JP2004260711A - Method for creating video search database and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for creating a video search database and a recording medium. <P>SOLUTION: Image data with superimposed characters are divided into a plurality of shots first. According to video in the plurality of shots, a plurality of key frames are extracted. An image browser search is set by the plurality of key frames. Continuously, a text region is captured according to image features in the image data. A text is divided with respect to the text region, and a plurality of text features are generated. The plurality of text features are compared with a text in the database, and text data are generated. A text index table is set by the text data. Finally, the video search database is created from the image browser search and the text index table. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to the production of language teaching materials, and more particularly, to an interactive video of language teaching materials with an image browser search and a text index table.
[0002]
[Prior art]
Under the trend of internationalization, various self-learning materials for English have been released, and interactivity has been particularly emphasized. However, it is difficult to produce language teaching materials, the manpower and cost are huge, and for consumers, the interactive image language teaching materials currently on the market have limited materials, are expensive, and in terms of content, However, learning cannot be continued effectively and without getting tired.
[0003]
[Problems to be solved by the invention]
An object of the present invention is to provide a method for creating a video search database that converts foreign language videos into interactive language teaching videos.
[0004]
[Means for Solving the Problems]
In order to achieve the above object, the present invention provides a method for creating a video search database, first dividing subtitled image data into a plurality of shots. Extract multiple keyframes according to the video in multiple shots. Set image browser search with multiple keyframes. Subsequently, a text area is captured according to the image feature in the image data. Split text into text regions to generate multiple text features. A plurality of text features are compared with text in a database to generate text data. A text index table is set according to the text data. Finally, a video search database is created using an image browser search and a text index table.
[0005]
BEST MODE FOR CARRYING OUT THE INVENTION
In order to further clarify the objects, features and advantages of the present invention described above, preferred embodiments of the present invention will be described below with reference to the accompanying drawings.
[0006]
FIG. 1 is a flowchart of a method for creating a video search database according to an embodiment of the present invention, and the operation steps are as follows.
[0007]
First, in step S10, for example, foreign language videos including Chinese subtitles and English subtitles are analyzed. In step S12, shot detection is performed automatically and quickly by a computer system, and the video file is converted according to the shot switching timing. Divide into multiple shots.
[0008]
In step S14, a plurality of key frames are captured during each shot based on a change in the screen of the frame. Keyframes represent one shot. In step S16, an image browser search is set up, and a video for one hour contains thousands of shots, and a key frame is representative of one shot. Can be completed, and the user can quickly find a video location to be viewed based on the time position and the content of the key frame.
[0009]
In step S22, the video in the image data is enhanced, that is, edges in the video are enhanced, and subsequently, a text region is captured by edge detection. The edges of the image portion in the video have a large radius and are irregular, and the edges of the subtitle portion in the video have many straight lines. Therefore, the text region in the image is captured according to this image feature.
[0010]
In step S24, the text area is subjected to text segmentation, the text area is detected, and the text in the text area is divided into a plurality of text areas according to the text length, width, height, line density, and structure. Then, the color in each text area is divided into two colors of black and white using binarization. Most of the text in general video is on a complicated screen, so remove the complex background, that is, convert each text area to black on a white background, separate the text from the background, and generate text features .
[0011]
In step S26, each text feature is compared with the text in the text feature database to perform text identification, search for similar text, and create text data.
[0012]
In step S28, a text index table is set using the correspondence between the text data and the video, and a video portion to be reproduced is searched.
[0013]
In step S32, a dictionary database is created from the text data.
[0014]
In step S30, the image browser search and the text index table are saved, and a video search database is created.
[0015]
FIG. 2 is a diagram showing a video search database according to an embodiment of the present invention. In the image browser search 50 in the search database 40, the user selects an arbitrary key frame during the image browser search, and selects a video having the key frame. Simultaneously with the reproduction of the location, an arbitrary search field in the text index table 60 can be selected using the text index table 60 in the search database 40, and the video location corresponding to the search field can be reproduced. In addition, the user can search for a video to be reproduced by using the keyboard input subtitle portion as search data.
[0016]
Also, when playing a video, the dictionary database function is executed, the screen screen is divided into two windows, and a video is paused, played, and played by a toolbar 67 in a video playback area 75 used for playing the video. Fast forward and rewind functions can be operated.
[0017]
In addition, the subtitle mask 65 blocks the text area in the video, that is, blocks the subtitle in the video, and the subtitle mask 65 is switched and controlled, so that subtitle shielding or display can be selected. The subtitle in the video is displayed in another window that becomes the subtitle display 80, and the text in the subtitle display 80 is text data obtained by analyzing the subtitle in the video, and the text data in the subtitle display 80 is selected. In addition, the dictionary database 70 displays the commentary of the text including the meaning of the text, the vocabulary, the idiom, and the like, and the image display 85 displays the shot corresponding to the text in the subtitle display 80.
[0018]
The method of creating a video search database provided by the present invention allows users to effectively convert foreign language videos into interactive language teaching materials, create index data based on subtitles in the videos, and provide users with index data. Provide the best video browser control method and convenient language teaching tool.
[0019]
Although preferred embodiments of the present invention have been disclosed as described above, they are not intended to limit the present invention in any way, and any person skilled in the art may make various modifications without departing from the spirit and scope of the present invention. Variations and hydrations can be added, and the protection scope of the present invention is based on the contents specified in the claims.
[0020]
【The invention's effect】
Get interactive language video teaching materials.
[Brief description of the drawings]
FIG. 1 is a flowchart illustrating a method for creating a video search database according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating a video search database according to an embodiment of the present invention.
[Explanation of symbols]
40 search database 50 image browser search 60 text index table 65 subtitle mask 67 toolbar 70 dictionary database 75 video playback section 80 subtitle display 85 image display

