AI Activity Tracker Office

This paper presents a real-time work monitoring system using YOLOv8 and OpenCV for detecting and tracking individuals in a workspace. The system identifies people and objects such as desks, laptops, and computers to determine worker activity status. Optical flow-based video stabilization is implemented to enhance detection accuracy. The model processes live video feeds, identifies workers, and tracks their proximity to desks and devices. If a worker is away from their workstation for an extended period, the system records the absence in a log file. The proposed approach combines object detection with time-based tracking, enabling automated attendance monitoring and workplace productivity assessment. The system is designed to provide real-time insights for employers and researchers studying workspace efficiency. Results demonstrate its effectiveness in tracking personnel movement and detecting absence duration with reasonable accuracy.