• Shital Dnyaneshwar Cheke
• Manjiri U Karande

The rapid expansion of the global population has led to an upsurge in public gatherings, giving rise to concerns about overcrowding and safety. Among these safety issues, crowd-smashing accidents have emerged as unexpected and swiftly escalating situations that pose significant risks to the general public. Manually predicting and managing such chaotic scenarios presents numerous challenges. As witnessed during the COVID-19 pandemic, maintaining physical distance is crucial in curbing the transmission of viruses from person to person. The World Health Organization (WHO) mandates limiting the number of people in a given space. To address these challenges, AI-based monitoring systems now incorporate cutting-edge object detection algorithms, with YOLOv8 being a prominent example. This advanced approach provides real-time crowd density assessments and instant alerts to authorities, enabling swift action and accident prevention. Human detection and crowd counting are fundamental tasks in computer vision, serving practical purposes such as surveillance, security, crowd management, and traffic analysis. Deep learning models, particularly the You Only Look Once (YOLO) approach, have achieved remarkable success in these domains. In our research paper, we delve into the current state of the art in human detection and crowd counting using YOLOv8, discussing both its advantages and limitations. Notably, our proposed model extends beyond crowd counting by detecting abnormal activities within the crowd, including weapons, fires, falls, and smoke. By identifying potentially hazardous crowd densities and promptly detecting abnormal incidents, our system not only prevents disasters like crowd-smashing, detecting physical distances but also strengthens overall security measures.

Yolo, Overcrowding, Convolutional Neural Network (CNN), Real- time video processing, ROI, LOI, Object Detection, Ultralytics

"Crowd Detection Using Artificial Intelligence", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 4, page no.l382-l386, April-2024, Available :http://www.jetir.org/papers/JETIR2404B52.pdf

"Crowd Detection Using Artificial Intelligence", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 4, page no. ppl382-l386, April-2024, Available at : http://www.jetir.org/papers/JETIR2404B52.pdf