Automated Attendance Using Face Recognition
Kapil Tajane, Vinit Hande, Rohan Nagapure, Rohan Patil, Rushabh Porwal · 2023
The term automation describes the use of a wide range of technology that is used in the creation of technologies to provide efficient, reliable and fast output without the intervention of human or human activities. Attendance management is one of such traditional processes which we propose to automate with this face recognition system. Thus, a Face Recognition based Attendance Monitoring System based on In-Out timestamps of a person is proposed. This face recognition system works at four major layers. First, images captured by cameras are passed through Haar-cascade based different classifiers like face, eyes, mouth classifier to detect facial regions in the image. Second, the facial features are enhanced using contrast adjustment, filtering images, and removing unnecessary features. This step increases the efficiency of the model. Third, these images are trained with a FaceNet model for face recognition. Then, this trained model is used to identify faces. Fourth, in-out timestamps of the identified person are noted. And, at the end, attendance is marked based on in-out timing of the person. Using the FaceNet model gives up an accuracy of 99.38%. And preprocessing of images improves features to be extracted. Overall, it gives automated, easy-to-use, high accuracy face recognition-based Attendance Monitoring System.