摘要 |
PURPOSE: A device for recognizing a voice by extracting 3D lip feature points based on a stereographic camera and a method thereof are provided to extract feature points from a lip area in real time, thereby accurately recognizing a desired voice using the extracted feature points. CONSTITUTION: A voice recognizing device includes a stereographic image obtaining unit(12) and an image processing unit(15). The image processing unit receives a stereographic image of a speaker from the image obtaining unit. The image processing unit extracts a lip area of the speaker from the received stereographic image. The image processing unit classifies a phoneme corresponding to a 3D coordinate value of a feature point of the lip area. The video color deflection recognizes the voice of speaker. |