Smart Surveillance System Using RESNET-50 and MTCNN

In recent years, advances in computer vision and artificial intelligence have led to the development of sophisticated surveillance systems capable of tracking and identifying individuals in a variety of situations. This research presents a new intelligent surveillance system designed to track multiple people in a video and generate a comprehensive log file to keep records of identified people. The proposed system integrates state-of-the-art techniques in face detection and recognition to achieve accurate and efficient identification of people in video streams. The system uses a custom dataset collected using a script, which captures images of individuals' faces in various conditions and environments, and fine-tunes a pre-trained ResNet50 model for face recognition tasks.In addition, face detection is performed using the MTCNN (Multi-Task Cascade Convolutional Neural Network) algorithm, which ensures robust face detection under various conditions. The intelligent tracking system works by analyzing each frame of the input video, detecting faces using the MTCNN algorithm, and then identifying individuals using a trained face recognition model. Identified individuals are logged with a time stamp, providing a comprehensive record of their presence in the surveillance area over time