Year: 2026
Type: article
Abstract: This project presents an AI-Based Human Activity Recognition for Surveillance System that automatically identifies and monitors human activities using Artificial Intelligence (AI), Machine Learning (ML), and Computer Vision techniques. Developed using Python, Flask, HTML, CSS, Javascript, SQLite, OpenCV, TensorFlow and YOLO, the system provides an interactive web interface with secure user authentication, real-time video processing, activity detection, and surveillance monitoring. The platform analyzes video frames captured from CCTV cameras or uploaded video files to recognize activities such as walking, running, standing, sitting and suspicious behavior. It uses deep learning models to classify human actions accurately and generate real-time alerts for abnormal activities. Designed for educational and research purposes, this platform demonstrates how AI-powered surveillance can enhance public safety, improve security management.
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7217833797