Design And Implementation Of Facial Characters Measuring System For E-learning Process

With the rapid development of multimedia technology, when there are no classrooms and no teachers, students can also use computers to study. In the process of e-learning, students can browse the courseware or watch the video. This scenario has bring a new problem on how to effectively conduct a process evaluation in study. Facial expression recognition based on facial image becomes an important component in the teaching and learning assessment. This paper makes an exploration on how to effectively get students’ facial information in the e-learning process.Using the collected students’ facial video as data source, the paper selects brow, eyes and mouth as the object of study. The paper designs and implements a facial characters measuring system for e-learning process. This system aims to identify frown, eyes closed and mouth opened in images. Then these facial characters’ frequency changes are measured and recorded per unit of time in the video. First, we use Adaboost algorithm to locate the face and use the RCPR facial shape detection technology to get facial landmarks. Then Gabor texture features of the facial organs are extracted and combine with geometrical features as cascading feature. Subsequently, using SVM classification identifies frown, eyes closed and mouth opened in each image. Finally, the facial characters of the video frames are associated with timing information to record video facial characteristics curves for the e-learning process.This system tests videos of students. The result shows the system can effectively capture the frequency of frowning, eyes closed and mouth opened. It meets the requirements of using as an application for evaluation of teaching process. The workof this paper gives an effective support for evaluating of teaching process.