and the labor of input work can be slightly reduced. However, this method requires timing information, for example, for each sentence, and thus has a problem in that it requires a high ability for the operator, as in the case of the conventional writing operation.
[0018]
Therefore, an object of the present invention is to provide a caption production system with a simple configuration that can reduce the burden of transcript work and can produce caption data from input data including non-speech information. I do.
[0019]
[Means for Solving the Problems]
According to the present invention, based on input data including text corresponding to the audio of the material program and the reproduced audio and reproduction time code of the material program, a display obtained by line-feeding and page-breaking the text for subtitle display In a subtitle production system for automatically producing broadcast subtitle data including unit subtitle sentence data and timing information for determining the timing of displaying the display unit subtitle data for each page, preliminary timing information is added as the input data. Thus, it is possible to obtain a caption production system characterized in that data including both pages that have been added and pages to which no preliminary timing information has been added can be used.
[0020]
Further, according to the present invention, from input data including text corresponding to the audio of the material program, display unit subtitle sentence data obtained by line-feeding and paginating the text so as to become a display unit subtitle sentence is generated. Generating, based on the display unit subtitle sentence data, the generated audio and playback time code of the material program, timing information for determining a timing of displaying the display unit subtitle sentence on a screen, In a caption production system for automatically producing broadcast subtitle data including data and the timing information, in the input data, for a page to which preliminary timing information has been added, text included in the page is line-feeded / page-breaked. When new timing information becomes necessary due to the above, the new timing information is obtained by the proportional distribution method using the preliminary timing information. Information is generated, and for the page to which the preliminary timing information is not added among the input data, when the timing included in the text included in the page is changed to a new line or a new page, new timing information is required. A new subtitle production system is provided in which new timing information is generated by an automatic audio synchronization process for synchronizing the unit subtitle sentence data with the reproduction audio.
[0021]
Furthermore, according to the present invention, from input data including text corresponding to the audio of the material program, display unit subtitle sentence data obtained by line-feeding and paginating the text so as to be a display unit subtitle sentence is generated. Generating, based on the display unit subtitle sentence data, the generated audio and playback time code of the material program, timing information for determining a timing of displaying the display unit subtitle sentence on a screen, In the caption production method for automatically producing broadcast subtitle data including data and the timing information, for a page of the input data to which preliminary timing information is added, the text included in the page is displayed in the display unit subtitle. After line breaks and page breaks to form sentences, new timing information is added as necessary by the proportional distribution method using the preliminary timing information. In the input data, for the page to which the preliminary timing information is not added, the text included in the page is line-feeded / broken down to become the display unit subtitle sentence, and then the display unit subtitle sentence data A new subtitle production method is provided in which new timing information is generated as necessary by an automatic audio synchronization process for time synchronization between the subtitles and the reproduced audio.
[0022]
Still further, according to the present invention, display unit subtitle sentence data obtained by inputting line breaks and page breaks from input data including text corresponding to audio of a material program to a computer so that the text becomes a display unit subtitle sentence And generating, based on the display unit subtitle sentence data and the generated audio and playback time code of the material program, timing information indicating a timing at which the display unit subtitle sentence is displayed on a screen, In a caption production program for automatically producing broadcast subtitle data including unit subtitle sentence data and the timing information, among the input data, for a page to which preliminary timing information is added, the text included in the page is converted to the text. After line breaks and page breaks to become display unit subtitles, necessary by the proportional distribution method using the preliminary timing information New timing information is generated in response to the input data, and for the page to which the preliminary timing information is not added, after the text included in the page is subjected to line feed / line feed to become the display unit subtitle sentence, A subtitle producing program for generating new timing information as needed by an automatic audio synchronization process for synchronizing the display unit subtitle sentence data with the reproduced audio.
