REAL-TIME DRIVER MENTAL STATE ESTIMATION BASED ON VISION INFORMATION

摘要

The driver mental state estimating research with vision information is described in this paper. Face detection is realized by FFCs (Face Feature Classifiers) with the method of machine learning first, and with the information of confidence coefficient, face tracking is realized. Thus the position information of face and eyes are obtained. For the face gesture parameter recognition, an algorithm of dimension reduction was introduced. A vector is formed to express the opening degree of eyes. Based on these, driver's mental sate can estimate efficiently.