Optical flow-based facial feature tracking using prior measurement

Cognitive informatics (CI) is a research area including some interdisciplinary topics. Visual tracking is an important research topic in computer vision and face expression recognition, in which domain-oriented facial features tracking is a very hot spot. In this paper, a robust facial feature tracking method is proposed. It takes Lucas-Kanade-Tomasi (KLT) optical flow as basis, and corrects the predictions by prior measurement which consists of pupils detecting, feature restricting and errors accumulating. Simulation experiment results show that the proposed method has better performance than the traditional optical flow tracking.