Surveillance System Based on Face Recognition Utilizing Deep Learning Technique

As the population continues to grow, ensuring the safety of individuals and places from theft or trespassing has become increasingly challenging. Traditional surveillance systems are no longer sufficient for modern environments, which often rely on operators for monitoring. To address this gap, the combination of deep learning and surveillance systems has become an essential tool for intelligent surveillance in various fields. This research introduces a single-shot YOLO algorithm, which enhances surveillance systems by utilizing face recognition capabilities based on the YOLOv8n algorithm. Our approach aims to develop a face recognition system that effectively distinguishes between authorized and unauthorized individuals. Such a system is essential because it can immediately notify nearby individuals or authorities, potentially saving lives and property. The system achieves an impressive mAP of 99.5% and has been tested under different lighting conditions to demonstrate its detection and recognition efficiency with 4 FPS. This technology is expected to be implemented in various facilities to meet the growing need for advanced visual surveillance capabilities in today’s dynamic landscape.