Real-Time Surveillance for Smart Home Security

In today's rapidly evolving digital environment, securing personal and organizational spaces has become more challenging, especially when traditional methods fail to identify intrusions. This project presents a Flask-based Face Recognition and Alert System designed to enhance security through intelligent face detection and real-time alerts. The system captures live camera feeds using OpenCV, performs facial recognition, and identifies whether the detected face is part of the authorized dataset. If an unauthorized face is identified, the system instantly sends alerts via WhatsApp and Gmail, along with an image of the unknown individual. The admin can review the received image and, if the person is trusted, train the model with the new face through the web interface. This helps the model improve over time and minimize false alerts. The system seamlessly integrates image processing, machine learning (face encoding and recognition), and real-time notification mechanisms, thereby creating a robust and adaptive security platform. It is particularly suitable for smart homes, office environments, and areas that require controlled access, without relying on physical keys or cards.