Image and Artificial Intelligence based Person Identification for Automation of Attendance Extraction for Classrooms

Currently, precious time of a teacher is wasted on taking a student's attendance manually. In some cases, students stand in queues to give their biometrics before entering the class through some biometric attendance systems. In both cases, precious time is wasted, and the result is not fruitful. Hence, an automated attendance system that uses image and artificial intelligence is proposed, which helps to solve all these current-day problems and allows institutions to take classes seamlessly. The proposed system concentrates on extracting multiple real-time face recognition without human intrusion through a surveillance camera. The traditional methods produce lots of false recognition due to the usage of Edge detection for face detection and pixel matching for recognition. The proposed methods combined the Viola-Jones algorithm for face detection with convolution neural network (CNN). CNN is used to refine the detected faces which results in removing false positives identification. This scheme produces more accurate results compared to the traditional methods and is quicker than an artificial neural network. In addition, Open CV, Emgu CV, Harcascade classifier, and principal component analysis with the Eigenface algorithm can improve the system's efficiency.