Principal Component Analysis Algorithm For Face Recognition in Kindergarten Students

Abstract

In kindergarten, understanding each student's activities is crucial for evaluating their learning and adaptation to the school environment. However, manually tracking individual student activities during class is challenging for kindergarten teachers. This paper proposes using face recognition for kindergarten students as a preliminary step to monitor and record their activities. The process involves converting video footage of students into digital images. Faces are detected using the Viola-Jones method, and feature extraction on the images is performed using the Principal Component Analysis (PCA) method. Euclidean Distance is then applied to recognize the students' faces. Our experiments utilize 70 images for training data, consisting of 5 different images from each of the 14 students. The experimental results demonstrate an accuracy of 91.42% when testing 14 new images of the students