Automated Attendance System Using Computer Vision

Abstract

Automated attendance system using computer vision is a model based on real-time facial recognition which works on deep learning algorithms. This automated system efficiently manages time with excellent accuracy. To reduce human effort, our model will precisely recognise the image and according to the label assigned to that particular image, the attendance will be marked accordingly. The concerned faculty can also retrieve the database of present students on a particular day. This reduces the time required for managing the manual attendance and also eliminates the chances of proxy attendance. Moreover, our model uses OpenCV and Dlib as the major libraries and Principal component analysis (PCA) algorithm. Traditional facial recognition systems only recognise one face at a time, but this paper deals with the automated attendance system which could detect multiple faces at a time and mark their presence simultaneously. In our paper, we have done multiple experiments with our model and the results are excellent with accuracy ranging from 75 to 100 %.