- RaspberryPi with RaspberryPiOS installed
- BME688 Breakout Board
| Step | Command |
|---|---|
| clone BME688CheeseMeatDetector repository | git clone https://github.com/pi3g/BME688CheeseMeatDetector.git |
| download the BSEC 2.0.6.1 package from the BOSCH website (you need to accept the license) | none |
| replace the BSEC_2.0.6.1_Generic_Release_04302021 placeholder directory by the BSEC 2.0.6.1 package | sudo unzip path/to/BSEC_2.0.6.1_Generic_Release_04302021.zip -d path/to/BME688CheeseMeatDetector/bme68x-extension |
| build and install the bme68x python extension | pip3 install -e path/to/BME688CheeseMeatDetector/bme68x-extension |
| install guizero (dependency) | pip3 install guizero |
| make sure the bme688 sensor is connect (should see 0x77) | i2cdetect -y 1 |
| run get_data.py | python3 path/to/BME688CheeseMeatDetector/get_data.py |
| run gas_estimates_gui.py (optional) | python3 path/to/BME688CheeseMeatDetector/gas_estimates_gui.py |
This example is using a pretrained algorithm by pi3g to differenciate normal air, meat and cheese. The get_data.py utilises the bme68x python module to read the sensor data and calculate the probability percentages for each class. The results are written to data.json. Optionally you can run the gas_estimates_gui.py script to read data.json and display the results in a fullscreen application. Leave the application via Alt+F4. To collect data for the algorithm we put the specimens into plastic boxes with a lid. This means the algorithm works best if you put the sensor and the specimen into a box or any relativly air tight container.
For questions or bug reports send an E-Mail to nathan@pi3g.com