CN109635154B - A method for automatically generating Internet graphic manuscripts based on manuscripts and news programs - Google Patents

Abstract

Translated fromChinese

本发明公开了一种基于文稿和新闻节目自动生成互联网图文稿件的方法，涉及广电媒体技术领域，本发明包括S1、提取文稿内容：读取新闻文稿内容，提取文稿中待生成互联网图文稿件各模块所需内容写入到图文稿件模板中的相应位置；S2、计算文稿对应视频段：根据文稿分段情况，结合节目视频，计算每一文稿段对应的视频段；S3、选取视频素材：从每一视频段中选取关键帧或关键帧序列加工作为对应文稿段视频素材插入到文稿中；S4、生成互联网图文稿件，本发明能够利用节目文稿和节目视频快速生成互联网图文稿件，免除工作人员对文稿和视频进行选取、核对、拷贝的繁琐过程，大幅提升生产效率。

The invention discloses a method for automatically generating Internet graphics and text manuscripts based on manuscripts and news programs, and relates to the technical field of radio and television media. The invention includes S1. Extracting manuscript content: reading news manuscript content, extracting Internet graphics and text manuscripts to be generated in the manuscript The content required by each module is written into the corresponding position in the graphic manuscript template; S2, calculate the video segment corresponding to the manuscript: according to the segmentation of the manuscript, combined with the program video, calculate the video segment corresponding to each manuscript segment; S3, select the video material Select key frame or key frame sequence processing from each video segment and insert in the manuscript as corresponding manuscript segment video material; S4, generate Internet graphic manuscript, the present invention can utilize program manuscript and program video to generate Internet graphic manuscript quickly, It saves the staff from the tedious process of selecting, checking, and copying manuscripts and videos, greatly improving production efficiency.

Description

Translated fromChinese

一种基于文稿和新闻节目自动生成互联网图文稿件的方法A method for automatically generating Internet graphic manuscripts based on manuscripts and news programs

技术领域technical field

本发明涉及广电媒体技术领域，更具体的是涉及一种基于文稿和新闻节目自动生成互联网图文稿件的方法。The invention relates to the technical field of radio and television media, and more specifically relates to a method for automatically generating Internet graphic manuscripts based on manuscripts and news programs.

背景技术Background technique

随着融合媒体的推进，电视台也越来越重视互联网渠道的内容发布以达到更好的传播效果，因此，在制作新闻节目时，常常需要同时制作互联网图文稿件。With the advancement of converged media, TV stations are also paying more and more attention to the content release of Internet channels to achieve better communication effects. Therefore, when producing news programs, it is often necessary to produce Internet graphic manuscripts at the same time.

现有的互联网图文稿件制作方法主要是通过人工进行，具体是：首先在互联网稿件模块挑选稿件模板，再拷贝文稿内容到模板的导语和正文位置，或者重新创作编辑导语和正文，然后浏览视频内容，根据正文内容选择视频画面截图作为稿件插图来完成互联网图文稿件的制作。The existing method of making Internet graphic manuscripts is mainly carried out manually, specifically: first select the manuscript template in the Internet manuscript module, then copy the contents of the manuscript to the lead and text positions of the template, or recreate and edit the lead and text, and then browse the video Content, according to the content of the text, select the screenshot of the video screen as the illustration of the manuscript to complete the production of the Internet graphic manuscript.

文稿来源于文稿系统，是新闻节目制作的重要环节(相当于新闻大纲+详细内容)，决定了这条新闻怎么制作和表达，这条几分钟的新闻每一秒要播什么内容，需要哪些部门的配合，剪辑的时候要剪哪些内容，是否有采访画面，是否需要配音，是否需要字幕，直播时主持人要说些什么内容等等，所以，文稿除了标题、导语和正文外，还包括同期声、配音、口播等信息，这就使得上述人工浏览文稿进行互联网图文稿件的制作较为繁复，需要耗费大量人力和时间成本，效率较低。The manuscript comes from the manuscript system and is an important part of news program production (equivalent to news outline + detailed content), which determines how to produce and express this news, what content will be broadcast every second of this news of a few minutes, and which departments are needed cooperation, what content should be edited when editing, whether there is an interview screen, whether dubbing is required, whether subtitles are required, what the host will say during the live broadcast, etc. Therefore, in addition to the title, lead and text, the manuscript also includes Information such as audio, dubbing, oral broadcasting, etc., which makes the above-mentioned manual browsing of manuscripts to make Internet graphic manuscripts more complicated, requires a lot of manpower and time costs, and is less efficient.

