Smart Surveillance: Real-Time Object Detection and Recognition with Interactive Polygon Region Using Deep Learning

The Traditional Surveillance system lacks the capability to focus on particular region at a time during the surveillance. This cause important detail in that region to be missed out and go unnoticed. This is a major issue as it causes security concern. Detection of any objects in that region should be done in real-time. The proposed system allows the user to create a polygon region in area of interest in the video and by using the capability of YOLOv8, which stands for You Only Look Once Version 8 this system can detect the object and classify them and then converting to csv file. By allowing the users to create the polygon region it gives the surveillance camera capability of highly precise monitoring. This system helps to monitor the regions that changes dynamically. Changing Traditional surveillance system as a whole can be expensive and time consuming. Our system helps to bring a cost-effective solution without the need of whole changing whole system. It provides extra security to current infrastructure without the need for a complete overhaul, making it a good choice for various security setups.