Research And Implementation Of Mobile Roll Call System Based On Face Detection

The roll call is one of the essential means of teaching management in universities. Follow the progress and development of science and technology, many universities use new way or system to call the roll. A school in Florida scan the fingerprints of students to call the roll, a community college in Erie use face recognition to call the roll by using infrared light to scan faces of students. In the domestic, some school call the roll by combining RFID(radio frequency identification) and smart card, some school call the roll by using gsm modem to receive the messages from students, and so on. These roll call systems are using additional equipment, not only cost too high that hard to spread the system, but also the operation is very inconvenient.To solve the above problem, this thesis found that take the picture of the whole class, then face detect the picture, is more suitable for roll call. The main contents are as follows:1) First introduced the research status of roll call system at home and abroad, and analyzed the problems of these roll call systems.2) Then to solve these problems, this thesis propose a solution that using a smart phone to take the picture of the whole class, then the system will face detect the picture, the faces of students will be "cut out", students choose their own face to confirm their attendances.3) Then in order to improve the face detection results, the variety of image processing technologies such as Gray World color equalization, color segmentation, histogram equalization, image stitching applied to the smart phone, and optimized and improved these technologies according to features of smart phones.4) Finally, this thesis achieved a prototype roll call system based on face detection, teachers and students through a few clicks on the phone to complete the roll call, low cost, and easy to operate.