发明内容Contents of the invention

本发明的目的在于：为了解决现有的互联网图文稿件的制作较为繁复，耗费大量人力和时间成本，效率较低的问题，本发明提供一种基于文稿和新闻节目自动生成互联网图文稿件的方法，能够利用节目文稿和节目视频快速生成互联网图文稿件，免除工作人员对文稿和视频进行选取、核对、拷贝的繁琐过程，大幅提升生产效率。The purpose of the present invention is: in order to solve the problem that the production of existing Internet graphic manuscripts is relatively complicated, consumes a lot of manpower and time costs, and has low efficiency, the present invention provides a method for automatically generating Internet graphic manuscripts based on manuscripts and news programs The method can use the program manuscript and program video to quickly generate Internet graphic manuscripts, eliminating the cumbersome process of selecting, checking, and copying the manuscripts and videos for staff, and greatly improving production efficiency.

本发明为了实现上述目的具体采用以下技术方案：The present invention specifically adopts the following technical solutions in order to achieve the above object:

一种基于文稿和新闻节目自动生成互联网图文稿件的方法，其特征在于，包括如下步骤：A method for automatically generating Internet graphics and text manuscripts based on manuscripts and news programs, characterized in that it comprises the following steps:

S1、提取文稿内容：读取新闻文稿内容，提取文稿中待生成互联网图文稿件各模块所需内容写入到图文稿件模板中的相应位置；S1. Extract the content of the manuscript: read the content of the news manuscript, extract the required content of each module of the Internet graphic manuscript to be generated in the manuscript, and write it into the corresponding position in the graphic manuscript template;

S2、计算文稿对应视频段：根据文稿分段情况，结合节目视频，计算每一文稿段对应的视频段；S2. Calculate the video segment corresponding to the manuscript: according to the segmentation of the manuscript, combined with the program video, calculate the video segment corresponding to each manuscript segment;

S3、选取视频素材：从每一视频段中选取关键帧或关键帧序列加工作为对应文稿段视频素材插入到文稿中；S3. Select video material: select key frame or key frame sequence processing from each video segment and insert it into the document as the corresponding document segment video material;

S4、生成互联网图文稿件。S4. Generate Internet graphic manuscripts.

进一步的，所述S1中根据待生成互联网图文稿件各模块所需内容相对应的标识符对文稿内容进行识别和选取，。Further, in said S1, the content of the document is identified and selected according to the identifier corresponding to the required content of each module of the Internet graphic document to be generated.

进一步的，所述S2中计算文稿对应视频段的方法，包括但不限于语音识别、机器视觉等智能识别技术。Further, the method for calculating the corresponding video segment of the document in S2 includes but is not limited to intelligent recognition technologies such as voice recognition and machine vision.

进一步的，所述S2中计算每一文稿段对应的视频段，具体为：Further, the video segment corresponding to each manuscript segment is calculated in the S2, specifically:

S2.1：利用语音识别技术，将视频中的语音转化为文字；S2.1: Use speech recognition technology to convert the speech in the video into text;

S2.2：利用文稿段匹配S2.1中语音识别转化得到的文字；S2.2: Use the manuscript segment to match the text converted from speech recognition in S2.1;

S2.3：当文稿段与所述文字的某段相似度达到设定的阈值时，定位到与该段文字匹配的视频段的首帧和尾帧。S2.3: When the similarity between the manuscript segment and a segment of the text reaches the set threshold, locate the first frame and the last frame of the video segment matching the segment of text.

进一步的，所述S3选取视频素材具体为：从S2.3中与文字匹配的视频段中任选一帧或多帧素材处理后作为对应文稿段的视频素材。Further, the selection of video material in S3 is specifically: selecting one or more frames of material from the video segment matching the text in S2.3 and processing it as the video material of the corresponding manuscript segment.

