An Intelligence Smart Attendance Recognition System using HOG with SVM Techniques

Maintaining attendance records in educational and organizational settings is critical for tracking participation and guaranteeing responsibility. This paper describes a full smart attendance system that integrates computer vision techniques such as Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) classifiers alongside real-time facial recognition and seamless cloud-based data storage via Google Firebase. The system collects live images using OpenCV, employs HOG and SVM algorithms to extract facial features and classify faces, and then compares them to existing database entries to ensure proper identification. The interface with Google Firebase, a NoSQL cloud database, enables real-time synchronization and access to attendance records across several devices and platforms. Facial encodings are saved in a pickle file for easy recognition during future attendance markings. The Date Time library uses timestamps and date tracking to ensure accurate record-keeping by connecting each attendance item to the corresponding date and time. The user-friendly interface, created with the CV zone library, offers an interactive experience by displaying live camera feed, recognized faces, and real-time attendance status. The system design, comprehensive methodology, implementation methods, and performance evaluation metrics are described, emphasizing its accuracy, precision, and scalability. The database design and organization of Google Firebase, and the stability and scalability of the system were examined. The proposed smart attendance system not only automates the attendance marking process but also helps administrators and users have a more secure, efficient, and user-friendly experience.