Automatic Face Tracking and Model Match-Move in Video Sequence using 3D Face Model

摘要: A stand-in is a common technique for movies and TV programs in foreign languages. The current that only substitutes the voice channel results awkward matching to mouth motion. Videophone with auto-automatic translation are expected be widely used near future, which may face same problem with-without lip-synchronized speaking image translation. In out this paper, we propose method track motion of from video image, one key technologies Almost all old tracking algorithms aim detect feature points face. However, these had problems, such as blurring point between frames occlusion hidden by rotation head, so on. detects movement head given three dimensional shape face, template using 3D personal wire-frame model. evaluation experiments carried frame measured reference data head. proposed achieves 0.48 angle error average. This posed result confirms effectiveness method.