[0023]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0024]
FIG. 1 shows a configuration of a caption production system according to an embodiment of the present invention. The illustrated subtitle production system includes a subtitle production control unit 10 that receives input data (digitized manuscript) 100 produced based on a material program and produces subtitle data for broadcast, and a subtitle production control unit 10 that is recorded on a recording medium such as an optical disk. A recording / reproducing unit 20 for reproducing a material program, a synchronization detecting unit 30 for synchronizing a display unit subtitle sentence extracted from input data under the control of the subtitle production control unit 10 and a reproduction audio from the recording / reproducing unit 20, A morphological analysis unit 40 that performs morphological analysis of text (character string) extracted from input data under the control of the production control unit 10, and a division rule for dividing (line feed / page break) the text extracted from the input data And a monitoring monitor 60 that displays both the subtitles based on the broadcast subtitle data produced by the subtitle production control unit 10 and the material programs reproduced by the recording and reproduction unit 20. It has.
[0025]
The subtitle production control unit 10 is realized by a personal computer, for example, and divides the input data 100 into processing unit subtitle data 110, and divides the text included in the processing unit subtitle data 110 so as to be suitable for screen display. The display unit includes a display unit subtitle conversion unit 12 to be combined and a timing providing unit 13 that adds timing information to the display unit subtitle data 120 to generate broadcast subtitle data 130.
[0026]
In the present embodiment, it is assumed that the synchronization detection unit 30 and the morphological analysis unit 40 are each realized by a dedicated computer, but may be realized by one computer including the subtitle production control unit 10. is there.
[0027]
The input data 100 is obtained by converting a character string representing a sound included in a material program into electronic data (text conversion) using a not-shown document digitizing device (for example, a personal computer). The transcription is performed by using a disk recording / reproducing apparatus capable of performing a non-linear operation.
[0028]
The input data 100 is assumed to be text data and XML (extensible Markup Language) data. The text data includes data (speech information) representing a character string (a character string mixed with kanji and kana) which transcribed the sound included in the material program and line feed information. The XML data is obtained by adding page information (page start / end information), timing information for specifying timing for displaying each page, and the like to the text data. In the case of XML data, the data may include data (non-speech information) representing a character string irrelevant to the audio of the material program, such as a speaker profile, and timing information for displaying the data. The input data 100 is provided to the caption production control unit 10 in a state recorded on a recording medium such as a magnetic disk or an optical disk.
[0029]
The production unit extraction unit 11 of the subtitle production control unit 10 reads the input data recorded on the recording medium for each line of text, and sends data representing the read one line of text to the morphological analysis unit 40.
[0030]
The morphological analysis unit 40 divides the text sent from the creation unit extraction unit 11 into morphemes, analyzes the morphemes, and specifies delimitable locations in the text. Then, the specified delimitable location information is added to the text and sent back to the production unit extraction unit 11. Note that the delimitable places are, for example, after a period, after a reading point, between phrases, or between morpheme parts of speech, and the order of priority is higher in this order.
[0031]
The production unit extraction unit 11 divides the text returned from the morphological analysis unit 40 into an appropriate length (about 70 to 90 characters) with reference to the delimitable location information added thereto, together with the delimitable location information. It is transmitted to the display unit subtitle conversion unit 12 as the processing unit subtitle data 110. However, pages of the input data 100 to which timing information (hereinafter referred to as “preliminary timing information”) has already been added are sent to the display unit subtitle conversion unit 12 as processing unit subtitle data 110 for each page.
[0032]
The display unit subtitle conversion unit 12 converts the text represented by the processing unit subtitle data 110 from the production unit extraction unit 11 into a format that is displayed on the screen, based on the division rule stored in the storage unit 50. By dividing and combining, the display unit subtitle data 120 is generated. That is, the display unit captioning unit 12 makes the number of lines and the number of characters of the caption displayed on the screen less than or equal to the maximum number of lines and the maximum number of characters per page specified by the division rule stored in the storage unit 50. As described above, the text represented by the processing unit subtitle data 110 is divided / combined (line feed / page break) into a display unit subtitle sentence, and the display unit subtitle data 120 representing the sentence is generated. At this time, when the processing unit subtitle data 120 includes preliminary timing information, the display unit subtitle conversion unit 12 inputs timing information (hereinafter, additional timing information) newly required by dividing the text. Is generated and added by the proportional distribution method using the preliminary timing information included in the processed subtitle data.
