This project focuses on collecting and analyzing smartphone sensor data to recognize prayer postures (Salat) such as Qiyam (standing), Ruku (bowing), Sujud (prostration), and Jalsa (sitting).
The ultimate goal is to identify patterns in sensor signals that can distinguish postures and detect whether the prayer is performed slowly or quickly.
Sensor data is collected using the SensorServer Android application.
The app streams or logs raw data from the following smartphone sensors:
- Accelerometer (x, y, z)
- Gyroscope (x, y, z)
Finally the data is visualized as plots.
- Setup python venv
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
- Run and start SensorServer while keeping both devices under same network and copy the address to address.txt
- Run: node track.js and start praying while keeping phone in pocket
- ^C to stop tracking and save record