进一步的，所述视频素材的类型包括但不限于：图片、多字幕拼接截图、短视频、GIF等。Further, the types of the video material include but are not limited to: pictures, screenshots spliced with multiple subtitles, short videos, GIFs, etc.

进一步的，所述S3中将视频素材插入到文稿中时采用“素材优先”或“文优先”等方式，所述“素材优先”为插入视频素材时，视频素材插入的位置在对应文稿段的前面；“文优先”为插入视频素材时，视频素材插入的位置在对应文稿段的后面。Further, when inserting the video material into the document in S3, the method of "material priority" or "text priority" is adopted, and the "material priority" means that when inserting the video material, the position where the video material is inserted is at the corresponding position of the document segment. Front; "Text first" means that when inserting video material, the insertion position of the video material is behind the corresponding paragraph of the document.

进一步的，所述S4生成互联网图文稿件后，可用于预览和编辑，然后将编辑后的最终互联网图文稿件审核后发布到互联网上。Further, after the S4 generates the Internet graphic manuscript, it can be used for preview and editing, and then the edited final Internet graphic manuscript is reviewed and published on the Internet.

本发明的有益效果如下：The beneficial effects of the present invention are as follows:

1、本发明利用文稿内容和节目视频的强关联性的特性，自动提取文稿段落对应的视频画面，再通过文稿的拷贝、文稿段与视频画面匹配实现互联网图文稿件的自动生成，并且可以预览和编辑，不仅将人力资源从繁复的操作中释放出来，并且大幅提升了生产的效率。1. The present invention utilizes the characteristics of the strong correlation between the content of the manuscript and the video of the program to automatically extract the video picture corresponding to the paragraph of the manuscript, and then realize the automatic generation of the Internet graphic manuscript by copying the manuscript and matching the paragraph of the manuscript with the video picture, and can preview And editing, not only frees human resources from complicated operations, but also greatly improves production efficiency.

附图说明Description of drawings

图1是本发明的方法流程示意图。Fig. 1 is a schematic flow chart of the method of the present invention.

图2是本发明新闻文稿的示意图。Fig. 2 is a schematic diagram of a news release of the present invention.

图3是本发明提取文稿内容的示意图。Fig. 3 is a schematic diagram of document content extraction in the present invention.

图4是本发明文稿段与视频段的对应关系示意图。Fig. 4 is a schematic diagram of the corresponding relationship between document segments and video segments in the present invention.

图5是本发明方法生成的互联网图文稿件。Fig. 5 is the Internet graphic manuscript that the method of the present invention generates.

具体实施方式Detailed ways

为了本技术领域的人员更好的理解本发明，下面结合附图和以下实施例对本发明作进一步详细描述。In order for those skilled in the art to better understand the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and the following embodiments.

实施例1Example 1

如图1所示，本实施例提供一种基于文稿和新闻节目自动生成互联网图文稿件的方法，包括如下步骤：As shown in Figure 1, the present embodiment provides a kind of method based on manuscript and news program to automatically generate Internet graphic and text manuscript, comprises the following steps:

S1、提取文稿内容：读取如图2所示的新闻文稿内容，提取文稿中待生成互联网图文稿件各模块所需内容写入到图文稿件模板中的相应位置，具体为：S1. Extract the content of the manuscript: read the content of the news manuscript as shown in Figure 2, extract the required content of each module of the Internet graphic manuscript to be generated in the manuscript, and write it into the corresponding position in the graphic manuscript template, specifically:

新闻节目制作中，文稿是通过文稿系统生成的格式化内容，文稿中信息的类别很多，例如：导语、配音、正文、同期声、口播等，通常是根据新闻内容进行配置的，而标题、导语和正文是一个新闻的主体，因此如图3所示，本实施例在生成互联网图文稿件时只提取了这三个信息；In news program production, the manuscript is the formatted content generated by the manuscript system. There are many types of information in the manuscript, such as: introduction, dubbing, text, simultaneous voice, oral broadcast, etc., which are usually configured according to the news content, while the title, Lead and text are the main body of a news, so as shown in Figure 3, present embodiment has only extracted these three information when generating Internet graphic manuscript;

