Table of Contents
- Title: Unobstructive Monitoring of Vulnerable Elderly
- Problem Statement: To mitigate physical and mental health risks faced by the elderly through remote monitoring that preserves the privacy of vulnerable elderly individuals
- Project Goal: To develop a non-invasive home-monitoring system for vulnerable elderly that provides unobtrusive fall detection and activity level tracking that preserves individual privacy and can be readily integrated into existing infrastructure
This repository contains the code for testing sensors and data analysis for the activity level monitoring subsystem. For the fall detection subsystem, refer here
Presence Detection (from Review 1)
Activity Level Analysis (from Review 2)
Background Subtraction Pipeline (Review 3)
Demo
Credits to Andrews Sobral for a GoDec Python Implementation and OpenCV for computer vision API. We experimented a lot with the different functions and parameters available before achieving this r
The Activity Levels Monitoring subsystem (as of Review 4) require the following devices:
- MLX90640
- RPI 3b
- NUC / Central Processor
For instructions on older sensors that we have experimented with, refer here.
- For general testing with the MLX90640, refer to MLX_SETUP.md
- For setting up the RPI, refer to RPI_SETUP.md
- For setting up the NUC, refer to NUC_SETUP.md