Dimensionality Reduction using PCA for Lecture Attendance Management System

EasyChair Preprint no. 4386

Abstract

Student class attendance record plays an important role in the universities to keep track of students and to maximize their academic performance. Keeping this in mind, an efficient attendance management system is proposed in this paper. This system makes use of Dimensionality reduction technique using PCA based Face recognition algorithm incorporating Euclidean Distance as the Distance Classifier. Faces of the students are segmented using Viola-Jones algorithm. Based on the recognized faces, Attendance is updated into an Excel database. The algorithm has been tested using two different student databases. Training database is created using 50 subjects including male and female with different facial variations (15 instances per subject). Statistical data shows that an accuracy of the algorithm is greater than 99% with normal lighting conditions.

Keyphrases: Biometrics, Eigen face, Principal Component Analysis, reduction rate, Viola-Jones algorithm

@Booklet{EasyChair:4386,
  author = {Ramaprasad Poojary and Mariyam Milofa and K. Shruthi},
  title = {Dimensionality Reduction using PCA for Lecture Attendance Management System},
  howpublished = {EasyChair Preprint no. 4386},

  year = {EasyChair, 2020}}