[0033]
Further, the display unit captioning unit 12 notifies the synchronization detection system 30 of the display unit caption text generated as described above. However, the display unit captioning unit 12 does not notify the synchronization detection system 30 of the display unit caption text generated based on the processing unit caption data 110 including the preliminary timing information.
[0034]
To the synchronization detection system 30, in addition to the data representing the display unit caption text from the display unit captioning unit 12, the playback audio and playback time code of the material program played back by the recording and playback unit 20 are input. Here, it is assumed that the recording / reproducing unit 20 starts reproducing the material program with a predetermined time delay from the start of reading by the production unit extracting unit 11.
[0035]
The synchronization detection system 30 performs a speech recognition process on the reproduced sound from the recording / reproducing unit 20, and compares it with the display unit subtitle sentence from the display unit subtitle unit 12. Then, time codes respectively corresponding to the start time and the end time of the reproduced sound corresponding to each display unit sentence are obtained. The synchronization detection system 30 outputs the obtained time code to the timing providing unit 13.
[0036]
The timing providing unit 13 inserts the time code input from the synchronization detection system 30 into the display unit subtitle data 120 from the display unit subtitle unit 12 as timing information for displaying the corresponding display unit subtitle sentence, It is output to the monitoring monitor 60 and an external storage unit (not shown) as broadcast subtitle data 130.
[0037]
The video reproduced by the recording / reproducing unit 20 is input to the monitoring monitor 60 with a delay of a predetermined time, and the video is displayed with broadcast subtitle data from the timing imparting unit 13 superimposed thereon. By monitoring the display content of the monitoring monitor 60, problems such as the display timing of subtitles displayed based on the subtitle data for broadcast output from the subtitle production control unit 10 are found and used for subsequent correction and the like. Is done.
[0038]
The subtitle production system of FIG. 1 operates roughly as shown in FIG.
[0039]
That is, input data is fetched line by line (step S201), and processing unit subtitle data is created (step S202). Next, the processing unit subtitle text is divided according to the division rule to generate display unit subtitle text data (step S203). Then, when new timing information is required due to the division (Y in step S204), it is determined whether or not preliminary timing information is added to the page including the display unit subtitle text (step S205). If the preliminary timing information has been added, additional timing information is generated and added by a proportional distribution method using the preliminary timing information (step S206). If the preliminary timing information has not been added, the synchronization detection unit 30 performs synchronization detection to obtain a timing code (step S207). Then, additional timing information indicating the timing code is generated and added to the display unit subtitle data (step S208). Finally, the subtitle data for broadcasting obtained as described above is output to the outside (step S209).
[0040]
Hereinafter, the operation of the caption production system according to the present embodiment will be described in detail based on examples.
[0041]
[Example 1]
Consider the XML data shown in FIG. 3 as input data. In this XML data, preliminary timing information (IN_TIME = "" and OUT_TIME = "";"" in all pages indicates the hour, minute, second, and frame number, respectively, two digits from the front). Is given.
[0042]
When such input data is provided to the subtitle production control unit 10, the production unit extraction unit 11 adds delimitable place information to text included in each page in cooperation with the morphological analysis unit 40, and performs processing. A unit caption sentence is generated and output to the display unit captioning unit 12 as processing unit caption data 110.
[0043]
The display unit subtitle unit 12 divides the text of each page (line feed / page break) with reference to the division rule of the storage unit 50. In this embodiment, since the preliminary timing information is added to all the pages, additional timing information newly required by dividing the text is generated by the proportional distribution method using the preliminary timing information.
[0044]
In the case of the present embodiment, since the existing timing information is included in all the pages, the information regarding the display unit subtitle sentence from the display unit subtitle generation unit 12 to the synchronization detection unit 30 is not output. Therefore, no time code is output from the synchronization detection unit 30, and no new timing information is added by the timing addition unit 13. Therefore, the output from the display unit captioning unit 12 is output as it is as broadcast caption data via the timing providing unit 13. The subtitle data for broadcasting obtained at this time is as shown in FIG.
[0045]
Here, generation of the additional timing information in the display unit subtitle conversion unit 12 will be described.
