real-time-person-elderly-fall-detection-system

A computer vision system that detects falls in real time from camera video, aimed at elderly care. It tries to avoid false alarms from sitting or sleeping and to cope with people hidden behind furniture.

Share on XNo license declared

Overview

This is a computer vision system that detects falls in real time from camera video, aimed at elderly care. It is designed to keep working when a person is partly hidden behind furniture such as tables, beds or sofas, and to avoid false alarms from everyday actions like sitting down, tying shoelaces or sleeping in bed. It uses a lightweight pose estimation model and runs on a standard CPU.

Key features

  • Triggers on rapid descent or sudden posture collapse
  • Tells resting or sleeping in bed apart from a fall
  • Four measures to cope with occlusion, including furniture tracking
  • Runs on a regular CPU with a model of about 6.5 MB
  • Interactive launcher that finds videos and supports drag and drop

Best for

Developers and students exploring fall detection from video, especially where furniture blocks the view. The README makes no claim about clinical or production validation, so treat it as a research project.

Upstream
AwaisShah75/Real-Time-Person-Elderly-Fall-Detection-System
Fork on GitHub
Guo-astro/real-time-person-elderly-fall-detection-system
Upstream stars
261
Category
Finance, data and research
Language
Python
License
No license declaredWithout a license, the author keeps all rights. Ask the upstream owner before reusing the code.
Forked
2026-08-16
Sync status
In syncLast synced 2026-09-29