Crowd counting using yolov8 and various tracking algorithms

In today’s urban landscapes, effective crowd monitoring is paramount for ensuring public safety, security, and streamlined urban planning. This research paper tackles the challenges associated with precise crowd counting and optimal tracking methodologies, aiming to enhance accuracy and efficiency. It proposes a holistic approach that integrates crowd counting with state-of-the-art person detection through YOLOv8, complemented by tracking algorithms such as DeepSORT, StrongSORT, ByteTrack, and BotSort to count the number of people in crowd. Through a comprehensive investigation, the study explores a myriad of deep learning techniques tailored for real-time crowd counting and individual tracking within dynamic crowd scenarios captured across various camera frames in video footage. Furthermore, the paper delves into the integration of these results to create a unified framework for efficient tracking within a crowd. The research aims to significantly enhance the capabilities of crowd monitoring systems, enabling applications in various domains, such as public safety, security, and urban planning.