[0046]
It is assumed that the input data (XML data) is as shown in FIG. 5 and a display unit subtitle sentence as shown in FIG. 6 is generated therefrom. In this case, the end timing of the second character string and the start timing of the third character string in FIG. 6 cannot be directly obtained from FIG. That is, the preliminary timing information included in the input data cannot be used as it is.
[0047]
Therefore, the display unit subtitle conversion unit 12 calculates the end timing of “forgot in the Legal Affairs Bureau” by calculation. The structure of this character string is 7 kana characters and 5 kanji characters, and the number of statistical readings is 16.3. From FIG. 5, the time (reading speed) corresponding to the number of readings for this character string is 0.12, so the required time for this character string is 16.3 × 0.12 = 1.89. Therefore, the end timing of the second character string in FIG. 6 is 29.04 obtained by adding the required time 1.89 to 27.15 which is the start timing of the character string. The start timing of the third character string in FIG. 6 may coincide with the end timing of the second character string.
[0048]
As described above, the subtitle creation system according to the present embodiment can automatically generate broadcast subtitle data suitable for screen display from input data in which preliminary timing information is added to all pages. In addition, the apportioning method can generate additional timing information required in a significantly shorter time than the voice synchronization processing. Therefore, when the preliminary timing information is added to all pages as described above, Thus, the subtitle data for broadcasting can be obtained in about 1/30 of the time required for broadcasting (reproduction) of the material program. In addition, the difference between the display of the caption based on the timing information obtained by the apportionment method and the corresponding sound is less than 1 second, and is considered to be within the allowable range.
[0049]
[Example 2]
The subtitle production system according to the present embodiment can use existing broadcast subtitle data as input data. That is, this subtitle production system can be used even when the display format of the existing broadcast subtitle data is changed.
[0050]
For example, since the aspect ratio of the screen (image) differs between the standard broadcast and the high-definition broadcast, when the subtitle data created for the standard broadcast is used for the high-definition broadcast, the caption 1 You may want to change the number of characters per line. In such a case, the subtitle production system according to the present embodiment can be used.
[0051]
As a specific example, as shown in FIG. 7, there are three pages of standard broadcast subtitle data for displaying subtitles in 15 lines × 2 lines, which are displayed in high definition for displaying subtitles in 22 lines × 2 lines. When it is desired to change to broadcast subtitle data, the subtitle production system according to the present embodiment operates in the same manner as in Example 1, and outputs output data as shown in FIG. In the case of the present embodiment, all the additional timing information required by the subtitle reshaping can be obtained by the proportional method.
[0052]
As described above, the subtitle production system according to the present embodiment can also be used for reshaping broadcast subtitle data that has already been produced.
[0053]
[Example 3]
The caption production system according to the present embodiment can cope with input data in which pages with preliminary timing information and pages without preliminary timing information are mixed.
[0054]
For example, consider a case where a caption display as shown in FIG. 9 is performed. In the text shown in FIG. 9, the italic character strings in the second to sixth lines are non-speech information, and the rest are speech information. In this case, for the non-speech information, the transcriber manually adds the timing information. As a result, the obtained XML data is as shown in FIG.
[0055]
When the input data shown in FIG. 10 is given, the caption production system according to the present embodiment supplies the additional timing information necessary for the page having the preliminary timing information by the proportional distribution method in the same manner as in Examples 1 and 2. obtain. On the other hand, for a page that does not have the preliminary timing information, the synchronization detection unit 30 performs automatic audio synchronization detection, and acquires a time code that matches the start timing and the display end timing of the playback audio corresponding to each display unit subtitle. , And generates additional timing information based on the acquired time code. Thus, broadcast subtitle data as shown in FIG. 11 can be obtained.
[0056]
As described above, in the caption production system according to the present embodiment, additional timing information is generated without performing automatic audio synchronization processing for a page of input data that includes preliminary timing information. For pages of the input data that do not include the preliminary timing information, additional timing information is generated by automatic audio synchronization processing. Therefore, by not adding preliminary timing information to speech information and adding preliminary timing information to non-speech information, timing information can be applied to input data in which speech information and non-speech information are mixed. Can be automatically provided.
