Detection and tracking of facial features

作者: Liyanage C De Silva , Kiyoharu Aizawa , Mitsutoshi Hatori , None

DOI: 10.1117/12.206649

关键词: Feature detection (computer vision)Feature (computer vision)Computer visionFace (geometry)GrayscaleArtificial intelligenceFrame (networking)PixelTemplate matchingComputer science

摘要: Detection and tracking of facial features without using any head mounted devices may become required in various future visual communication applications, such as teleconferencing, virtual reality etc. In this paper we propose an automatic method face feature detection a called edge pixel counting. Instead utilizing color or gray scale information the image, proposed counting utilized to estimate positions eyes, nose mouth first frame moving image sequence, variable size template. For remaining frames, is carried out alternatively deformable template matching One main advantage that it does not require condition high inter correlation around areas matching. Some experimental results are shown demonstrate effectiveness method.© (1995) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading abstract permitted personal use only.

参考文章(0)