In this cool project we work with BGT60TR13C radar chip.
We learn how to find the difference between moving and static objects, by the data that the radar chip gives us. When the person is classified as static we make analysis over the vital functions of the person by measure breathing as well as heart rate of the person using the cheap raw data.
Clone the project from this repository
git clone https://github.com/ml-lohi/hackathon-private
cd hackathon-private
The libraries needed for the project are defined in requirements.txt
To install needed libraries in your virtual environment run the following commands:
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
Bug: unfortunately the provided library ifxdaq only works with Windows and Linux. We tried as well Ubuntu Docker, but there comes up the issue, that the docker doesn't have an easy ability to work with usb ports of the host. This is definately a limitation (in our team 2/3 are using MacOS). Moreover it appears that the library ifxdaq doesn't work with the newest version of the Python (3.10).
We implemented all the needed utilities for working with data.
Firstly we made an overlook over the data, that the radar chip delivers us. We measured for 30s three different data sets: static person, static person but further away from the radar and moving person. Our overlook can be found in data_overlook.py.
First step is to classify the objects as moving or static.
The detailed overview is provided in the notebook models.ipynb. In the notebook we measured for 100s each of the team membrs in static as well as in moving state. This delivers us 600s of total data for the analysis. In the notebook linear_discriminant_ananlysis.ipynb we can find the analysis of the data with help of linear boundry.
For the vital functinality we analyse breathing as well as heart rate. The detailed analysis is implemented in fourier.ipynb
The final state of our project was to develop an app, which can be started, when the radar is connected to the laptop. This app is implemented in main_app.py. You can start it by simply calling
python main_app.py
The applicatino plots the phases every second, updates the value of the heart rate every 5s and breathing rate every 10-15s.