图文稿件模板从互联网生产系统取得，互联网图文稿件的结构组成与文稿的结构相似，两者的区别在于各个模块展现的样式不同以及图文稿件正文部分支持插入视频素材，当根据新闻标题、导语和正文各自对应的标识符识别和选取新闻标题、导语和正文后，将标题、导语和正文写入图文稿件模板的相应位置，能够过滤掉采访内容等同期声说明信息，使互联网图文稿件的主体正文内容连贯，并且当自动生成图文稿件后可以根据内容对图文稿件模板进行更换；The graphic manuscript template is obtained from the Internet production system. The structure of the Internet graphic manuscript is similar to the structure of the manuscript. The difference between the two is that the styles displayed by each module are different and the text of the graphic manuscript supports the insertion of video materials. When according to the news title, The corresponding identifiers of the lead and the text identify and select the news title, lead and text, and then write the title, lead and text into the corresponding positions of the graphic manuscript template, which can filter out the synchronous explanatory information such as the interview content, and make the Internet graphic The content of the main text of the manuscript is coherent, and the template of the manuscript can be replaced according to the content after the manuscript is automatically generated;

S2、计算文稿对应视频段：如图4所示根据文稿分段情况，结合节目视频，计算每一文稿段对应的视频段，所采用的方法包括但不限于语音识别、机器视觉等智能识别技术，具体为：S2. Calculate the video segment corresponding to the manuscript: as shown in Figure 4, according to the segmentation of the manuscript, combined with the program video, calculate the video segment corresponding to each manuscript segment. The methods used include but are not limited to voice recognition, machine vision and other intelligent recognition technologies ,Specifically:

S2.1：利用语音识别技术，将视频中的语音转化为文字；S2.1: Use speech recognition technology to convert the speech in the video into text;

S2.2：利用文稿段匹配S2.1中语音识别转化得到的文字；S2.2: Use the manuscript segment to match the text converted from speech recognition in S2.1;

S2.3：当文稿段与所述文字的某段相似度达到设定的阈值时，定位到与该段文字匹配的视频段的首帧和尾帧；S2.3: When the similarity between the manuscript segment and a certain segment of the text reaches the set threshold, locate the first frame and the last frame of the video segment matching the segment of text;

S3、选取视频素材：从每一视频段中选取关键帧或关键帧序列加工作为对应文稿段视频素材插入到文稿中，具体是从S2.3中与文字匹配的视频段中任选一帧或多帧素材处理后作为对应文稿段的视频素材；本实施例中视频素材的类型包括但不限于：图片、多字幕拼接截图、短视频、GIF等；对一帧或多帧素材进行处理可以利用插件或智能处理方法，生成图片、短视频、GIF等，或采用手动截图、剪辑、编辑等形式形成视频素材，针对不同类型的素材采用现有的处理手段即可；S3. Select video material: select key frame or key frame sequence processing from each video segment and insert it into the document as the corresponding document segment video material, specifically select a frame or frame from the video segment matching the text in S2.3 After the multi-frame material is processed, it is used as the video material of the corresponding manuscript segment; the types of video material in this embodiment include but are not limited to: pictures, multi-subtitle splicing screenshots, short videos, GIF, etc.; one-frame or multi-frame material can be processed by using Plug-ins or intelligent processing methods to generate pictures, short videos, GIFs, etc., or use manual screenshots, clips, edits, etc. to form video materials, and use existing processing methods for different types of materials;

将视频素材插入到文稿中时采用“素材优先”或“文优先”等方式，所述“素材优先”为插入视频素材时，视频素材插入的位置在对应文稿段的前面；“文优先”为插入视频素材时，视频素材插入的位置在对应文稿段的后面When inserting video material into the document, adopt the methods of "material priority" or "text priority". When inserting a video material, the insertion position of the video material is behind the corresponding document segment

S4、生成如图5所示的互联网图文稿件，便可以直接审核发布到互联网上，或者生成互联网图文稿件后，经过预览和编辑进行修改加工美化后，再审核发布到互联网上。S4. Generate the Internet graphic manuscript as shown in Figure 5, which can be directly reviewed and published on the Internet, or after the Internet graphic manuscript is generated, modified, processed and beautified through preview and editing, then reviewed and published on the Internet.

