Facial feature point tracking method based on human eye preferential fitting

The invention provides a facial feature point tracking method based on human eye preferential fitting. The facial feature point tracking method based on the human eye preferential fitting is characterized by comprising the steps of estimating and updating face attitudes, detecting and tracking human eyes, and performing human eye preferential fitting and rest face part fitting based on a reverse synthesis algorithm. The facial feature point tracking method based on the human eye preferential fitting is used for solving the problems of facial feature point tracking loss, low tracking accuracy and the like due to large-amplitude deflection of face attitudes and large deviation of the initial position of a face model and a target position during facial feature point tracking, and therefore, the accuracy, instantaneity and robustness of video facial feature point tracking are improved.