ANALYSIS METHOD FOR CHUNK AND KEY WORD BASED ON VOICE SIGNAL OF VIDEO DATA AND SYSTEM THEREOF

摘要

The present invention relates to a method for analyzing a chunk and a keyword based on a voice signal of video data and a system thereof. The method comprises: a first step of extracting, by a chunk/keyword realization application (100), a chunk unit based on a preset duration time in which a level of the note division intensity lasts higher than or equal to a preset intensity level among the note division intensities by performing an analysis on note division intensity of recognized English-based voice information after recognizing the English-based voice information in video data including the English-based voice information stored in a database (140) or input from the outside; a second step of analyzing a syllable and a phoneme on each chunk unit by the chunk/keyword realization application (100); and a third step of analyzing, by the chunk/keyword realization application (100), a keyword by extracting a voice signal with threshold voice volume information or higher within the chunk unit through noise filtering by a noise gate algorithm (130a). Accordingly, characteristics of an English voice signal varied by various environments can be accurately recognized through a common analysis method.