Sign language translation system and its control method that translates voice of video into 2D and 3D avatar and animation

摘要

An object of the present invention is to allow a hearing impaired person to easily and clearly recognize a conversation of an image. That is, the present invention provides a voice extracting unit 100 for separating and extracting a voice signal from an image, a text converting unit 300 for converting the voice data extracted from the voice extracting unit 100 into text, and the text converting unit 300 ), a text refining unit 400 that corrects according to the spelling of the text data generated by the text refining unit 400, a sign language translation unit 500 that translates and generates the text refined in the text refining unit 400 into sign language, and the sign language translation unit 500 ) A sign language animation conversion unit 600 that converts the generated sign language into animation, and a sign language image mixing subsign image mixing unit 700 that mixes the animation generated by the sign language animation conversion unit 600 to one side of the image In addition, as a sign language translation system configured with a speaker classification unit 220 that classifies speakers of voice data, it provides the hearing impaired by translating video conversations into sign language rather than existing subtitles, so that hearing impaired people can easily hear the video And it relates to a sign language translation system that translates the audio of an image with a clearly recognizable effect into 2D and 3D avatars and animations.