Sistem Absensi Berbasiskan Pengenalan Wajah Secara Realtime Menggunakan Webcam Dengan Metode PCA

Face can be considered as part of biometric system which is used in computing processes for tracking, recognition, expression and pose estimation in computer vision. Face recognition has become an attraction and a lot of dedicated research in this field has been done since a decade ago. This is mainly because face recognition does not require direct physical contact interaction between human and computer input sensor, problems in face recognition system are quite dynamic and complex. This is caused by several factors, including: indoor lighting conditions, angle of view of the camera, the speed of head movement and facial expression changes. Several methods for face recognition, for example: Gabor filter, PCA, LDA, subspace LDA, Neural Network and Fuzzy. Differences of these methods depend on the end result that is desired, such as computing speed, accuracy and durability. This study applied the PCA method that is commonly referred to as the application of the Eigenface. The PCA method was chosen due to that realtime attendance system required simplicity, speed and accuracy. The results showed that realtime facial recognition to the input face examined in ten different actors have avarage confidence values 0.97 with an average accuracy of 97.67%. The speed of detection for each image is 0.25 ms