Claims (16)

Dividing image data with subtitles into a plurality of shots;
Extracting a plurality of keyframes according to a change in video during the plurality of shots;
Setting an image browser search using the plurality of key frames;
Capturing a text area according to the image features in the image data;
Subjecting the text region to text segmentation to generate a plurality of text features;
Comparing the plurality of text features and text in a database to create text data;
Setting a text index table using the text data;
Creating a video search database using the image browser search and the text index table;
A method for creating a video search database, comprising: The method of claim 1, wherein the subtitles are Chinese and English subtitles. The method of claim 1, wherein the plurality of key frames are generated according to a video having a large screen change. The method of claim 1, wherein selecting an arbitrary search field in the text index table plays a video location represented by the search field. 2. The method according to claim 1, wherein a corresponding video portion is obtained by using a keyboard input device and using the input character data as search data. The method of claim 1, wherein a shot with the key frame is played by selecting an arbitrary key frame in the image index table. 2. The method according to claim 1, wherein shielding or displaying the text subtitle is selected by shielding the text area. The method according to claim 1, wherein a dictionary database is created using the text data. A recording medium for recording a program that causes a computer to execute a method of creating a video search database,
Dividing image data with subtitles into a plurality of shots;
Extracting a plurality of keyframes according to a change in video during the plurality of shots;
Setting an image browser search using the plurality of key frames;
Capturing a text area according to the image features in the image data;
Subjecting the text region to text segmentation to generate a plurality of text features;
Comparing the plurality of text features and text in a database to create text data;
Setting a text index table using the text data;
Creating a video search database using the image browser search and the text index table;
A recording medium comprising: The recording medium according to claim 9, wherein the subtitles are Chinese and English subtitles. The recording medium according to claim 9, wherein the plurality of key frames are generated according to a video having a large screen change. 10. The recording medium according to claim 9, wherein a video portion represented by the search field is reproduced by selecting an arbitrary search field in the text index table. 10. The recording medium according to claim 9, wherein a corresponding video portion is obtained by using the input character data as search data using a keyboard input device. 10. The recording medium according to claim 9, wherein a shot having the key frame is reproduced by selecting an arbitrary key frame in the image index table. The recording medium according to claim 9, wherein shielding or displaying of the text subtitle is selected by shielding the text area. The recording medium according to claim 9, wherein a dictionary database is created using the text data.

Images