Student’s Identity Positioning And Recognition In The Classroom

With the rapid development of information technology, information technology has played a great role in promoting the modernization of education. In the teac hing process, the mastery of student status information is the first step of the tea chers to understand students’ learning status. Only effectively grasp each student’s learning state, teachers can give timely and effective instruction. In recent years, t he convenience and non-intrusive face detection and recognition technology has be come a hot research topic, which has been successfully applied in some fields su ch as video surveillance. However the face image resolution in classroom is low, only through the face recognition technology is difficult to achieve precise positio ning of the student’s identity. Therefore, in the case of the student’s clothes not c hange, this paper puts forward garment feature matching and face recognition tech nology combined to achieve the identity of students in the classroom. The camera capture the student’s face images in advance at the entrance of the classroom, th en get the student identity through face recognition and generate the class list of personnel; At the same time, the characteristics of the students’clothing were extr acted and the students’identity and the identity of the simultaneous extraction of clothing characteristics of the students and realize the student identity association. The character of the student’s clothing in the classroom scene image is extracted, and through the garment feature matching to determine the initial list of candidate s in the list of persons, then in the candidate list accurately locate the student ide ntity by face template matching. This method improve the recognition accuracy ef fectively and reduce the performance requirements of the camera.In order to solve the problem of student’s identity positioning, the main work of this paper includes:(1) Face detection based on AdaBoost algorithm is used to solve the face det ection in classroom. Based on the classroom scene, the face detector in the existi ng OpenCV library is not good, this paper intends to create a big scene under di fferent direction of face image database, and according to the frontal face and sid e face classification training face detector.(2) Improved the CamShift algorithm to track student’s face in the classroom scene. In the classroom scene, traditional CamShift algorithm face tracking is easy to be lost, this paper through skin color detection by the training skin color mo del to determine whether the target tracking is lost, and detect again in the local scope to lose face and tracking. This method ensure the continuous monitoring of the target in the whole teaching process.(3) Put forward a way based on a clothing feature and face recognition techn ology combination to realize the identity of the students. A student identity recog nition based on the characteristics of auxiliary clothing is proposed. First of all, b ased on the face location of tracking predict the clothing area under the face. The n, using the skin color detection and GrabCut segmentation algorithm realize to th e segment the clothing image of students automatically. Finally, extracting matchin g clothing characteristic information and combined with the face template matchin g to achieve precise positioning of the students identity.This study solved the problem of the student’s identity positioning in the clas sroom. First, face detection and tracking in the scene of the classroom; then using motion detection, skin color detection and segmentation algorithm realized autom atic clothing segmentation of the students image, and according to the face templa te matching for precise positioning of the target identity. Good recognition rate is obtained by using this method in the test data, and the effectiveness of the metho d is verified.