[0057]
Moreover, the generation of the additional timing by the apportionment method requires much less processing time than the automatic voice synchronization processing, so the time required to automatically add timing information to input data that does not include non-speech information is reduced. Automatically adding timing information to input data in which speech information and non-speech information coexist in almost the same processing time as described above.
[0058]
[Example 4]
When timing information is automatically added to input data in which a page having existing timing information and a page having no existing timing information are mixed, additional timing information added to text of a page having no preliminary timing information does not match the preliminary timing information. Cases can happen. This occurs, for example, when the synchronization detection unit 30 recognizes noise included in the reproduced sound as speech.
[0059]
Here, consider the text data shown in FIG. The italic character string in the second line is non-speech information, and the transcriber manually gives preliminary timing information to this non-speech information. Then, when such input data is input, it is assumed that the result shown in FIG. 13 is obtained as a result of adding the additional timing information based on the time code obtained from the synchronization detecting unit 30. In this case, the second page must be displayed while the first page is being displayed, and it is clear that any timing information is inappropriate.
[0060]
In order to avoid such a situation, the timing providing unit 13 determines whether or not the automatically provided additional timing information matches the preliminary timing information. If the display times of the consecutive pages overlap each other, it is determined whether the automatically added additional timing information can be changed to adjust the display times of the two pages so that they do not overlap. That is, when the automatically added timing information is changed so as not to overlap the display time, it is determined whether or not an allowable display time remains as a time for displaying the page. In the case of FIG. 13, the display time of the first page is 21 seconds, and there is a remaining 11 seconds after subtracting 10 seconds which is the display time of the second page. Therefore, it can be determined that there will be no problem even if the display time of the first page is reduced. As a result, the timing providing unit 13 determines that the display end time of the first page is incorrect, and matches the display end time with the display start time of the second page. Thus, the timing providing unit 13 automatically corrects the timing information and outputs the broadcast subtitle data shown in FIG.
[0061]
In addition, when the timing imparting unit 13 determines that it is not appropriate to perform the adjustment so as to eliminate the overlapping portion, it does not automatically correct the timing information. In this case, the correction is manually made later.
[0062]
As described above, the present invention has been described in accordance with one embodiment, but the present invention is not limited to the above embodiment.
[0063]
For example, in the above embodiment, it is determined whether or not the display unit subtitle sentence is transmitted to the synchronization detection unit 30 in page units based on the presence or absence of the preliminary timing information. When it is desired to display speech information and non-speech information on the same page as in the case of speech, for example, only a character string sandwiched between predetermined symbols may or may not be transmitted to the synchronization detection unit 30. Specifically, only the character string enclosed by "" (angle brackets) is transmitted to the synchronization detection unit 30, or the character string enclosed by "<>" is transmitted to the synchronization detection unit 30. You should not do it.
[0064]
【The invention's effect】
According to the present invention, of the input data, the page to which the preliminary timing information is added is divided by the proportional distribution method using the preliminary timing information, and the page to which the preliminary timing information is not added is converted to the text by the voice synchronization processing method. By obtaining new timing information required by line feed and page break, subtitle data can be automatically produced from text data including non-speech information. As a result, it is possible to reduce the burden of the subtitle creation operation.
[Brief description of the drawings]
FIG. 1 is a schematic diagram showing a configuration of a subtitle production system according to an embodiment of the present invention.
FIG. 2 is a flowchart for explaining an operation of a caption production control unit in the caption production system of FIG. 1;
FIG. 3 is a diagram illustrating an example of input data input to the caption production system of FIG. 1;
FIG. 4 is a diagram showing broadcast subtitle data produced from the input data of FIG. 3;
5 is a diagram showing a text included in other input data input to the caption production system of FIG. 1 and related information related thereto.
6 is a diagram showing a display unit subtitle sentence obtained from text included in other input data in FIG. 5 and its start / end time.
FIG. 7 is a diagram showing still another example of input data input to the caption production system of FIG. 1;
8 is a diagram showing subtitle data for broadcast produced from the input data of FIG. 7;
FIG. 9 is an example of a text including speech information and non-speech information.
