In this paper we will represent the vision-based methods for automatic detection of face feature and tracking their movements across image sequences. The developed component for the Analysis of Facial Mimic Image (AFMI) is involved in a whole system aimed for the monitoring of face mimics. The goal of our research is to develop a system for face mimic analysis and synthesis that can be used in areal-time under a variety of conditions and in different scaled applications. Therefore the development of stable and time efficient methods of image processing and analysis is of significant importance in the whole framework . The main tasks of this analysis component AFMI is the automatic detection of face features, tracking of their displacements and classification.