Live Video Face Recognizing Smart Attendance System Using Harr Cascade Algorithm

Live Video Face Recognizing Smart Attendance System Using Harr Cascade Algorithm

Prof. Anand Dhawale1, Ayush Yeole2 , Shivani Tuplondhe3, Vidya Take4

1Department of Computer Engineering, MES Wadia COE, SPPU, Pune-411001

2Department of Computer Engineering, MES Wadia COE, SPPU, Pune-411001

3Department of Computer Engineering, MES Wadia COE, SPPU, Pune-411001

4Department of Computer Engineering, MES Wadia COE, SPPU, Pune-411001

Abstract - Face recognition is a significant image processing application widely used for authentication, particularly in student attendance systems. Traditional attendance methods are time-consuming and prone to manipulation. The proposed system automates attendance tracking using face recognition technology, employing machine learning techniques such as Haar Cascade classifiers, KNN, CNN, SVM, and LBPH. The Haar Cascade algorithm, based on the Viola-Jones method, is used for real-time face detection, leveraging Haar-like features and the AdaBoost algorithm for efficiency. The system ensures high accuracy by testing under various conditions, including illumination changes, head movements, and distance variations. Attendance reports are stored digitally in an Excel or cloud database, making the system cost-effective, secure, and resistant to proxy attendance.

Key Words:  Haar Cascade, Face Detection, Face Recognition, AdaBoost, Machine Learning, Viola-Jones Algorithm, Open CV, Attendance System, Computer Vision, model training, image reading.