Non-Contact Attendance Monitoring System for Chartered Institutions Using Face Recognition

Resumen

The study develops a non-contact attendance system for government employees using face recognition. This is an alternative solution to fingerprint scanning following health-related protocols for COVID 19. Computer Vision's Face recognition biometric system is being used as a way of attending the marking system. The system uses python programming language and integrates cascade classifiers from OpenCV library. LBPH algorithm is applied for face recognition. Employees will face the camera, capture the image, and record the CSV format's time and date. PHP programming language is used to fetch the record from CSV file, transferred to the MySQL database, and generate CSC Form 48 (Daily Time Record). The system was evaluated based on the ISO 9126 standard. Analysis of result was done using mean and standard deviation. Results showed that the system is very high usability (4.27), very high functionality (4.35) and very high maintainability (4.23). The use of python programming language, openCV as classifier, LBPH algorithm as recognizer and PHP for interface in attendance monitoring is then believed as a very convenient way to capture store, train, analyze and retrieve data. Accurate detection and recognition of face corresponding to the employee’s identity was obtained using OpenCV and LBPH recognition. Thus, this study provides timely mechanisms to address both threat of COVID-19 and ensuring and monitoring employees’ attendance.