نظام ذكي لاكتشاف سقوط الأشخاص
Human Fall Detection System using YOLO and Pose Estimation
An AI-powered fall detection system that uses YOLO pose estimation and Supervision (sv) for real-time human posture analysis. The system detects potential falls by analyzing human keypoints from video streams and highlights them with visual alerts.
Project Overview
This project implements an intelligent fall detection system using YOLO-based pose estimation to monitor human movement in real time. By analyzing keypoint coordinates such as ankles, knees, and torso, the system determines whether a person has fallen based on vertical position relationships between body joints.
Key Features
Real-Time Pose Estimation
- 🧍 Detects human pose keypoints using YOLO pose models.
- ⚡ Works efficiently on live video streams.
Keypoint Analysis Algorithm
- 📉 Analyzes vertical distances between ankles and body joints.
- 🧠 Identifies abnormal or fallen postures.
Custom Sensitivity Control
- ⚙️ Adjustable detection threshold via config.yaml.
- 🎯 Easy tuning for different environments.
Live Visualization & Alerts
- 🦴 Draws skeleton edges and keypoints using Supervision.
- 🚨 Displays red alert box with “Fall Detected” label.
Technology Stack
- 🐍 Programming Language: Python 3.x
- 🧠 Pose Estimation: Ultralytics YOLO
- 🎥 Computer Vision: OpenCV
- 🧩 Visualization: Supervision (Edge & Vertex Annotators)
- ⚙️ Configuration: YAML (config.yaml)
Use Cases
- 🏥 Elderly Care Monitoring – Automatic fall detection.
- 🧑⚕️ Patient Safety – Instant alerts for healthcare staff.
- 🏠 Smart Home Safety – Senior monitoring systems.
- 🏭 Workplace Safety – Detect accidents in industrial sites.
Conclusion
The Fall Detection System using YOLO and Pose Estimation provides a robust, real-time AI solution for human safety monitoring. By combining deep learning pose detection with analytical keypoint logic, the system accurately detects falls and enhances safety in healthcare, homes, and workplaces.