An automated classroom attendance system using video based face recognition

This paper proposes and compares the methodologies for an automated attendance system using video-based face recognition. Here input to the system is a video and output is an excel sheet with attendance of the students in the video. Automated attendance system can be implemented using various techniques of biometrics. Face recognition is one of them which does not involve human intervention. In this paper, attendance is registered from a video of students of a class by first performing Face Detection which separates faces from non-faces, and then Face Recognition is carried out which finds the match of the detected face from the face database (collection of student's name and images). If it is a valid match then attendance is registered to an excel sheet. Face recognition is performed and compared on the basis of the accuracy of recognition using Principle Component Analysis (PCA) and Linear Discriminant Analysis (LDA) algorithms.