Dynamic Facial Expression Feature Extraction and Classification Based on Candide-3 Face Model

Face recognition involves artificial intelligence, pattern recognition, image processing, psychology and other fields. It is a critical research topic in human interaction. The facial expression recognition based on dynamic image sequence is becoming more attractive with the improvement of the computer speed. It is different between the facial expression recognition based on the dynamic image sequence and the facial expression recognition based on static image. More dynamic and static feature information is involved in the dynamic features. Expression feature extraction describes by static information and the changed trend of expression due to the collection of multiple images. This paper provides the dynamic feature extraction algorithm based on Candide-3 face model parameters to extract the dynamic feature with matching facial model.