Attendance Marking System Using Face Recognition

Abstract

Background/objectives: To propose a suitable method for marking attendance automatically. Automating the attendance is necessary these days because many of the universities/institutions are particular about their student's attendance in the classroom. But the usual method of marking attendance is proven to be tedious, inaccurate and time-consuming when the number of students in a class is more. Proxy attendance is another big issue for the teachers. Methods: This study proposes a method for marking attendance automatically using Principle Component Analysis (PCA) as a face recognition technique. The image of all the students seated in the classroom is captured and then compared with the student's database, based on which attendance of individual students is marked. An e-mail is sent to those parents whose wards were not present in the class. Findings: The students are successfully detected using Viola Jones Algorithm and recognized/identified using PCA analysis method. The database constantly updates the status of attendance of every student. This system works well even in different light settings. It can identify a person even if some of his/her features are changed. For example, it could identify a person with different haircuts, a person with or without spectacles, a reasonably old photo vs new photo etc. Novelty/improvements: Thus, the classroom time can be effectively utilized without compromising attendance. Viola Jones algorithm accurately detects faces. PCA analysis algorithm utilizes the detected faces to recognize the students by comparing it with the database.