摘要 |
PROBLEM TO BE SOLVED: To provide a video-image extraction device that is capable of generating a video-image extracted partially from a video-image, based on the video-image contents. SOLUTION: The video-image extraction device 1 is provided with a subtitles information extracting means (video-image addition unit data generation means) 10 which divides text data of subtitles into predetermined units and extracts a CC sentence corresponded to a video-image section; a speech recognition means (video-image addition unit data generation means) 11 which divides speech data added to a video-image into predetermined units using the data as text data and generates a speech recognition sentence corresponded to the video-image section; a calculating part of a similarity to introductory sentences of other programs (means of calculating a similarity to other video-images) 14b which analyzes a predetermined amount of characteristics of a CC sentence or a speech recognition sentence and calculates a similarity between the program introductory sentences of other video-images and this amount of characteristics; and a program introductory video-image extracting part (partial video-image extracting means) 14c which selects a CC sentence or a speech recognition sentence based on the similarity and extracts a video-image of a corresponding video-image section. COPYRIGHT: (C)2008,JPO&INPIT
|