FIG. 10 is a diagram showing XML data generated from the text of FIG. 9;
11 is a diagram showing broadcast subtitle data obtained by inputting the XML data of FIG. 10 to the subtitle production system of FIG. 1;
FIG. 12 is another example of a text including speech information and non-speech information.
FIG. 13 is a diagram showing XML data in a state where the manually input preliminary timing information and the automatically added additional timing information do not match.
FIG. 14 is a diagram showing XML data in a state where preliminary timing information manually input and additional timing information automatically added match.
[Explanation of symbols]
10 Subtitle production control unit
11 Production unit extractor
12 Display unit subtitle conversion unit
13 Timing giving section
20 Recording and playback unit
30 Synchronization detection system
40 Morphological analyzer
50 storage unit
60 Monitoring monitor
100 digitized manuscript (input data)
110 processing unit subtitle data
120 display unit subtitle data
130 Subtitle data for broadcasting

Claims (4)

Based on the input data including the text corresponding to the audio of the material program and the reproduction audio and reproduction time code of the material program, the display unit subtitles and the page breaks of the text for subtitle display are displayed. In a subtitle production system for automatically producing broadcast subtitle data including display unit subtitle data and timing information for determining the timing of displaying each page,
A subtitle production system, wherein data including both a page to which preliminary timing information is added and a page to which preliminary timing information is not added can be used as the input data. From input data including text corresponding to the sound of the material program, display unit subtitle sentence data obtained by line-feeding and page-breaking the text so as to become a display unit subtitle sentence is generated, and the display unit subtitle sentence data and Generating, based on the generated audio and the reproduction time code of the material program, timing information for determining a timing at which the display unit subtitle sentence is displayed on a screen, including the display unit subtitle sentence data and the timing information In a subtitle production system that automatically produces broadcast subtitle data,
Of the input data, for a page to which preliminary timing information has been added, when new timing information is needed due to line breaks and page breaks of the text included in the page, the distribution using the preliminary timing information is performed. Generate the new timing information by the method,
Of the input data, for a page to which preliminary timing information has not been added, when new timing information is required due to line feed / page break of the text included in the page, the display unit subtitle sentence data and New timing information was generated by an automatic audio synchronization process for time synchronization with the playback audio,
Caption production system characterized by the following. From input data including text corresponding to the sound of the material program, display unit subtitle sentence data obtained by line-feeding and page-breaking the text so as to become a display unit subtitle sentence is generated, and the display unit subtitle sentence data and Generating, based on the generated audio and the reproduction time code of the material program, timing information for determining a timing at which the display unit subtitle sentence is displayed on a screen, including the display unit subtitle sentence data and the timing information In a subtitle production method for automatically producing broadcast subtitle data,
Of the input data, for a page to which preliminary timing information has been added, after the text included in the page is line-feeded and page-breaked so as to become the display unit subtitle sentence, by a proportional distribution method using the preliminary timing information. Generate new timing information as needed,
In the input data, for a page to which the preliminary timing information is not added, the text included in the page is subjected to line feed / page break so as to become the display unit subtitle sentence, and then the display unit subtitle sentence data and the reproduction New timing information is generated as needed by automatic audio synchronization processing that takes time synchronization with audio,
A subtitle production method characterized by the following. A computer generates input unit subtitle sentence data obtained by line-feeding and page-breaking the text so as to become a display unit subtitle sentence from input data including text corresponding to the audio of the material program, and Based on the sentence data and the generated audio and reproduction time code of the material program, the display unit generates timing information indicating the timing at which the display unit subtitle sentence is displayed on the screen, and the display unit subtitle sentence data and the timing information In a subtitle production program that automatically produces broadcast subtitle data including
Of the input data, for a page to which preliminary timing information has been added, after the text included in the page is line-feeded and page-breaked so as to become the display unit subtitle sentence, by a proportional distribution method using the preliminary timing information. Generate new timing information as needed,
In the input data, for a page to which the preliminary timing information is not added, the text included in the page is subjected to line feed / page break so as to become the display unit subtitle sentence, and then the display unit subtitle sentence data and the reproduction New timing information is generated as needed by automatic audio synchronization processing that takes time synchronization with audio,
Caption production program characterized by the following.

Images