本实施例的S2中，由于每一个文稿段和对应的视频段均具有强关联性，而且新闻节目中的视频片段非常具有代表性，原则上正文对应的视频片段中的任一画面或任一连续画面都能作为插图、短视频，因此，计算文稿对应的视频段以及选取视频素材的计算方法各式各样，包括但不限于利用语音识别、机器视觉等智能技术，找到文稿和画面的大致或精确的匹配关系再按需截取视频素材即可。In S2 of this embodiment, since each manuscript segment has a strong correlation with the corresponding video segment, and the video segment in the news program is very representative, in principle, any picture or any Continuous pictures can be used as illustrations and short videos. Therefore, there are various calculation methods for calculating the video segment corresponding to the manuscript and selecting the video material, including but not limited to using intelligent technologies such as speech recognition and machine vision to find the rough outline of the manuscript and the picture. Or precise matching relationship and then intercept the video material as needed.

以上所述，仅为本发明的较佳实施例，并不用以限制本发明，本发明的专利保护范围以权利要求书为准，凡是运用本发明的说明书及附图内容所作的等同结构变化，同理均应包含在本发明的保护范围内。The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention. The scope of patent protection of the present invention is subject to the claims. Any equivalent structural changes made by using the description and accompanying drawings of the present invention, All should be included in the protection scope of the present invention in the same way.

Claims (7)

1. A method for automatically generating Internet image-text manuscripts based on manuscripts and news programs is characterized by comprising the following steps:

s1, extracting manuscript contents: reading the content of the news manuscript, extracting the content required by each module of the internet picture and text manuscript to be generated in the manuscript and writing the content into the corresponding position in the picture and text manuscript template;

s2, calculating a video segment corresponding to the manuscript: according to the document segmentation condition, combining with the program video, calculating a video segment corresponding to each document segment;

s3, selecting video materials: selecting a key frame or a key frame sequence from each video segment to be processed as a corresponding manuscript segment video material to be inserted into the manuscript;

and S4, generating an internet image-text manuscript.

2. The method for automatically generating internet text based on manuscript and news program as claimed in claim 1, wherein in S1, the manuscript content is identified and selected according to the identifier corresponding to the content required by each module of the internet text to be generated.

3. The method of claim 1, wherein the method of calculating the video segment corresponding to the manuscript in S2 comprises voice recognition and machine vision intelligent recognition technology.

4. The method according to claim 3, wherein the step of S2 is to calculate a video segment corresponding to each manuscript segment, and specifically comprises:

s2.1: converting voice in the video into characters by utilizing a voice recognition technology;

s2.2: matching characters obtained by voice recognition and conversion in S2.1 by using the manuscript segment;

s2.3: and when the similarity of the text segment and a certain segment of characters reaches a set threshold value, positioning the first frame and the last frame of the video segment matched with the segment of characters.

5. The method for automatically generating internet teletext manuscripts based on the manuscript and the news program according to claim 4, wherein the step S3 of selecting the video materials specifically comprises the following steps: and selecting one or more frames of materials from the video segment matched with the characters in the S2.3, and then using the processed materials as the video materials of the corresponding manuscript segment.

6. A method for automatically generating internet teletext manuscripts based on text and news programs according to any one of claims 1-5, characterized in that the types of video material include: pictures, multiple subtitle splicing screenshots, short videos, and GIFs.

7. The method for automatically generating the internet image-text manuscript based on the manuscript and the news program according to claim 1, wherein a 'material priority' or 'text priority' mode is adopted when the video material is inserted into the manuscript in the step S3, and the 'material priority' is that when the video material is inserted, the position where the video material is inserted is in front of the corresponding manuscript segment; "text-first" is the position where the video material is inserted when the video material is inserted, behind the corresponding document segment.

Images