International Journal For Multidisciplinary Research

E-ISSN: 2582-2160   •   Impact Factor: 9.24

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Call for Paper Volume 8, Issue 6 (November-December 2026) Submit your research before last 3 days of December to publish your research paper in the issue of November-December.

AI Based Real Time Crowd Monitoring and Risk Detection Using Computer Vision

Author(s) Ms. Thanisha G, Ms. K Jyothi, Ms. Rachana M, Ms. Gaanashree B N, Prof. Raghavendra S
Country India
Abstract Monitoring crowd conditions is a challenge with just CCTV monitors. Whilst the cameras can capture footage this requires monitoring to identify when the platform is getting busy or when people are rushing. An automatic monitoring system has been created that detects people in video footage from CCTV or IP cameras. The model uses YOLOv8 to detect people and ByteTrack to track people between frames. The scene has been split into zones so crowd monitoring can be done to see how many people are in each zone, how many people are coming into and out of the zone and how fast and direction people are moving. A YOLOv8 model has been fine-tuned with the D-Fire dataset to detect fire and smoke. An LSTM Autoencoder has been used to detect movement. The fusion stage will classify each zone as Normal, Caution or Critical. Debouncing has been used to stop multiple alerts being sent. All the information will be displayed on a dashboard created with Flask and all events will be stored in MongoDB.
Keywords Artificial Intelligence and Computer Vision are important. Crowd Monitoring is vital. Risk Detection is key. YOLO.
Field Engineering
Published In Volume 8, Issue 6, November-December 2026
Published On 2026-11-15
DOI https://doi.org/10.36948/ijfmr.2026.v08i06.92168

Share this