Intelligent Attendance System using Machine Learning for Face Recognition

Traditional attendance methods face several challenges, like time-consuming recording, the risk of forgery, and difficulties in maintaining and organizing error-prone paperbased records. Traditional attendance methods often face several challenges that can be effectively addressed by real-time digital attendance systems. This paper describes the creation of an attendance system based on image processing techniques. We employ machine learning algorithm utilizing Python as our development platform to create an application. The attendance is taken in real- time through a live-feed from a camera. The identification process involves detecting faces in real-time using the Haar-Cascade machine learning algorithm. It creates intermediate images of detected faces for recognition. These images consist of data about pixel intensity that aids in edge detection. The processed images are subsequently matched against a database of recognized individuals. If a match is found, the system logs the individual’s identity and timestamp. The application provides an interactive interface to register the user and initiate attendance. The tool captures a group of images to be stored for comparing. An Excel file, consisting of identity and timestamp, will be generated. This file will then be mailed to the educator. This allows to increase the efficiency of attendance-taking process by reducing manual efforts.