Automated Virtual Attendance Using Image Recognition Techniques (Review Paper)

Abstract

Conducting routine attendance is a very essential and obligatory task for smooth functioning of day-to-day administration. It may normally become a laborious and redundant activity, even pushing itself to inaccuracies. The rather old and traditional approach of making roll calls to determine whether the student is present or not proves itself to be a statute of many limitations since it is very strenuous to call names and maintain the records especially when the ratio of students to faculty is not good. Every organisation has its way of ensuring efficient measures for the attendance of present students to confirm highest accuracy and utmost precision. Some organisations use a document-oriented approach whereas others have implemented many digital and virtual methods such as biometric fingerprinting techniques, face identification techniques, card swapping techniques, etc.

However, these methods prove to be a hindrance as it subjects the attendees to wait in a time-consuming and tardy queue. There are multiple situations of adversity which may arise such as, consider when the student fails to bring his identity card, then he will not be able to mark his attendance, even though he is physically present. In this model, we have proposed a framework which aims at automating the process of conducting attendance of students.

In this paper we will be using an SVM classifier for taking attendance of a large audience and will be evaluating its performance.